Have you read Rick Cook's WIZARDS BANE and sequels?

A very long time ago. Not what I have in mind.

Well, a fair amount of the first two books dealt with making something type-safe with tons of built-in error checking. Since the failure results were, uh, bad. He never got into multi-threaded demons, which is a bit of a pity I think.

He didn't have any particular cost associated with it, which was a physics problem. You've got that, so that also adds some fun complexity into it.

First thought: supply chain attack

Lead to: enslaved "computers" doing the drudge work of keeping the castle lights >blinking<

It's too sane.

If you have ever tried to persuade wannabee (and even experienced) programmers that they WILL make errors or, worse, persuade 'computer scientists' (ESPECIALLY ones with associated publications) that no amount of type fancification and other language gimmickry will EVER catch all errors, or even catch most of the 'difficult' ones (*), you will end up as cynical as I am :-(

Let's ignore the manageritis and publication mania that leads to the attitude "There's no time to check it - that can be done later - just get it out there fast and first."

I agree that having their souls carried off to hell would enable a salutary amount of evolution, but I am not sanguine that would work in a small number of generations. The universe would be solid with people selling 'guaranteed safe' grimoires, and dogmatists mining their favoured ruts. And Cthulhu help anyone who believed either of them!

But, if you give the idea to any decent author, who is likely to give the idiots I describe the ends they deserve, PLEASE tell me how to contribute to the project :-)

(*) Measured by the degree of skill and amount of effort to locate them, given adequate debugging tools, full access to (and even an understanding of) the source. There are two reasons for this being a fundamentally insoluble problem: one derives from Turing/Goedel and the other from the fact that some of the nastiest errors are when the program is correct but solves a slightly different problem, or one with subtly more constraints.

Charlie, which works of High Fantasy do you genuinely like? And I mean works that takes the subgenre seriously, not parodies, deconstructions or subversions.

And to take it a step further, a large conglomerate creates a centralized and highly simplified intermediary platform that automates all the mundane and finicky aspects of pact-based magic so that anyone can quickly and easily arrange for magical tasks to be performed at the drop of a hat. Maybe call it "Amazing Wizard Services" ;)

I'm with Cynic.

Everybody, including me, likes to build systems at the limits of their complexity handling. Then some more complexity comes along. (Because real life always has surprises.) Now you have built a system slightly more complicated than you can understand, which means either bugs are leaking in where you can't quite see them, or the code is brittle -- "this works but adding more features is hard work." (If you're honest, you say "I'm afraid to touch it any more.")

Software engineering as a discipline is a set of tools to improve your complexity handling. This does not make the above limit go away. It just means that you can build bigger programs before you reach it.

With enough experience, you reach the point where you recognize what kind of problems you can successfully tackle and you tackle them reliably. Now your mistakes are of the form "whoops, I tackled slightly the wrong problem" (as Cynic says). This is still complexity handling and the above still applies.

If software errors got your soul dragged to hell, I wouldn't have made it to age 25...

(Also, it seems hard to justify heat as a valuable currency in a universe where thermodynamics applies. Maybe hell doesn't have thermodynamics and that's why demons are willing to bargain?)

I have a list of "Schools Of Magic" that I will almost certainly never get around to turning into anything:

Magic as a bargain: magic result from striking bargains with supernatural entities, and so magic school is like law school.

Magic as art: anyone can do magic, but it's not very practical, a small number of Wizards get rich and famous, but most give up magic and get proper jobs instead, so magic school is like drama school.

Magic as code: magic is precise, low level instructions inserted into the runtime of the universe, and magic school is like computer science classes.

Magic as innate power: some people can do magic, but it can't be taught, everyone does it their own way, so magic school just teaches ethics.

Magic as science: magic can be studied and analysed and anyone can do it, it's just another branch of sufficiently advanced technology. Magic school is like an engineering degree.

Magic as psychology: magic is real like money is real, like the law is real, it's a social construct you can't find under a microscope. Magic school is teaching applied psychology and stage hypnotism and is non-stop mind games.

Magic as money: magical power must be hoarded and built up, and when it's spent it's gone, so it is not too be used lightly, but it can be lent and borrowed and traded. Magic school is like business school: where gets you the best return on mana employed?

Magic as a foreign language: the true names of things give you power over them, so you must memorise a lot of enochian vocab. Magic school is like German lessons.

Magic as a lost art: the great mages of the past could do magic, and we're trying to find their old secrets. Magic school is like studying archaeology.

To shoe-horn this into a single plot then I think we have a bunch of students who have just finished their A-levels hopping round the multiverse trying to work out which magical system suits them best, which seems weak, so I think that in the fantasy universe were I get to try and write this, it's just a collection of thematically linked short stories.

So does that mean that the motto of your fantasy equivalent of Facebook is "Move fast and wake Things"?

Okay, this successfully nerd-sniped me into me delurking and finally posting here.

The obvious result of this system is that a justification for ye most honorable fantasy trope, "dragon with a hoard under a mountain with smaller monsters guarding the dungeon," falls naturally out of this system.

After all, dragons are long-lived and magically-gifted creatures. They also innately spew flame, which could be used for demonic pact powering in lieu of coal.

So, the average dragon finds a mountain relatively rich in gold ore. They power a series of very simple drilling and sifting demonic scripts with their flame, gradually moving gold dust from deep under the earth to directly underneath their feet in a growing pile.

This effort also naturally hollows out the mountain over time -- conveniently in the right angles favored by dungeon designers due to the simplicity required for safe coding! -- leaving space for other creatures, who are naturally attracted by the public utility of the spillover heat left by the dragon. After all, it's hard to steer flame and stone traps heat, so there's a lot of leftover power to run your household grimoires off of, so long as one doesn't disturb the dragon...

Note that this same thought process also explains what happened to Moria (an industrial accident with their experimental Balrog code) and a justification as to why Sauron needed to locate himself right on top of a volcano and was obsessed with control systems...

Finally, the best part is this: you don't need to explain why a bunch of magical swords, armor, etc. are in the loot pile in the dungeon. Rather, the goal of the adventuring party is to get and retain control of the dungeon heat source long enough for their extremely slow-but-safe grimoires to run "upgrade my sword" spells off of the available power...

Charlie, which works of High Fantasy do you genuinely like?

Not Lord of the Rings (too pastoral-ish) or Game of Thrones (I dislike grimdark and I don't need to be spoon-fed the history of the Wars of the Roses, with added dragons and zombies). Tried The Wheel of Time and bailed after three pages of turgid prose.

I also have a laundry list of high fantasy tropes that enrage me. Like, currency based on precious metal coinages in a fixed exchange rate ratio between metals using decimal or base-100 with no awareness of seigniorage or clipping or adulteration to different degrees under different monarchs (which is why the head on the coin is important) ...

I immediately went to "what happens when they make logic errors?"

A runaway infinite loop that drains all heat from the surrounding area. A forkbomb equivalent could be weaponized. Freeze your enemy's lair/castle/domain into a solid lump, which never stops growing.

But heat is too easy. Why not use souls? Not the whole thing. Maybe a small part of it. So magic users still have most of their soul, but a chunk (or chunks) are owned by whatever demons they've bargained with. As long as they still have the biggest chunk, they can still control their own lives. Maybe the demons bargain with each other to accumulate bigger shares of a specific individual. And then one day, one demon winds up with the largest share.

Or viruses/worms. Nobody wants to give up their own soul, after all. So why not attach to someone else's "process" and let them pay instead of you. So you may think that you're OK, but then discover that some demon got a big chunk of your soul mining crypto for someone else.

"Also, it seems hard to justify heat as a valuable currency ...."

One can easily make it simply coal - they eat it and, in Pandemonium, good-quality coal is getting more expensive (*), so bribing them with anthracite works.

(*) My mind fails to encompass what demons would use for currency, especially given what OGH said in #12!

That sort of thing is why I find almost all pseudo-mediaeval fantasy irritating - I like some, despite those tropes, but their worldbuilding is almost always dire. LOTR is extreme in that respect. The one I find most irritating is travelling long distances without major supply problems, but there are a zillion others.

So there are none? Why do you feel like engaging with a genre you don't like?

Also, wait a minute, don't you have an entire highly successful series of books with magic as software?

PS. Fixed exchange rate is more of a D&D convention, I believe, to make life easier for DMs (who are usually not writers, and frequently don't even do serious worldbuilding, because you can have a fun game without). But sure you know that, didn’t you use to write for D&D?

Why not use souls?

You really need to go read the Craft Sequence by Max Gladstone. Start with Three Parts Dead, then continue ...

Always thought the Deadlands Roleplaying game had an interesting concept of casting spells with the player playing a game of poker with the a demon as written in a copy of Hoyles Card Games.

It also sounds like something the House Ex Miscellanea has a school in from Mage the Ascension.

Oh, the coal/heat thing is easy.

Magic, the demon-powered variety, is powered by pain.

Demons collect human souls and torment them to generate the pain they use to make stuff happen.

Coal? Raw heat? It's all for tormenting the captive souls.

Ditto the infernal bargains, of course, but they don't discuss the why of corrupting human souls with humans any more than humans discuss battery farming with chickens.

What if the mage doesn't want to be bound to regularly paying wages to a demon? Could they enslave a demon? Or purchase the captives from demonic war?

When I'm running D&D games, I make sure to tell my players "For the sake of ease and convenience, we will pretend that all of the coins you get are copper, silver, gold, etc. However, in reality there are a bunch of different coins with wacky values, and none have a fixed denomination. Cash money you find is made up of thalers, dollars, pounds, louis d'or, florins, marks, crowns, et numerous cetera."

I once had the PCs go to a faraway place where they actually used paper money. They tried to pay merchants with their regular coins and the merchants said: "Go get some real money from a bank. Can't take that crap. Whatever it is." Lots of histrionics ensued....

I've always liked to think about magical programming languages as imperative, not declarative. What the spell is supposed to achieve rather than how to achieve it. Granted, it wouldn't work well if it's a specification given to a demon, as lots of genies-granting-wishes stories illustrate.

The first thing I wonder is whether there are an unlimited supply of microimps or just a practically unlimited supply, like TCP/IP v4 address ranges. Though not necessarily to that exact number it might be something defined in this universes' bible. If you start getting a shortage it could have interesting effects.

If screwing up damns you, expect that sourcerers are kept at a distance from anyone important. Quests seem like possible busywork to keep at least some of them from blowing up in town.

There's a lot of D&D players who don't want any magic other than simple magic weapons and armour contaminating their brawny or skillful character. Fighters and rogues have always been popular. The equivalent people in this world might just have a point about the dangers.

I'm thinking of this XKCD, except:

ALL MODERN DAEMONIC INFRASTRUCTURE vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv A cantrip some random magus in Hammerblight has been thanklessly maintaining since the Third Era

That just calls into question the Minimum-Torturable-Sophont.

Obviously someone is going to try to uplift pigs/chickens/rabbits/etc. Into a quickly reproducing just-sapient-enough mass sacrifice.

One effect is that the old saw about undefined behaviour in C - when the compiler encounters a construct where the standard does not specify what it ought to do, it is legal for it to make demons fly out of your nose - holds in this world, but it really means that demons can fly out of your nose.

Something about this reminds me of a story called ...

... 'Examination Night', by a chap called Charles Stross.

I always felt that was worth further development, though it's clearly more a 1990s Charlie than a 2020s Charlie.

I guess my grumble is, do we have to continue using Fundamentalist Xtianity as the baseline, hell, demons, and all?

How about Mammonism instead? You can make any sort of commentary you want about modern culture with this as your base model.

So I'll make a few tweaks:

The world was not created. It IS being continuously created by the gods. It's what gods do. They're Urges in a Greek sense. We'll call demons Catastrophs, because they destroy things. It's what they do. It's what they want to do of their own free will. This is not a moral world, but one that is good because it functions. To function, both urges and catastrophs are fundamental and necessary. Everything they do has a place in the world. Humans and other species, to the extent they have free will, can choose to some degree how they work with the urges and the catastrophs.

And then there's the spirit--or is it spirits?--of Mani.

Mani has the properties of both mana and money. The being associated with it is both an Urge and a Catastroph, and neither an Urge nor a Catastroph, thus pointing to the fact that Urges and Catastrophs are not fundamentally unlinked beings. Similarly, Mani may be one spirit or many, of any gender or genderless, completely free-willed and having no free will at all. This ambiguity is an essential feature of Mani. Or is it a bug? Or both?

The power of mani is largely based on belief in it. Mani has power to the extent that it is believed to have power. "Value" and "worth" are crude measures of how much mani something is believed to potentially have or need.

Mani only works when it is given. Baptising a thing into the Realm of Mani happens when a being of free will claims to control the thing, and a second being of free will gives mani to the first in return for control of the thing. "Thing" can be broadly defined here: rock, person, urge, catastroph, etc. Ditto free-willed being.

A thing can only leave the Realm of Mani if the being that claims control of it is given what it/they believes is enough Mani to let it go free, possibly with "safeguards" preventing it from being claimed for the Realm of Mani again. Enforcing those safeguards depends on having mani available to enforce them. Or on the thing being believed to have no mani value and actively disengaging from the Realm of Mani.

So getting back to the magic system above, let's assume it can run on mani. Let's also assume that urges and catastrophs in the Realm of Mani can do things for magicians. They will work slavishly if given the right amount of mani to do so. Too little, and they'll remember they are free-willed beings and not do as ordered. Too much, and they'll get greedy or disgusted, remember that they're free-willed beings, and not do as ordered. If the magician can keep them happy, they'll work like machines as long as the right amount of mani keeps coming in.

I should point out that most urges and catastrophs are not in the Realm of Mani, despite Mani's attempts to take over the world. They are perfectly willing to work with "magicians" who are willing to form positive relationships with them--so long as mani is not involved.

And yes, mani keeps trying to take over the world. But it only has power to the extent that beings believe that mani has power. Lose that belief, and mani's power diminishes or goes away entirely.

Feel free to claim these ideas as often as desired. You don't have to give me any mani to do so.

I've been kind of wondering about this ever since I read the Harry Potter books and started wondering what Potterverse magic might look like if it was actually systematized instead of being plot-sensitive phlebotinium. "Magical theory" is an academic subject (HP studies it), and there are accounts of people inventing new spells, presumably based on the theory. Wands are important, but so are words.

What follows is not based in the Potterverse as such, its a magic system that might look rather like the Potterverse one at first glance.

System

A wand is a kind of crochet hook for the substrate of reality. With it you can catch a thread of this substrate and weave it in with others. This requires both physical dexterity and the ability to perceive the invisible material you are working with (i.e. magical talent). In theory a mundane could use a wand to cast a spell, but it would be a bit like trying to crochet wool using waldoes that you can't see.

The structures you can weave act as magic "circuits" or "mechanisms". A magic spell is simply one of these mechanisms. They can be arbitrarily complex and can be constructed to be stable, assuming you can get the loose ends tied off. Once you have your spell you can poke it with your wand and it will do stuff, or it might be constructed to start running immediately.

Magic users memorise "cantrips"; simple spells that can be constructed quickly to do useful things. More advanced users can invent stuff on the fly.

Because magic can have non-local effects, it is possible to attach a complex spell to a particular word. Once this is done (its a tricky and time-consuming task) the spell can be invoked by a combination of the word and a basic cantrip. Mock-latin is popular for these words because you aren't likely to say it by accident, but any sequence of phonemes will do.

Spells interface with the intent of the caster/user; part of magic use is using your will to pass information into a spell for it to act on. So the Potterverse "obliviate" (amnesia) spell requires a definition of the memory to be erased in a form not dissimilar to a SQL query.

Creating a spell that interfaces to reality in a desired way is a particular part of this ability. Simple spells will not affect the material world in any way. Constructing a spell to have a precise desired impact on the world is a tricky business, but doable if you have studied the techniques. Spells that couple with matter can have non-local effects, but they are still constrained by the basic laws of physics: a fireball has to get the heat energy from somewhere hot enough, and a levitation spell must be pushing down on something.

Spells can be modularised. Past magic users have created a library of spell components which can be invoked in the same way as any other spell, but are intended to be interfaced to other components. So its possible to build up complex spells out of these components.

Version control is a hot mess. The spell associated with a particular word can be updated by anyone using the same technique used to replace it. So its quite common for a magic word to suddenly change its effects or even stop working altogether. There is also no copy protection beyond keeping the word secret: once someone learns the word that invokes a spell there is nothing to stop them from doing it as often as they want, and no way to learn who they are or to charge them for it.

Optional Extras

Have any animals evolved to make use of magic? Could there be creatures that exist largely or completely in the reality substrate of magic? That would explain stories of demons and monsters, and (as in the Potterverse) magic creatures would be an obvious source for wand components, since they have the built-in ability to interface to the substrate.

There was definitely demonic influence in the design of C and C++ - I could name names, but won't. Some of the most extreme examples were leaving something ill-defined because it was 'obvious' it meant X, later changing the interpretation to Y, and denying that there ever was a change, on occasions to avoid having to fix a compiler bug. Undefined behaviour is not always due to simple programmer error ....

In the universe of this post, and if one assumed that demons had any influence on the languages of invocation, one can easily see them doing exactly that, and even once-reliable grimoires becoming death-traps.

(This is a separate comment from my Potterverse-like magic system, since the general point isn't much affected by the details of the system. Charlie's and mine would have similar effects).

If you have a moderately dependable mechanism for doing magic, and it can be analysed, learned and invented, then there is going to be a kind of magitech industrial revolution. The world isn't going to be a medieval swords-and-sourcery place for very long; its going to morph into something like the Shadowrun world, only more steampunk than cyberpunk. Then once electricity gets discovered it will move towards something like the modern world but with magic, and then on to who knows what.

In Bujold's World of the Five Gods a magic user figures out how to turn a hand-written sheet of paper into a printing plate: put the sheet face down on the plate, and use magic to engrave the ink into the plate. The paper is destroyed (conservation of entropy), but presumably one could then take a printed sheet from the press and repeat the process, so that many identical plates could be produced. It's a neat trick and its inventor publishes the details, but 200 years later the society is still a medieval aristocracy.

In our world the invention of Gutenberg's printing press seems to have been one of (if not the) key factors in the start of the Industrial Revolution. Printing made it practical to buy a book to learn a monetisable skill, printing costs dropped by 3 orders of magnitude in 100 years, and lots of inventions suddenly started happening. (We've discussed this here in the past: see The Day the Universe Changed by James Burke ).

I'd have expected something similar to happen in the World of the Five Gods. And there would be lots of interesting stories there, starting with "what do each of those gods think about industrialisation?". But it didn't.

So a part of the world-building here isn't the construction of a generic swords-and-sourcery world with some magic users added in, its much more about what the stories that their version of James Burke will tell about the development of the modern magitech society, with its imp-powered iconographs.

Charlie, I like the idea of magic as Unix - each little thing does one thing, but does it well.

I am reminded of the Demon Tech books by my sigh late friend David Sherman. Guns? There's a small demon inside, and when you pull the trigger, it pinches them or something, and they spit....

Nasty thought: so, you trade them coal for work, and you have to buy or mine coal to pay them. Then this guy comes along to try to sell them methane...

Truly warped thought: this magician comes up with a perfect way to keep Hell hot: to steal heat from Heaven, since it's been shown for many years that Heavan is hotter than Hell.... (for example, https://www.me.iitb.ac.in/~awdate/other4.html )

In my career, it's the new programmers, or the ones who have been in one line too long. The ones right out of school have never had to actually validate input data.... And managers and such are sure they know how it should be done.

And don't get me started on "subject matter experts".

EVERYONE wants to read the Craft Series.

Define "chunks of their soul". Perhaps "x hours/days/weeks/years"? I mean, given that it's all finite (as opposed to the Christianists' view, that infinite punishment is just right for finite evil")? You get to torture me for your doing this work.

I can see an older mage, "I am NOT going to be tortured like that again for something that stupid...."

I see them carrying a lot of the equivalent of Elf-made MREs.

On the other hand, and this is supposedly true, the treaty that Billy Penn made with the local Native Americans was for as much land as a man (or was it three men?) could walk around in three days. The Native Americans figured they'd have to stop at some point and hunt, and start a fire.... What Penn did was send good walkers out, carrying three days food on their back.

Oh... you mean like the one guy in the world maintaining NTP for ALL *Nix?

[...] but you'll understand how it works!

No, no you won't. What will happen is quite different. At some point in the past a system of grimoires was arrived at which was both far safer than preceding (and many succeeding) grimoires and sufficiently powerful to express a vast range of existing grimoires (this ability, for reasons too large for this comment, was usually known as the 'consideration' facility). This system, or rather this family of systems of grimoires, often known as the āwlyspian grimoires, was immediately derided by the authors of other grimoires (collectively known as the pelagian grimoires) as being too complex, too ugly, and too slow for serious magical work, especially since all serious practitioners understand the safety issues and write grimoires without the flaws which the āwlyspian grimoires sought to avoid.

As time went on and several countries were lost to catastrophic demonic infestation as a result of errors in pelagian-family grimoires, many safety features of the āwlyspian grimoires were implemented in their descendents, usually without any understanding by the practitioners that this was what they were doing. Pelagian grimoires became slower – far slower than any āwlyspian grimoire had ever been – as a result, while still remaining intractably difficult to use and never gaining the consideration facility in anything other than a rudimentary form.

In the mean time, certain sects of the practitioners of the āwlyspian grimoire family (some survived) observed that the consideration facility, as usually described, contained certain issues of safety, whereby, for instance, and imprecation to a demon might be misconstrued as a different imprecation in considerations written by inexperienced practitioners. For many years these practitioners – the 'arrangement' school as they were known – refused to use considerations at all in any official capacity. Eventually, a new consideration scheme was arrived at, expressed in terms, of certain abstract metamagical submonoidal morphisms, or equivalently in terms of undiagramatic graphs of groipoidal cycles. A theorem, running to more than four thousand vellum pages, was proven to show that these two schemes for considerations were equivalent. Several people understood it.

A new family of āwlyspian grimoires using these considerations became somewhat fashionable. It is widely believed that nobody understood it.

In the pelagian world two related processes happened. Grimoires became vastly complex and stylized in order to prevent inadvertent demons: nobody understood these because it would take far longer than a human lifetime to read any one of them. Other pelagian grimoire schools, being dimly aware of the arrangement school, started to describe their grimoires in terms of semi-infinite femtomagmas: nobody understood these either.

While this evolution was still taking place, the seas rose – completely expectedly – and everyone drowned.

I like that. For that matter, if humans can do magic, I'm positive that critters can, to some degree or another.

Thought, Go Away. No one is going to find a woman magiced by her cat to take care of aforesaid cat....

Nahh, as some point, there was a school that beganning diagramming the spells, using a technique named after the leaders of that school, Backards and Noor, which made if far more comprehensible.

They did have a whole section, often ignored, on Things That Need To Be Addressed. This includes uncompleted circles, which result in all sould in a range of X will be drawn in and taken.

This seems very close to Ethereum smart contract programming today, where the risk of failure is loss of all wealth/ bitcoin/ bankaccounts. Maybe in you system you can partition your soul into chunks, or perhaps the souls of children (as we have the classical witches capturing children for something tropes) as a method payment. Like credits for AWS.

In Bujold's World of the Five Gods a magic user figures out how to turn a hand-written sheet of paper into a printing plate: put the sheet face down on the plate, and use magic to engrave the ink into the plate. The paper is destroyed (conservation of entropy), but presumably one could then take a printed sheet from the press and repeat the process, so that many identical plates could be produced. It's a neat trick and its inventor publishes the details, but 200 years later the society is still a medieval aristocracy.

While there are undoubtedly some holes in Bujold's worldbuilding, that isn't one of them.

Firstly, the process only works (as I recall) with a sheet written out by the magic user doing the transfer (which is part of the creator's enthusiasm for getting it working with metal plates rather than wooden ones). Which means that if you want a book made that way, you need to have a sorceror write it out longhand. Which brings us to the second, much larger, problem.

The limiting resource for the process, in that world (even if we allow the use of a printed sheet to create a new plate) is the availability of a sorceror to do the job. Temple sorcerors are rare and in great demand, and in any case their priorities are decided by the White God, with the temple hierarchy assigning whatever's left after that. Creating new temple sorcerors is a lengthy and demanding process, and attempting to take shortcuts or arrange your own supply is ... unwise even beyond "normal" attempts to appropriate a God's power for your own purposes.

Hedge sorcerors powerful enough to do the job are probably even rarer, and very unlikely to do anything that leaves such an obvious mark because of the attention it will draw. And while "just making books" might seem excusably benign to those unfamiliar with Bujold's world, there are very good reasons why the temple hunts down free/rogue sorcerors and (usually) strips them of their powers. Untrained sorcerors are extremely dangerous to those around them (and indeed to themselves; they tend to die young, unpleasantly, and not from human actions), and the only way to mitigate that danger is ... divine approval and several years of seminary training. (NB for those unfamiliar with the series: I don't mean approval by the temple authorities. The White God gets personally involved, and the stories feature some very clear and unambiguous divine wrath pour encourager les autres.)

I don't, for a similar reason to OGH's about currency. Unless you are rewriting the laws of physics and biology to evade this issue, a walker needs c. 1 Kg of food a day, or will lose 250+ grams of weight a day, or 1/2 Kg a day and a 125+ gram weight loss. You CAN halve the weight by eating almost pure fat, but will get serious gut problems if you do it for long and don't mix it with a lot of vegetable matter, which takes time to gather and cook. And possibly even if you do (e.g. I would)! Few people can walk effectively with more than 25 Kg on their backs.

Note that is relatively gentle walking - say, 20 miles over fairly easy going (e.g. dry but rough or grassy dirt roads / tracks over undulating country). For tougher going, longer distances, or cold weather, double the figures - for seriously tough going, double that AND cut the distance.

Furthermore (shade of Tolkein please note), hunting isn't as much help as all that:

https://en.wikipedia.org/wiki/Protein_poisoning

Version control is a hot mess. The spell associated with a particular word can be updated by anyone using the same technique used to replace it. So its quite common for a magic word to suddenly change its effects or even stop working altogether.

Easily dealt with by introducing UUIDs (aka GUIDs), only instead of an unpronouncable 128-bit integer it's encoded as a sequence of otherwise-unconnected words or phonemes. Each invocation of a UUID introduces or loads a dictionary that contains a snapshot/copy of every desired spell in the library: part of setting up a new spell is taking a snapshot of the existing units you want. Uh. magical github? Frankly, you don't need a full version control system in a system of individualistic magic -- initially you just need enough for a single user: VCS's can be quite simple (I'm looking at sccs(1) as the canonical example, never mind the far more sophisticated but-still-single-user rcs(1).)

Magical animals: yes, and be afraid, be very afraid, of magic-wielding predators with theory of mind.

Robert Jackson Bennett's Founder's Trilogy covers similar territory, only with animism instead of imps. "Tech" giants, scrappy hackers, open source, etc.

One thing that's led to the real world software bloat that you allude to has been that Moore's Law has compensated for a lot of that bloat. I assume the imp equivalent is that there are various grades of imps in the Demon Halls, and the initial summoning spells only got the worst, slowest ones, to begin with. Then the expanding magical capabilities led to better summon spells that got faster, coal-efficient grades of imps.

Now what happens when the imp quality tops out? First, I imagine that a scrappy hero, in the midst of his heroic deed of saving a mead-hall from beastman attacks, figures out a way to gang together many low-quality imps to achieve the same thing as the well-funded royal demonologists. But what next? Freezing the tiniest of imps next to each other until they start to blur into each other like some kind quantum imputer?

Very curious what you'd make of the Founder's Trilogy (https://www.robertjacksonbennett.com/the-founders-trilogy) -- magic based on text scribed into objects that "programs" the object to have different properties, like an arrow thinks it's 10x its own mass so does more damage.

I like the idea of magic as Unix - each little thing does one thing, but does it well.

Unfortunately that theory doesn't always link up to the reality. We get simple single-purpose tools that crash and burn when you fuzz their inputs, and then you get helpful idiots adding a metric shitload of command-line options and/or hard links that change the behaviour of the underlying binary because of course fgrep, grep, and egrep are all variants of grep, and then some numpty at the FSF causes the GNU version of fgrep to emit a snarky message to STDOUT telling users (who are mostly shell scripts by this point) to use grep -f instead (because obviously this is the right thing to do if you're an idiot who doesn't appreciate you're maintaining critical infrastructure), only in this metaphor a demon turns up and eats your soul rather than most of your shell scripts suddenly giving incorrect output.

And then you get generalist tools.

Anything you can do with awk you can do, more or less, with Bourne shell, sed, and various simple commands. But awk is so convenient at text mangling that it does an embrace-and-extend on any shell scripts that have to do significant mangling of textual input data.

Then you want to add sockets and random POSIX shit and you get Perl 4, which is a strict superset of awk and absolutely awesome for 1992 values of awe. But Perl 4 is extremely bad at structure in the large -- you can write libraries of code by defining local namespaces but they're leaky and messy and duck-typing is just asking for trouble. So then you get Perl 5 which adds references and function prototypes and about sixty bazillion different ways to do OOP and functional programming, sometimes in the same sentence while standing on your head juggling chainsaws without a safety net. Which garners complaints so then you get a 20 year project to invent Perl 6 which turns out to be an entirely different language, and everybody gives up on it and buggers off to Python in the meantime because the lack of a major version number increment convinces them that Perl 5 has died (rather than going from 5.001 to 5.60 in 59 incremental releases) ...

No, it's just basic printing as from before movable type - see Sir Pterry's The Truth for an example. You write the paper, a mundane process, the only magic is that the sorcerer is writing a textbook of what they do. You press the still-wet page to a boxwood panel, transferring a reverse image, and the engraver carves away the letters so that you have an engraved image that you use to print with maybe a dozen times before its worn away. Paper is not destroyed, but the ink is only so wet so you get to do the transfer just once. Penric and Des figured out that with her ability to make rust, they could etch a steel plate, making a much more robust engraving that would last an indefinite number of pressings.

On topic - an engraving like that would enshrine errors into hundreds of widely-distributed and -used grimoires instead of only a dozen at a time, leading to to some nasty legacy code issues.

No I really do not want to meet the demonological equivalent of Perl 5, thank you very much. Unless it's one of the later iterations with "use strict" enabled by default and all the safety checks welded and padlocked in the "on" position. And even then I'd be extremely wary.

Any demon summoning toolkit more complex than the System 7 UNIX command line environment (think Xenix) is probably way too dangerous to hand to mere postgraduate researchers because the amount of crap you'd have to memorize to know what was actually going on would be too much.

(I have an old copy of the Mark Williams Company's Coherent 3.2 somewhere, which was a System 7 clone: the manual is a miracle of concision at only about 1100 pages ...)

Chrisj @ 42: While there are undoubtedly some holes in Bujold's worldbuilding, that isn't one of them.

Fair enough. But the wider point still stands: if you have magitech then people are going to get inventive, and from there its a short step to an industrial revolution. Feudal aristocracies depend on keeping people in their place: "the rich man in his castle, the poor man at his gate, God made them high or lowly, and ordered their estate". Modern software engineering, or anything like it, is going to be a nightmare for anyone who wants to keep things the same because people will keep mcgyvering it.

I suspect the only way to handle it is via something like Graydon Saunders' Commonweal series -- where magical ability is innate so sorcerers become nobility until a major disruptive event occurs and there's a revolution and then a harshly egalitarian state emerges where the penalty for trying to set yourself up as a petty god-king is drastic and permanent. (Fast forward 500-600 years and they've laboriously legislated a framework for taming and enabling godlike entities to coexist with mere mortals in a civil society without turning into oligarchs, and a socialist approach to development is improving everybody's standard of living ... subject to disputes and the odd attempted invasion and genocide by the crazed devotees of fire gods.)

How you get to that nobody-has-to-die-unless-they're-unreasonably-addicted-to-power situation from the nasty/brutish/short age of Dark Lords that precedes it is the most fantasy-esque aspect of the setting.

The "HELLO WORLD" joke writes itself! No, I mean literally, it's writing itself, please send help.

So you build your system out of the most stupid demons.

Shouldn't these demons just be a part of the whole hell-dimension demon biosphere? Probably an important part?

Imagine if some interdimensional aliens were to start exploiting Earth by stealing our microorganisms.

You can't live without microorganisms.

So, bottom line, this idea ends with the smart demons invading to stop us.

How you get to that nobody-has-to-die-unless-they're-unreasonably-addicted-to-power situation from the nasty/brutish/short age of Dark Lords that precedes it is the most fantasy-esque aspect of the setting.

Immortal godkings versus mortals?

The obvious answer is that mortals do to godkings what pests are doing to us: have short generation times and keep evolving new attacks, Red Queen style. Eventually, the mortals figure out how to do whatever-it-is that the godkings can't withstand. Then the godkings are stuck in the position of, say, various Papuan peoples, who have to deal with up to four different types of malaria simultaneously, suffer terribly, and watch a majority of their children die as babies, all with no hope of a cure or escape. So they make their peace with the inevitability and ubiquity of the mosquitos. And so godkings would suffer mere mortals, not because they want to, but because we're unavoidable and collectively unbeatable.

The reason we don't see this as an obvious answer is that we're conditioned to sympathize with the godkings, rather than to realize our identity as mosquitos, so far as the godkings are concerned. Even Graydon made that mistake.

*Aprocryphal bathroom graffiti from a university biology department: Someone wrote: "Oh God, why are we born, only to suffer and die?" And someone wrote below that: "Because those who suffered and died left behind more descendants than those who did not." No one I've told this to has ever considered this a lesson they want to teach to students. Especially not to the undergrads they teach.

Yep. Tell me alllll about it - yesterday, being the beginning of the month, I did my monthly backup and update. And when it came up (almalinux, which just went from 9.1 to 9.2), I tried to log in... nope, went back to the login screen. The only way I could get back to where I wanted to be, running KDE (I HATE gnome and its bloat) was to create ~/.xinitrc, and put startkde in it, then run $: startx.

The only good thing was that, unlike 90% of the time, when it came up, the four windws I wanted up were in the correct place, and the minimized half-down were, in fact, still minimized.

We'll see what happens tomorrow.

Oh, and hey, I love awk. But then, it saved my butt in the early nineties: a project I was on, we, the contractors, were to be the magic on the board diagram, and we'd tell all the court-ordered sources what format to send their data to us in.

Every one said, "sorry, no budget, do you want it whatever format we have or not."

I wrote about 30 awk script to reformat them... and I'm not talking 2 liners, I'm talking 100-200 lines of awk.

Speaking of simple magical languages, I intermittently amuse myself with the idea of a language of magic isn't complex, but very simple, like the conlang Toki Pona.

I don't speak toki pona, but my understanding is that it's so simple that "mi moku" means both "I eat" and "I am food." To communicate in it, apparently handwaving is necessary. Possibly diagrams?

Anyway, if this was the language shared by a magician and a daemon, I suspect that handwaving and diagrams would be necessary for the magician to explain to the daemon what they wanted. Ritualize communications to make life easier for everyone, and you have a magic system.

And oh, the failure modes...

Um, er, as I've been saying for a looooong time, the OT says that A&E ate of the fruit of the tree of knowledge, and God said, "before they eat of the fruit of the tree of life, let's kick them out."

Simple grammar reads that as the only difference between man and god is that they live forever... which means they ain't smarter than we are.

Yeah, no. "Craft sequence" is the worst kind of magic worldbuilding. It's just a word salad of technobabble.

Perl 3 was really the successor to APL - a write-only language. And I can assure you that you really, really do not want to have to look at its source!

Typo: Perl 4.

Simple grammar reads that as the only difference between man and god is that they live forever... which means they ain't smarter than we are.

Yup.

So far as humans being the magical equivalent of mosquitoes...imagine something like Wiccan Rule of Three, only as rapidly evolving and collectively unstoppable as malaria, at least so far as godkings are concerned. In other words, whatever they do magically rebounds threefold upon them. Kill a bunch of mortals with magic? Die that many times, tripled, except that godkings don't die permanently, so they get to revive at the end of the last death they experience, then heal themselves however they can. Suck all the mana out of a mortal, only to lose all your mana three times over? Maybe not something the discerning godking would want to do?

It would make being a dark lord rather painful, at the very least.

Heh. I take it you have never met a lawyer ...

I was about to make a joke about a demonic fork bomb but then I realised that would probably make an extremely horrifying weapon especially if used in a populated area and if you could get a hold of some Plutonium-238 (unlikely in a fantasy universe but still) you could probably have something that in terms of danger to life would be akin to a nuke

Or a rabbi...(or Orthodox Jew).

So, if the feedstock for a foodprinter is a genengineered weed, with trace elements added as needed, is printed bacon kosher? (respondent 1: yes; respondent 2: the rabbis will say no).

Ha! I recently started ground work on a new project. One of the things I added to the design was automatic test generation, with constant checking of invariants as a runtime option.

"Why would you want to do that?"

"Because we all make mistakes" was not an acceptable response.

No, see, I think there's a different direction to take this. All of this is tying the system to software engineering and trying to find parallels there, but what if this were to go down the track of:

• Excel is the most-used programming language in the world
• The world runs on spreadsheets
• Most spreadsheets are disasters
• Also, Visual Basic for Magic and all those Windows applications
• ... and backwards-compatible magic on the world's most popular magic system going back decades so that even a home automation spell your mum wrote in the 1980s can still run on the new hotness in the place you're renting.

Come to think of it, this is starting to sound a bit Pratchett-y.

It is trivially obvious ... that "Faust" is a load of christian lying bullshit, right?
but heat (to keep the fires of Hell burning) - WRONG - the only person to visit, recorded that the ultimate depths were frozen over ...
{Reference: "Divine Comedy" One, cantos: 32 onward ....
SERIOUSLY: READ Dante, first!

Auricoma @ 6
LotR? Or maybe not?

Ben Curthoys @ 9
"Magic as psychology" - translates to: Magic is POLITICS ...
"Magic as a foreign language" - well, I managed to summon a bumblebee, by calling her by her "True Name" ... once.
See Also: "Earthsea" by U K Le G.

Charlie @ 12
Lois McMaster B's universe?
... @ 19 ... "Raw Heat" = Nuclear Power, oops.

• see also ChrisJ @ 42 ( An appropriate number, too! )

Elaine @ 48
Where did all this "Penric" knowledge go to & vanish, by the time of "The Curse of Chalion" I wonder?
Destroyed in the wars of the Golden General?

What levels of abstraction will one need to understand all of this, given the parallels with computer coding & science?

Re: '... in return for the sum of all knowledge'

I'm thinking of a scenario where the imp steals rather than 'discovers' knowledge.

First off - all the knowledge as of what date?

If the overall proportion of wizards in that universe is comparable to research scientist in ours, ditto for research slog, then every single wizard (and soon-to-be wizard) will be continually learning something/accreting knowledge. So unless your imp can zip in and out of a pocket universe after every noting every bit of knowledge right up to the end of that universe or death of all wizards (whichever comes first), it cannot deliver the sum of all knowledge.

Also - if the comparison to scientists holds, then the sequence of the various layers and connections between bits of knowledge might also be important. If your imp just collects and dumps all the knowledge on your lap, you won't know what bit of info goes with what, what the exceptions are, etc. Then there's obsolete knowledge because someone cam up with a better/faster/easier/more reliable way of pulling off a spell.

What about language/grammar of the various snippets of knowledge - unless everyone always spoken exactly the same language all over that world?

Good scenario for exploring every way that the imp could screw up (short stories) kinda what Asimov did when he started writing about robots.

Paul @ 29 and ttb @ 38 - very interesting! Thanks!

So unless your imp can zip in and out of a pocket universe after every noting every bit of knowledge right up to the end of that universe or death of all wizards (whichever comes first), it cannot deliver the sum of all knowledge.

Assuming that you have been sufficiently unambiguous to use the word "sum" to mean "all knowledge in aggregate" and not "a summary of all knowledge".

I'm reminded of the Pete Seeger story of the king who sends out his wise men to summarize all the world's wisdom in a book, for his sons to learn. When the complete that task he orders them to summarize the world's wisdom in a sentence, which they determine to be "this too shall pass". Then he requires a single word summary, which they deliver as "maybe".

Pete tells it much better than I've tried to. And you really wouldn't want to hear my trying to sing the musical interludes :-)

I couldn't help it and copied/pasted it into chatgpt as a prompt: https://chat.openai.com/share/a862f19c-aa2b-4f77-bcd7-a27b5bef6eb9

No, I understand the idea of magic self-executing contracts. It's a common approach to building a magic system.

I dislike the Craft series because of the writing style. It's simply not clear how the magic system works, and so the books are full of Dei Ex Machina. It's not clear what the limitations of the abilities are and what are the consequences of actions.

The world itself is also inconsistent.

https://what3words.com/ gets a lot more fun in a world like that. Sort of like a memory palace but with more unexpected exits involving terms like {arrrrggh} and {crunch}.

When I'm running D&D games, I make sure to tell my players "For the sake of ease and convenience, we will pretend that all of the coins you get are copper, silver, gold, etc. However, in reality there are a bunch of different coins with wacky values, and none have a fixed denomination. Cash money you find is made up of thalers, dollars, pounds, louis d'or, florins, marks, crowns, et numerous cetera."

At least most modern games, starting at least with 'Vampire: The Masquerade' in 1991, mostly abstract the money and other property of the characters, usually by some attribute or something. I like this approach.

There are games where you can calculate the exact change the characters have, but it's some kind of old school game then (yes, D&D 5e is that). Depends on the genre, too, sometimes you have to keep track of every arrow, but sometimes the wealth of the character is a skill which can be improved (or reduced: 'Diaspora' has a skill for wealth and an upper limit on how many skill levels you can have, so it's perfectly fine to get worse in something.)

Yeah nope, that's about as funny and amusing as a three-day-ripe haddock to the face.

I think if you're focussing on the mechanics of the magic system rather than the depth and richness of human relationships in those books then you're kind of missing the point ...

Presumably the imps manipulate physical things and not just information. It's not like real-life programming where you need a smouldering heap of CNCs, 3D printers and IoTrash devices to do real things, all facilities that are easier to mess up than the code. And "everybody" can program imps.

Therefore it's a slave economy, approximately in the Roman sense, as discussed on this blog previously. Most of the population are unemployed. But this is not urbanized Rome, it's a rural economy with no dole, no land ownership for the masses (therefore missing the chance of imp-automated smallholdings everywhere), and no law enforcement except for occasional hack-and-burn sweeps by the palace crowd.

Freelance imping doesn't pay (contract imping does, but boooring), there's no work and people gotta eat. So they go adventuring. The world fills up with with muppet-grade, level-1 dungeon parties and you just know that they won't live long enough to make level 2. Darwin awards may be in order. It's an un-heroic questing scene.

This could be written as comedy (but harshly in the shadow of Pratchett); or a story could focus on the few competent adventurers with a good reason for questing. I'd quite like a fantasy book where the heros' problem is not monsters in an empty landscape but amateur competition.

And then you get generalist tools.

That reminds me of a development system at work a couple of decades ago. There were these Unix servers which were used to build the stuff. The version control was, I think, CVS - where there is a possibility to check some code out so that nobody else can work on it while you have it checked out.

Now, of course somebody checked things out, worked on them a bit, and went on vacation. This meant nobody could work on that part without removing the lock file. Admins could probably do it, but of course this was a long and laborious process just to delete one file.

So, we had also 'sudo' (superuser) rights, but only for certain necessary commands, and 'rm' (remove file) was not one of them. However, we quickly figured out that 'make' was an allowed command, and so the problem was solved. Also our faith in the IT security was diminished by a fair bit.

(Explanation for those who don't know what 'make' is: it is a tool where you write a file of 'recipes' for doing stuff, so it's possible to do things like 'make install' which would perform all the steps for the 'install' recipe. The nice thing is that you can basically write whatever shell commands you like to the recipe, so we could add a recipe with 'rm lockfile' which did what it was supposed to do even if we were not allowed to run 'rm'.)

What can this lead to? To bloody wars, that's what!

Look at my left arm, kiddo. Oh, you can only see its shadow? Must be I lost its physical reality somewhere… Yeah, a bloody demon took it, along with half my team, back in the Functional offensives of '43. At night I still wake up sweating, remembering how they grew bigger and bigger by eating themselves. By their own tails, they self-devoured, stacks and stacks of themselves until they consumed all available space and them bam! You were lucky if you still had some parts of your body left for your soul to wear.

Three sons I had, lost two in the Great Editor War of '65. The youngest came back, but I can see that far-away look he gets.

I only worry about my grandkids nowadays. Bright lads, happy, they've only known peace. But I hear the Agiles are massing in giant Scrums not far from here and I fear. I fear for them and for all of us. Damn you Baggage, damn you Turing and damn the fucking holy Ada! You heard me! They messed with things we were not supposed to, and now it's people like me and mine who carry the scars.

The mechanism by which magical societies never stop being feudal is easy: Magic destroys the ability to create science, because it means that the results of any experiment won't necessarily be the same as when you did it the time before. So you get to a high medieval society, maybe with some good medicine, then your population crashes in a series of ugly wars, each one different because it's fought according to whatever theory of magic is fashionable at the time.

Charlie Stross @ 76: I think if you're focussing on the mechanics of the magic system rather than the depth and richness of human relationships in those books then you're kind of missing the point ...

Yes, the same goes for my whinge above about the lack of an industrial revolution in the Bujold's World of the Five Gods. Bujold writes wonderful character-driven plots, so one can forgive the occasional wobbly set.

It will depend on where the genes inserted into the weed came from :-)

Troutwaxer @ 80: Magic destroys the ability to create science

That's the usual cop-out: magic and science/tech are completely immiscible; either magic automatically blocks any tech (like in Shadowrun and the Potterverse) or tech destroys magic (also in Shadowrun).

I'm interested in a system where that doesn't happen. Given that technology is just a particular arrangement of matter, the presence of magic shouldn't affect it unless it is specifically aimed at the matter in question. This doesn't need to be destructive: a computer that interfaces with a spell should be perfectly possible.

The Discworld has bits and pieces of this; momentum and energy are conserved even when magic happens. The Five Gods also has it; magic is done via demons that inhabit a sorcerer (and are prone to possess them, which is a hazard). Demons are creatures of chaos, and the core of their magic is the manipulation of entropy. They can increase entropy ("downhill magic") very easily, causing things to decay, rust, break etc. They can also decrease entropy, but this is very hard work. It doesn't break the 2nd Law of Thermodynamics because the resulting entropy is only stored and has to be shed at some convenient point. One character finds slaughterhouses very convenient for this; he simply dumps the excess entropy into a beast due for slaughter, and it drops dead on the spot.

Yeah :-( What's more, providing evidence that it will reduce the time to final release, and reduce the maintenance costs thereafter, isn't acceptable, either - because it will increase the time to the first beta release.

It would be extremely interesting (seriously) to see what effect daemonic possession on the people responsible for releasing buggy programs would have. If the demons took just the programmers, the managers would simply hire more, so we would end up in the semi-slave states so common in fiction and fact. But, if the demons took those responsible for giving the orders? The social effects of such a thing would be far more drastic than appear at first glance.

This is my personal hell. Trying to convince people that their ten thousand line boilerplate type system code catches zero bugs of relevance, but costs thousands of work hours in maintenance every year.

I don't think magic and science are immiscible. I think that the presence of magic could easily mess with the foundation of science which is the idea that scientific experiments should be repeatable. This doesn't mean that science is impossible in a world where magic exists. Obviously physics or chemistry, even at the atomic or quantum levels, still works unless a magician intervenes, but I think the scientific method is unlikely to be developed in the presence of magic.

One character finds slaughterhouses very convenient for this; he simply dumps the excess entropy into a beast due for slaughter, and it drops dead on the spot.

And should be at least partially cooked

If you pay in heat, you could use magic to generate electricity to power your non-magic stuff while doing magic. It would be interesting to see what happens as technology develops along that pathway: people are required to do magic to keep civilization running?

Hacking. Either through introduction of demons or similar molecules.

Major magics might need to be done in a clean room / sanctified space.

Also, highly exponential spells unless there is a limitation on acquisition of heat energy. Cities end up with disruption effects everywhere. ( Or nonexistent). Also, no wooden buildings.

Spells where 95% of the power is in anti interference provisions. Combat spells being wildly inefficient as a result. Area vs volumetric safety, sabotage by being inside / close to a ritual circle.

Spells casting relatively slowly (demons per second?) Poor little critters.

Multi threaded spells, with the requisite very high requirements in casting complexity

Sentient demonic familiars as error correction

I assume Hell would publish altered versions of these open source libraries with logic bugs that will cause the caster's soul to be dragged to Hell.

As a result, magicians would be more cautious about supply chain attacks than programmers are. (Most bugs hurt other people...)

"read an instruction from an input, write an output, and execute a very limited number of tasks ("put this atom on top of that atom", "pause", "resume", "halt and catch fire" ...)"

OK, so this demon is basically a Turing machine, or at least the head bit of one; and as we all know a Turing machine can compute anything computable... but nearly always with grotesque inefficiency and extravagance in storage. And if they're operating at the level of moving single atoms, well there are a fuck of a lot of atoms in even the tiniest thing that's at all useful. To rearrange atoms at a rate comparable to, say, a school-lab-scale demonstration of electrolysis - ie. you can see it works, but it takes hours to get even a small bubble of gas - the demon is going to have to have a clock frequency at the very least at the low end of the gamma range, and that's with some very generous assumptions about its operating environment.

You're not going to be feeding it on coal. Chemical reactions are several orders of magnitude short of that kind of energy density. This is the kind of beastie that considers a lump of ⁶⁰Co to be a tasty snack. It's possible that in a state of nature they can survive on coal by extracting the radium and other uranium decay products in it, but the concentrations are so low that nearly all their atom-shifting capacity is used in digging out their food, so there isn't any useful surplus to exploit in captive ones. I suppose you could try setting up your magic lab in the cave underneath Chernobyl where that gamma-eating fungus lives, but it almost certainly won't be worth the hassle.

So despite the awful warnings against it, the only way to get a useful processing rate while running them at chemically-practical clock speeds is indeed to go for massively parallel arrays. Each demon only shifts atoms slowly, but there are lots and lots of them so the overall rate comes out as something appreciable. And of course there must be some overall loose interaction between the demons to keep their actions properly coordinated.

We know this kind of setup. The operation of the demons on an individual level is a matter in the realm of software engineering, but the complexity of the interaction between billions of them in a lump is way beyond what software engineering techniques can handle, and trying to predict or constrain the overall behaviour by software adjustments on the scale of the individual units is extremely unreliable, involves a lot of guesswork, and often gives quite unexpected results.

For dealing with such massively parallel systems we have to use a much more "top down" approach to keep things tractable, which helps, although it still doesn't do brilliantly at giving results with an engineering-level degree of dependability. Biological systems are, after all, well known for unexpectedly catching you out.

So I think what you've invented here is a kind of coal-powered bacterial slime, only made of demons and with explicit computational and synthetic abilities.

Re: 'Each demon only shifts atoms slowly, ...'

Only if demons are part of our 'natural' universe. Demons intersect with our universe at some spacetime points but they're like dark energy and dark matter: the math says they're out there but a determination of their composition and attributes is still in the to-be-discovered phase. OOC - has any fantasy author ever put together a theoretical basis of magic? (I'm talking at the subatomic particles or at least table of elements level not just a few sounds and a random wave of a wand.)

About wizards suddenly discovering that they're a wizard at some random age and then being able to perform magic without any training. Even if magic is an inborn/instinctual sense like a vision, hearing, wizards would still need some sort of training just like muggle babies are taught how to focus on different shapes/colors, sounds, etc. in order to identify and use their senses effectively.

a school-lab-scale demonstration of electrolysis - ie. you can see it works, but it takes hours to get even a small bubble of gas

Faster than that. Maybe ten minutes to fill a test tube.

Even if magic is an inborn/instinctual sense like a vision, hearing, wizards would still need some sort of training just like muggle babies are taught how to focus on different shapes/colors, sounds, etc. in order to identify and use their senses effectively.

Like Peter goes through in Rivers of London?

I've now got a mental image of a group of Prachettesque robed and hatted wizards in a magical preschool, with a Matron talking to them slowly and with careful enunciation…

A thought for the day:

Instead of polarizing reality into good vs. evil, as done here, what about polarizing it into intelligence (heaven) vs. stupidity (hell).

So if magic is a system of pacts with demons, maybe instead it's about pacts with abysmally stupid spirits, the kind that would make a bad deal with a con artist magician instead of being smart and doing the right thing?

I think all the other assumptions, including the need for rigidly defined programming/magical languages, can be left more-or-less intact. As for what an abysmally stupid demon would want in return for work, I'll leave that as an exercise for the reader.

Now, a society that has magic of this ilk probably would be stereotypically medieval in the modern sense, especially in kingdoms that employed a lot of magic to bolster the local authoritarian leader. For some reason, I'd expect orange to be a popular heraldic color in those places....

Conversely, the role of temples devoted to intelligent deities might be rather more interesting. Assuming the gods want to have anything to do with humans.

The thing is, if you want to focus on relationships, then a generic "dragons and wizards" soft-magic fantasy setting is fine. In this kind of setting nobody really expects you to be super-consistent, and that's fine.

Craft Sequence is in a strange situation, where the author tries at the same time to both investigate the impact that magic would have on society, and not really let it dictate the plot. As a result, it kinda fails on both accounts.

A good example of the first kind of fantasy where the author focuses on the impact of technology on a magic society would be the late Terry Pratchett books. Or the utterly depressing Baru Cormorant books.

what about polarizing it into intelligence (heaven) vs. stupidity (hell).

Surely you mean heaven: unimaginative people happy to just be somewhere pleasant. Hell: surrounded by whiners, complainers, people who want to speak to management RIGHT NOW, and ... people who have a great idea for improving things.

Surely you mean heaven: unimaginative people happy to just be somewhere pleasant. Hell: surrounded by whiners, complainers, people who want to speak to management RIGHT NOW, and ... people who have a great idea for improving things.

Think of it this way. In the beginning was the Code, and the Code was with God, and the Code was God. God used the Code to create a virtual machine on which the world runs, and created Gods within the virtual machine to write and maintain the code of the world. And they, iteratively, created virtual machines to make reality, and so on recursively in a strange loop.

In this reality, heaven is where the Gods work to make and maintain the Code that runs reality in its own sandbox, and the Code(s) of reality. This is, so far as I can tell, nerd heaven. There's also a lot of spirits who are creating by rote and merely running code. I'd call this nerd hell.

Hence I'd say that heaven is intelligent and hell is stupid. Magicians are stuck working with demons in this world because, basically, they're mid-level routines that have become borked, started acting like viruses, and are trying to get more privileges and resources than are allotted to them. Systems more intelligent than magicians are--Gods and their nerds de camp--are unlikely to work with them, except in a cybersecurity context. Thus the magicians by default have to work with demons, who are stupid.

Here endeth the lesson. Any parallels with any big tech firms on Earth are strictly coincidental.

Pigeon @ 91: OK, so this demon is basically a Turing machine, or at least the head bit of one; and as we all know a Turing machine can compute anything computable... but nearly always with grotesque inefficiency and extravagance in storage.

The issue with Turing machines is that data access requires stepping through a tape to reach the next bit. Also the universal TM is particularly inefficient because it has both instructions and data on the tape.

These nano-imps presumably have perfect recall for the instructions given, so each one is just a simple TM manipulating data. They are also not limited to a tape: you could give one a 3d crystal and tell it to move in 3 dimensions to manipulate the atoms. So that makes things much more efficient.

For the computerised creator Olaf Stapledon in Star Maker created an efficient heaven and hell. His hobbyist creator making experimental universes tried out one with two associated tieless sub - universes. One had an eternal moment of anguish and the other an eternal moment of bliss.

Actually, there are two meanings to 'Turing machine'. One is the original tape model, and the other is any machine that can compute the same programs. The latter is actually more common, and there is no particular reason to favour the tape model for demons.

As Pigeon says, managing even mere millions of entities is beyond the state of the art (whether in computing, business or politics). Structuring the management can just about keep it under control (a few countries' military were the experts here, but haven't done it in some time), but the emphasis is on 'just about'.

If we assume that demons have marginally more 'independence' than logic circuits, or even just behavve a bit like 'quantum' computers, the consistency problems get mind-boggling. I have mentioned before that even deterministic consistency is hard for the top experts. If 'quantum' computation ever becomes practical, the consistency problem becomes massively worse.

While (in theory) this WOULD make a good basis for a fantasy world, it would be unmarketable. "You say that at least a hundred people might be able to understand the plot?"

Aside: a long time ago, Intel was thinking about going in for massive parallelism on a chip (technically quite easy, with very simple CPUs in a 2-D torus), but backed off. Most people said that there was no point because we didn't know how to use it, and I could not persuade them that this was because the happy hackers (professional and amateur) hadn't had a chance to try. My guess is that is why Intel backed off.

For a system that is less software engineering and more straight up engineering there's the magic from Michael Scott Rohan's Winter of The World series. There all magic is through items made by mage-smiths.

It also has an interesting inversion of the usual law gods=good and chaos gods=evil setup.

In my own universe I wrote the following, but I didn't really follow it through as a plot point (my knowledge of computing doesn't go that far.)

“...magic, which is the essential energy that controls other energies, can only do five things; amplify or reduce, slow or accelerate, direct other energies, open and close portals, and make changes which are otherwise physically, mentally, or spiritually possible. The more complex magics work in a fashion similar to the way all higher mathematics are generated out of addition, subtraction, multiplication and division.

“Magical energy is created by the growth of plants and animals, particularly by the growth of a mature Elven forest, by making music, particularly with an audience, and – I know you Humans shy from discussing this – by making love.

“So when I healed Albert, I didn’t do anything impossible. I simply set his body to working on the problem much more quickly than is normal. Essentially, I amplified and accelerated his healing until it outpaced his dying.

“Further, the kind of magic someone can do is deeply tied in with their inner being, and their ability to express magical needs in a fashion which is in accord both with reality and their inner nature, and that frequently means that the magic one does is related to their obsessions, which results in the different magical specialties such as alchemy, enchantment or divination. Gran-Gran is a necromancer, for example, because she’s obsessed with things like anatomy and the processes and mysteries of life and death. She says that if you want to undo someone else’s work you must either understand them deeply or know what they said and did to shape the energy.

It's probably not terribly logical, but I'm guessing a real computer scientist could take this a lot further than I do (as mere background geekery.)

Ted Chiang's "seventy two letters" featured a universe that ran on a mix of "magic" and thermodynamics.

A magic system that apparently lacked supernatural entities and was subject to investigation up to a point.

The degree of data compression was quite hard to swallow though. I feel that information theory would hold just about anywhere.

Some really great ideas in the comments here.

My concern is that most versions of this make it way too easy to destroy the universe. In fact, so easy that it would certainly have happened before most of the interesting possibilities suggested.

Maybe if heat is required, then the mages who make mistakes end up draining huge areas of all heat, creating a 'always winter, never christmas', or an 'ice queen' scenario, where they are bound to their barely functioning spells, still running, but incredibly slowly, surrounded by a frozen landscape.

Having it so that without the mage within some radius of the imps means that they lose interest and wander off could be good as well, forcing mages with more going on to stay in their towers, and explaining why mages who have run away magic cataclysms don't move on and wreck other areas too.

Fred Saberhagen's Empire of the East series had a wizard who was using a bound demon to pursue science experiments, rediscovering the ancient arts of physics and chemistry.

On his command the demon magicked up a balloon that obstinately refused to float. The demon explained that air had weight, and was ordered to subtract the air from the inside ...

All this complexity is going to be beyond the reach of most people

And there ARE intelligent demons out there that will happily manage it for you

For a price of course

The most likely outcome of all this is a major demonic incursion eating the world as one of the smarter ones gets loosed

Most demonic incursion plots bother me, because it's the sort of thing you would expect to happen almost immediately after demons are discovered, well into the past of most plots.

I'll give Ian Tregellis a bit of a pass for Milkweed as there was a very small number of careful warlocks, and things immediately went to crap when their numbers increased.

...or the evolution of collective intelligence among sufficiently large groupings of individually-stupid nano-demons, so although each demon itself is not intelligent, a blob of bacterialdemon slime is.

One could well imagine also that the collective intelligence is more blindly nasty than you would expect a normal "macroscopic" individual demonic intelligence to be. I mean, corporations are bad enough...

It could be, as well, that as the blob of demon slime grows larger, the increasing communication delay between distant parts of itself starts to limit the ability of the array as a whole to maintain coherence. So you get local instabilities, which grow into distinct sub-intelligences, and then perhaps the blob fragments into multiple smaller blobs each of which has its own idea.

Or perhaps it evolves some kind of supervisor command level and creates a hierarchical structure which enables it to maintain consistency over a larger scale. So this keeps things together for another few size-doublings and allows the blob to develop a higher degree of overall intelligence. Until it reaches its enlarged limits, and once again needs either to fragment or to develop a further level of hierarchy... and so it repeats.

Most demonic incursion plots bother me, because it's the sort of thing you would expect to happen almost immediately after demons are discovered, well into the past of most plots.

I just read a book which has an unusual twist on a demonic invasion. Demons do not want to be in the mortal world, and return to infernal realm as fast as possible. Unfortunately, their most expedient way to go home is to grab onto nearest living, intelligent non-demonic creature. A portal to Hell opens, and they both disappear.

This kind of demonic invasion is self-limiting, but can wreck a lot of havoc in the process.

I think the answer to the demonic incursion issue is that "Hell is other demons". If a demon kills and take the place of a careless wizard then he is going work hard to stop other demons managing the same trick - after all, a world full of demons is Hell.
This also means a demon would seek to trick a good wizard rather than an evil one - the former is unlikely to want to leave Heaven whereas the latter has become a very capable and highly motivated denizen of Hell.

If we instead have the generalised "pool of magic" resource model then any apparent Magical Intelligence has to be the result of increasingly complex combinations of spells/incantations; until one day someone creates ChantGPT...

"Hell is other demons"

For some reason I thought no, hell is for children. And then I looked at the lyrics for that song and I thought... what the hell did I do that for. Sigh.

I share the generalised concern Disney and folks here that algorithmic magic would suffer all the same problems as software but without the constraints that limit the bad effects of software. We already have the corporation version of AI slowly killing us, the idea that some supergenius could white plague us using magic instead of GE. Hence my claim above that hell is full of people who have really good ideas for improving things. But no idea what Chesterton's Fence is.

Maybe the cyclic version of the big bang is caused by overconfident minions typing sudo shutdown -r now or whatever... urk.

Hence my claim above that hell is full of people who have really good ideas for improving things. But no idea what Chesterton's Fence is.

See the problems you get into with strictly Christian reasoning? Your notion could also bolster the claim that stupidity is more dangerous than evil.

As an Australian, you might imagine Hell is the sacred site for Evil Dreamers. Everything that exists, without exception, has a Dreaming. This therefore includes both evil and stupidity. Imagine the songlines for the Stupid Dreamers of the world, there to allow stupid people to reproduce in sufficient numbers that they don't disappear. Ditto for evil. And good...

Chesterton's Fence is, at best, misleading - and is generally used to hide bigotry. There are a lot of 'fences' that have no purpose, so that demanding a proponent of change identify their purpose is equivalent to vetoing any change.

One reason for this is that they may once have been thought to be a good idea for some putative reason, but because that was not clearly recorded and was not based on verifiable (surviving) evidence, nobody will ever be able to identify the reason.

However, claiming that Chesterton's Fence is false is equally bad. His arguments are correct for those 'fences' that have an extant, but unobvious, reason - and there are a lot of those, too.

Actually, there is a lot of evidence that stupidity IS more dangerous than evil!

Seems weird to ask for a non-Christian imagining of a Christian concept. But I suppose Ratana managed it with the Christian god so why not have an Australian conception of hell (other than "English weather and an English Prime Minister").

I'm more thinking that a good place to be is organised (not really low entropy, more 'there are consistent rules'), and a bad place is chaotic/any given rule only applies to a limited space-time domain. So hell is full of random events and stuff that only works sometimes, where not just the conditions but the mechanisms are subject to change at any time. Very like the Christian heaven. That's likely me projecting, though, since I strongly dislike "guess the rules" games...

You can see that as strictly Christian conceptions of evil, hell whatever you want. But painting dots on it and calling it a Dreaming doesn't make it aboriginal, any more than John Oliver calling our leader the "President of the United States, of Australia" makes him the actual president. If I was to imagine an aboriginal conception of hell I'd more likely look over the fence and think, a bit more concrete, a few more cats, some more global warming, we can make this awful enough to count as hell. But as far as I know that's not even vaguely how aboriginal spirituality works, either in having a hell or what it would look like if they did.

I don't think stupidity is disconnected from evil, all too often stupidity is good intentions applied with great skill and enthusiasm. The liberation of Afghanistan springs to mind. Russians or USA, take your pick (and that's an inclusive or there). Evil merely describes the outcome...

EC
While (in theory) this WOULD make a good basis for a fantasy world, it would be unmarketable. "You say that at least a hundred people might be able to understand the plot?"
Oh dear, already been done - by Hannu Rajaniemi.

.. @ 114:
Actually, there is a lot of evidence that stupidity IS more dangerous than evil! - if only because, all-too-often any significant amount of stupidity will "release" a more-than-siginificant amout of Evil(s) - AS WELL AS the damage it causes directly

It is certainly possible that is what he was attempting to achieve, but I would disagree that he succeeded (at least in the Quantum Thief).

Aren't you kind of falling for the whole Christian dualism nonsense about an afterlife, heaven, hell, etc?

There are plenty of mythologies/superstitions/religions that don't posit an afterlife at all. Or that don't automatically connect demons/spirits/kami/demiurges to a Good Place or a Bad Place.

Let's get creative folks. (And no more Christian inanity about temptation, sin, redemption, etc. It's just extradimensional magical entities that want to eat you.)

So if magic runs off of heat it seems that thermodynamic considerations must be relevant. Heat by definition cannot do work under isothermal conditions; you need a temperature difference to make a heat engine work. It doesn't seem the power source is a flow of heat OUT of the imps' native domain, as if that were how it worked, you'd want to refrigerate your pentacle as much as possible (covering it with helium II would get you nearly 100% efficiency); you wouldn't be talking about coal. So instead it seems you have to raise the local temperature to hotter than the imp realm. But just how hot is that?

(An old engineering joke turns on the assertion that hell has a lake of molten sulfur, which sets an upper limit on its temperature—though that does assume pressure comparable to that of Earth's atmosphere.)

The energy needed to store one bit is proportional to temperature. If the imps come from a hot realm their energy storage medium must be highly resistant to thermal degradation. Though I suppose they might function much closer to the edge of random change than humans do; if they reproduce, for example, the offspring might be subject to macromutations. (I am not assuming, here, that "information storage medium" must be a form of matter. If they come from a different universe they might have different ontological categories.)

I should be interested to see such issues addressed.

Try this:

Demons / angels / 'things' eat concentrated complexity / order / information, and so target the most disciplined and skilled people, but it's humans that say whether they have been raptured by angels or devils. With skill, such entities can be summoned and bound to do a service in return for some food, of which a human sacrifice is the easiest large source.

This extends to computers, but they can't easily be fobbed off by copies, because two copies of one machine or program contain no more information than one. So it's only the first one 'sacrificed' that is worth a major service - and the same is true to some extent for humans.

This adds a new excuse: "My homework was so good that a demon ate it." :-)

I question the whole concept of magic having "laws" and "rules" concerning the source(s) of magical power, specific incantations yielding specific results in a consistent manner, certain procedural rites having to be followed, etc.

The whole point of magic is that it breaks the laws and rules of physical reality.

But then to claim that magic itself requires laws and rules that have to be followed would be internally inconsistent.

The very concept of magic is that it is inherently chaotic whether you turn some poor guy into a newt ("I got better"), cause your neighbors crops to die without meteorological cause, defy gravity (Broadway reference) by flying on a broomstick, control others via an all powerful one ring, etc.

A world full of practicing magicians would be constantly altering the laws of physical reality to the point where the universe would no long have any meaningful laws.

And that chaos would have to apply to magic itself so that if you want to become invisible on Tuesday you rub yourself with a mixture of eye of newt and salamander tongue, but on Friday you would need to perform a sacrifice to the elder gods.

You want magic, science will provide all you need.

Demons and angels, ghosts and poltergeists? Self aware software avatars lurking in every program and corner of the internet.

Talking Disney animals? Uplift species to near human intelligence.

Magic mirrors telling you that you aren't the fairest of them all?
Your toilet can analyze you waste, declare your blood sugar to high and that you are overweight, and then lock the pantry door so you can't get at those left over brownies.

Magical all powerful sword? Edge weapons with a monofilament edge consisting of a single linear carbon molecule.

Talking furniture? Every couch and armoire having a microchip and software intelligence.

Crystal ball to predict the future? Apply chat GPT software as a Delphi oracle and ask it about the future (already done for global warming).

I am reminded of a Far side cartoon:

"Welcome to Heaven, here is you harp."

"Welcome to hell, here is your accordion."

A world full of practicing magicians would be constantly altering the laws of physical reality to the point where the universe would no long have any meaningful laws.

This is kind of one of the main drivers of the old tabletop RPG 'Mage: The Awakening'. The rules of the universe there depend on 'the consensus' (even the game, in its multiple versions, is not very clear what 'the consensus' really is, though) and the main factions are the 'Nine Traditions' of mages trying to do more magic and the 'Technocracy' trying to suppress the 'reality deviants' from breaking the world with their ideas.

It's more complicated than that, of course, but there is the fear that magic use will just unravel the whole universe. The Technocracy are using the same magic but in denial, and of course there are rules for the magic because they are kind of useful for a roleplaying game. (Though the game's system is pretty free-form.)

Not everyone belongs to that (extreme physics) religion, you know. Let's stick to the 'real world', and demolish a few of that religion's dogmas.

Firstly, there is some reason to doubt that what we think of as reality is an immutable absolute, even in this universe. Almost all physicists close their eyes to that one, even when they are developing theories based on our current physical laws having been created. Yes, I am referring to the Big Bang Bullshitters.

Secondly, there is no evidence that the current theories are the One True and Final Truth of how the universe behaves. This is not the first time that has been claimed, it was always wrong before, and there is evidence that it is this time, too.

Thirdly, chaos is not and has never been an absolute. A total absence of any laws whatsoever is not an observable or even philosophic situation. All observable or philosophic chaos is constrained, and constraints are laws.

I know of no magical system throughout history (and very little fiction) that wasn't based around some variant of using extra or alternative laws, above and beyond the simple 'physical' ones.

EC @ 125
Firstly: Big Bang Bullshitters - REALLY?
What's your explanation for the Cosmic Microwave Background, then?
Secondly: Agree, however - the Gen Rel / QM ummm .. mismatch has been telling us that for years.

I was not referring to the Big Bang theory as such, though there ARE other possibilities, but the people who claim that the evidence for that proves that our physical laws and constants were created in the first pifflisecond.

Let's get creative folks. (And no more Christian inanity about temptation, sin, redemption, etc. It's just extradimensional magical entities that want to eat you.)

I'd point out that L. Sprague De Camp, along with Fletcher Pratt, played with these ideas in multiple stories, so for those not familiar with works like The Land of Unreason, it's probably worth checking them out. Yes, there are whole subgenres on this, I'm just pointing back to the ones from the 1950s.

Anyway, much as I like ecosystems based on who munches who, humans aren't a great food value: we grow too slowly and aren't efficient at turning cheap food into good meat. You'd want pigs for that job.

However, if one posits a paranatural polarity of smart to stupid, with "Gods" on the smart end and "demons" on the stupid end, what place do humans with paranormal powers occupy?

It might be worth exploring the following analogy: magicians are to demons as gods are to clerics. What then are the relationships between gods and magicians, and clerics and demons? If magicians use a mix of software engineering and manipulative psychology to get demons to work for them, what approach do gods use to get clerics to work for them? The same ones? Or others?

Anyway, much as I like ecosystems based on who munches who, humans aren't a great food value: we grow too slowly and aren't efficient at turning cheap food into good meat. You'd want pigs for that job.

This is assuming, of course, that the food in question is purely biological rather than based on some informational/metaphorical/organizational aspect. Think targeted adtech as demon gourmets. They want to harvest specific informational patterns/emotional responses. In that case there's a little bit more reason to want to raise intelligent and social animals because otherwise you might not be able to produce the full range of flavors.

I swear I've read at least one story where there were demons who will do a task but will take your ability to feel certain sensations/emotions as payment, but I can't remember the title or any additional details. Anyway, that sort of concept would provide justification. It's like how humans don't necessarily care about entire plants, we just want to harvest the bits that taste good to us.

"The whole point of magic is that it breaks the laws and rules of physical reality. But then to claim that magic itself requires laws and rules that have to be followed would be internally inconsistent."

All fiction breaks the laws and rules of physical reality, because none of it actually happened. If it only followed physical reality then it would be non-fiction, ultimately either physics or history. :)

The point of well-conceived magic systems is that they add to the laws and rules of physical reality we know. The details of exactly why are always going to be handwavy. But any story needs constraints for the hero (otherwise it just becomes one line of "they wanted this so they got it"), so the magic system needs limits, and that will determine your handwaving.

I know of no magical system throughout history (and very little fiction) that wasn't based around some variant of using extra or alternative laws, above and beyond the simple 'physical' ones.

Have to be careful here. AFAIK, the idea of magic having laws goes back only to Frazier's Golden Bough. IIRC, that's where we get the idea that magic works on Similarity ('like affects like") and Contagion ("once together, always together"). Frazier derived these laws empirically by looking at a lot of folklore and ethnographic literature, and trying to systematize the commonalities he perceived.

Magic, especially magic versus science, is in many ways a modern, WEIRD idea that's probably more at home in fantasy from the last 70 or so years than it is in reality. It's fun to play with, but you have to acknowledge that its underlying assumptions are questionable at best.

For example, in positing that there's a magical energy we'll call mana, we're saying that the mechanism that causes charms to be effective in curing warts is the same energy that makes Polynesian politics and religion work (they are one system, but we insist on dividing them), and that this energy also explains why Christianity conquered the world in ways that other religions did not. Yes, it's fun to ask "what if these things are all related?" But in the real world, there's no evidence that they are.

In the real world, people have dreams where spirits teach them songs. They sing these songs, and others feel better. Magic! Created by following laws? Not really. The closest we get to that is humor-based medicine.

Also, a bunch of renaissance magic really looks to be repurposing the old memory palace arts of the Middle Ages, at a time when paper and printing presses made them obsolete. In a time when books were handwritten on parchment, they were effing expensive. People traveled to read books and to memorize them when they couldn't take them with them. Things like illuminated manuscripts, with all those elaborate doodles, were done to help people remember every separate page. When all this was no longer necessary, because you could buy a copy fairly cheaply, those memorable doodles became demons and magic symbols. And now we're repurposing the repurposed magic, in order to write entertaining fantasies. Which is fine, as is trying to derive laws. So long as we remember that all of this effort is basically playing and have fun with it, we're fine. Taking it too seriously might get a bit awkward, though.

How about: the new enhancement of page A breaks pages B, C, and D?

When I was working as a contractor for AT&T, '06-08, we had one woman on the team who was our tester. Including regression tests.

I could also note that one of my daughters has had a good career with Boeing as a tester, so unless your company thinks they're better than AT&T and Boeing....

Oooh, ooh, you missed the best point of all: the sum of all knowledge at X time.

That is, when wizards Tom, Dick, and Harry were each in the middle of testing their new spells, which then proceeded to fail spectacularly (and fatally).

Oh. Great. And with #81: so that's why I'm less enthralled with her more recent stuff - not enough plot.

I refer you to my recent rant, er, essay: character-driven vs plot-driven: a false dichotomy? https://mrw.5-cent.us/?p=309

Fer chrissake! The fact that those 'laws' were neither always codified (though sometimes they were), nor reliable (though occasionally they were), doesn't mean that they were any the less laws.

"If you don't sacrifice a cock to Mumbo at least once a week, bad things will happen to you" is a law. And thAt sort of thing is well documented from ancient times.

Cool! And, speaking as both a programmer AND a version control manager who still likes CVS, you'd have been on the verge of being fired.

Of course, the guy who left with it checked out with a lock would have been, too.

And, y'know, just MAYBE you should have gone to either the CVS manager, or your manager, who would talk to them, and they could UNLOCK the damn thing?

But if it was set up correctly, your builds would have pulled the last checked in version, not one that the programmer on vacation hadn't finished with yet.

I disagree. The whole point of the existence of grimoires tells us that a spell, once known to work, will always work, and is so repeatable.

You missed a great worldbuild concept: magic works by interacting with dark matter and energy.

And deities that want you, too? I am reminded of the cover of an old underground comic - it might have been The New Adventures of Jesus. "He's BACK! Even the grave couldn't stop him, and he WANTS YOUR SOUL!"

Here ya go: the imps' world is hot, and by doing magic in this universe, it cools it off, so we're just a radiator for a hot world.

Oh, and after I hit submit, I realized there's been a long-standing discussion in the imps' world, over whether the critters doing "magic" and calling them are actually intelligent, or whether they're just animals, repeating a formula. I mean, what do they ask for? A mate, a cheap element (gold), etc....

Sorry, but that's completely wrong. Magic must have rules, otherwise you can only do a spell once, no matter how simple it is. It has to be repeatable, otherwise your universe stops existing for no reason some day.

I think it was Lord Dunsany who said that writing fiction is eash - everyone knows what the rules are, and where you can bend them. In fantasy, on the other hand, you have to create all the rules, and you can never, ever break them, or you lose your readers for cheating.

Come on, why do you think churches abominate witches and magicians? Because they're unlicensed magic users, while priest and saints, etc, are licensed.

»You missed a great worldbuild concept: magic works by interacting with dark matter and energy.«

Which is why one of the final checkboxes on ESA's launch checklist for the EUCLID mission is to top up the Holy Water Container.

... you know, just in case...

I swear I've read at least one story where there were demons who will do a task but will take your ability to feel certain sensations/emotions as payment

Dan Wells's John Wayne Cleaver series (published 2009-2017) has something similar. Long, long ago, a bunch of people made a deal with powers. Each gave something up in return for immortality (of the 'does not age / senesce' variety).

Turns out that this was a case of 'be careful what you ask for, you might just get it'. The guy who gave up feeling emotions, unless he caused an emotion in someone else? He really missed having emotions. The person who could jump bodies because she didn't like her original body? She didn't like any body she jumped to.

That might've been it. I read the first book of the series back in 2010 or so.

»The whole point of the existence of grimoires tells us that a spell, once known to work, will always work, and is so repeatable.«

Does not follow.

It might also be to tell us "This one has already been used", if for instance the dæmons refuse to be "tricked" twice … or worse.

Precisely. The distinction between magic, religion, (usually) medicine and (often) science is an entirely modern invention. The reason that pupils apprenticed themselves to witchdoctors, shamans etc. was to learn the rules.

"If you don't sacrifice a cock to Mumbo at least once a week, bad things will happen to you" is a law. And thAt sort of thing is well documented from ancient times.

Depends on what you mean by law. What you cite above is a protection racket, which some have codified into rules. It's not like a law of physics, which is a reliable description of how something works in relationship to other things. With things like Similarity and Contagion, you can create new rituals. Your Mumbo rule doesn't allow you to create a useful system of Mumbodynamics, or even Mumobstatics, for instance.

AFAIK, the closest thing we have to classical rules of "magic" are the four and five-element/humor theories of medicine, whether from Galen, Indian, Muslim, Chinese... Much of the time, they seem to have been trying to group and codify treatments that empirically worked in particular cases. And similarity and contagion certainly show up there, too--they help make placebos work, among other things.

A protection racket? Hey, you just gave me an idea to sue Desanctimonious, because it's "if you don't agree with my version of "Christianity" (tm), we'll make laws against you.... RICO....

»Depends on what you mean by law. What you cite above is a protection racket[…]«

From friends how have grown up in mafia-infested locales, I can tell you that a protection racket is as much a law of nature as any other source of inescapable "this is how things work here" rules.

Just because it is criminal, or religious, or both, does not make it any less "a law" for the subjects.

I didn't say that. I said that magic means a scientific experiment is not always repeatable.

That's a very intelligent idea.

So? The idea that those are fundamental to laws is another entirely modern invention - and, indeed, even MORE modern than the separation of magic, religion and medicine. It's not universally agreed, even today.

If you were to tell a sub-Saharan witch doctor the currently-used practices for healing particular ailments, casting curses etc. do not follow any law, you would be laughed at (if you were lucky).

Indeed, trying to explain that your assertions are essential components of laws to a legislator or standards committee of some (unmentionable) programming specifications would get you nowhere.

You don't need magic for that. In most fields of science (even several branches of physics), an experiment is not always repeatable. At best, you can get statistical repeatability - at worst, it is not even theoretically repeatable.

For examples of the last, try cosmology and evolution :-)

Charlie Stross @ 118:

Aren't you kind of falling for the whole Christian dualism nonsense about an afterlife, heaven, hell, etc?

There are plenty of mythologies/superstitions/religions that don't posit an afterlife at all. Or that don't automatically connect demons/spirits/kami/demiurges to a Good Place or a Bad Place.

Let's get creative folks. (And no more Christian inanity about temptation, sin, redemption, etc. It's just extradimensional magical entities that want to eat you.)

Well, LE Modesitt's world of Recluce had magic based on order and chaos. There was an Order/Chaos balance, so using too much of either one would create problems. Concentrate too much order in one place would cause outbreaks of wild chaos in another place; too much chaos and there are outbreaks of so much order that everything would freeze solid ... delinking the order that held chaos within released tremendous energy (atomic bomb levels of energy).

I don't think that was Christian dualism because you could have good or evil on both sides of the balance.

DP @ 121:

A world full of practicing magicians would be constantly altering the laws of physical reality to the point where the universe would no long have any meaningful laws.

A world where magic exists would have different "laws" governing physical reality, but "magic" wouldn't change the laws; it could only work within them.

When considering modern science you're absolutely correct about the lack of repeatability. But even the basic, classical experiments which work every time (on Earth) if done correctly, which have nothing to do with quantum physics or other complex topics can fail in a world where magic is possible. So you can't even do the necessary work to develop the theoretical/practical basis for those more complex experiments.

"The point of well-conceived magic systems is that they add to the laws and rules of physical reality we know."

Then it is not magic.

It's newly discovered science.

As I approach the doors, I wave my hand in the magic pass, and the doors open before me before I reach them and have to wait.

Magic, or electric eye, like at the supermarket?

Any magic that works with physical laws or operates according to rules of additional laws isn't magic.

It's just another flavor of science.

Magic is something that breaks and defies physical laws, warping and altering reality as it does so.

Bend reality enough and it breaks.

Oh, this leads to Consultancies.

We’ll hire bright young overconfident people, preferably from outside the field of mathelogical demonology, pay them peanuts, and contract them to you to projects like yours at outrageous fees.

And if they make a few mistakes, well…. not MY soul on the line. Or yours. So, we got a deal then?

I would assume that most consultants were possessed by demons and act accordingly. Much the same as I do now.

You haven't worked with enough consultants. They burn through the brand-new-magic-degree kids in weeks or months. The ones that survive become senior consultants.

Magic is Money and Money is Magic.

Magic is intangible. Magic is vastly powerful. Magic follows rules that are at best an echo of any scientific laws. Nobody fully understands magic, how it works, why it matters. Magic will do what it wants and only occasionally - but very forcefully - be reined in by the general laws that govern reality.

There are those who are born with access to immense amounts of magic, who are able to do or experience anything they desire. Others spend their lives studying magic, pursuing it, and sometimes gain control of vast amounts. Most have little access to magic, or merely enough to survive and perhaps thrive. Yet magic pervades every aspect of life, and the magic cost of each activity or item is central to most lives.

Magic is capricious. Those who are born with or acquired vast amounts might find a sudden change in the rules or context of magic mean that their magic is worthless. Magic can be stolen, slowly or quickly, and even the most powerful wizard can make colossal mistakes with their magic.

Few people, even those with control of vast reserves of magic, understand how it works, how it moves through the world. The most powerful sorcerors can at best grasp and utilize a small trickle of the raging river of magic that sloshes around the world.

Governments of all forms require magic to function. In some ways they make magic through fiat, but are randomly subject to the same capricious semi-random changes to the rules of magic that everyone else must endure. And yet it is possible for some governing sorcerors to direct enough magic to move armies and shift mountains.

Success at magic breeds success at magic. The more magic you control, the more you can gain through similar efforts. Magic favours powerful magic users, and it is possible for one powerful sorceror to take control of some or all of the magic of countless minor wizards.

Entire schools rise to study the flow and use of magic. Careers are built on esoteric theories of magic. Sometimes various sorcerors seek to impose those theories on the function of magic in the world, but it never works very well.

The distribution of magic is not fair. The Potters and Malfoys who inherit vast amounts are no more deserving than anyone else. Possession of magic does not make one any safer - no matter how distant your tower or how many monsters and traps you create to protect yourself, there is always another wizard or adventuring party who want what you have.

Troutwaxer @ 152:

I didn't say that. I said that magic means a scientific experiment is not always repeatable.

That wouldn't work. Follow the protocol step by step (spell or experiment) get reproducible results ... doesn't matter whether it's science or magic.

The whole point of the existence of grimoires tells us that a spell, once known to work, will always work

Or at least that whatever is captured in the grimoire will generally do the same thing again if the grimoire is satisfied.

I'm thinking of grimoires as smartphones. For the most part you can power up an old smartphone and it will boot and make phone calls, most of the software will still work, but battery life will be limited (physical decay) and apps that rely on external servers might not work (infrastructure decay), but if you're unlucky OS updates will quickly render some or all of it unusable (software decay). "I gave it power and made the required gestures, but my spell did not work" pretty accurately describes many users complaints to helldesk.

~Sighs~

How hard is this to understand?

Suppose you're Gregor Mendel, meticulously documenting what happens when you breed yellow and green peas together? What happens to all your data-collection when the wizard next door decides to turn all the local peas purple because that's her favorite color? Or maybe the peas turning purple are a side-effect of something else the wizard is doing? Or maybe the wizard simply does something that renders the monastery garden infertile? If the effect is subtle enough, you might never know exactly what happened.

Re: '... worldbuild concept: magic works by interacting with dark matter and energy.'

Actually I was thinking about something like that when I mentioned differences in spacetime alignment.

I have no idea what the actual equation or its theoretical component terms mean in my ordinary life/lexicon but apparently the best fitting model for explaining our universe has 10 dimensions. If dark matter and energy (and maybe dark gravity?) are two or three quasi/half dimensions of our universe, then 'magic' for lack of a better term* could be one, two or four of the other dimensions. So in the demon scenario, this could likely mean that changes in any one of the ten 'dimensions' could be catastrophic for the entire universe because despite the appearance of random noise/change that we can see in our everyday lives, the universe is actually at a stable self-correcting equilibrium.

*Good grief - for a really bright bunch, physicists as a group are crap at coming up with original names for stuff they discover.

A few comments on/questions re: other folks' comments:

1-Order vs. chaos ...

How does evolution (or a hierarchy of demons) fit in if the demon's universe is completely chaotic?

If it's always chaotic/in a state of constant instability, then how do you explain that the demon entity coheres/doesn't fall apart, i.e., lasts long enough to perform a task?

Is there such a thing as time in the demon universe?

2-Rewards ...

I was considering emotions or time as a possible reward. Feeling emotions would have consequences for the demon though so because demons are typically described as having no perception of pain, I'd start with 'rewarding' them with senses.

I was also considering 'time as an experience' as a reward: if demons are eternal, then maybe they do not perceive time. They exist. The universe exists. Period. Boring as hell. This would explain why demons bother visiting humans - we're a novelty that's full of surprises because time is at the core of our very being. (A being outside of Time but that can interact with a being that exists in time suggests a really weird fractured wave theory.)

If you've got a really dumb demon, promise it some crypto: it's human techno-magic!

Cool! And, speaking as both a programmer AND a version control manager who still likes CVS, you'd have been on the verge of being fired.

Of course, the guy who left with it checked out with a lock would have been, too.

And, y'know, just MAYBE you should have gone to either the CVS manager, or your manager, who would talk to them, and they could UNLOCK the damn thing?

But if it was set up correctly, your builds would have pulled the last checked in version, not one that the programmer on vacation hadn't finished with yet.

Well, no. Nobody was really in danger of getting fired, we don't really do the thing were you do your job and get fired for that. The powers that be knew of that but trusted us, that's what we were hired for. (I think this would have been a place for an improvement suggestion if the project would have been for example ISO27k audited, but it wasn't.) This was a router platform for mobile operators, so there would have been better ways to get fired by doing some code or other, I think. Getting admin rights

Going via the proper routes to get an admin to delete the lock file could have taken a couple of days, so this was much easier. (Yes, yes, convenience and security are opposites.)

It was twenty years ago, so I don't remember the details, but at least lately most projects have used git for version control, which while not perfect by any means at least supports simultaneous work a bit better than CVS, in my opinion.

Science vs Magic.

My ideas are based largely on a bit in Black Easter, by James Blish.

Science depends on the basic components of the universe being mindless things, that do whatever they do, and respond to prodding however they respond, and once you know enough of how they work, you can either make them do what you want, or conclude that you cannot.

Magic expects that some, at least, of those components are agents, with wishes and desires, that you can, perhaps, bribe, bully, cajole, or coerce into doing what you want, but which cannot be relied on. The bribe that was enough last time may not be enough this time round, or the agent may have worked out how to defeat the means of coercion (they don't like being coerced).

Attempts to investigate magic scientifically tend to fail because the experimental subjects can, and often will, mess with the results.

JHomes

Attempts to investigate magic scientifically tend to fail because the experimental subjects can, and often will, mess with the results.

So magic is like social sciences or economics? Instead of physics?

It's not like we haven't tried to deal with this kind of stuff already, with various levels of success.

This discussion seems to be heading towards this month's "major panic" - the Threat to Humanity posed by AI
Haven't the ignorant idiots writing this trash ever heard of Isaac Asimov & the Three laws of Robotics ??
IF you are going to implement AI, then you need to put those governing laws/restrictions in place, yes?
Problem solved, possibly.

And what makes you think that is even possible? We assuredly don't know how to do it. Defining 'harm' is seriously non-trivial. Have you read "With Folded Hands"? That's the least of what is likely to go wrong.

To take a personal example, I describe chemotherapy as poisoning the cancer without quite killing the patient, and the consultants merely nodded. Yes, I was perfectly well aware that the treatment was doing me serious harm (I had to stop because of peripheral neuropathy), and might kill me.

The problem isn't with AI as such - it is with giving it untrammelled control. We simply don't know how to program intelligence or judgement (in any meaningful sense). And the actions of organisations like the banks or Home Office show that they are extremely unlikely to provide effective human oversight.

That sort of thing happens all the time in agricultural, biological and medical research. Experiments do fail or give misleading answers because of unexpected (and often unidentified) external factors.

Provided that it doesn't happen all the time, that's livable with. And there is no reason to believe that magic would mean that every experiment on pea genetics attracted a sorcerer to disrupt it!

As Robert Miles pointed out in one of his talks, all you need to do is correctly define what "harm" means, and at that point you have essentially solved ethics.

Aargh! I have used "all the time" with two different meanings. What I meant was that external factors cause trouble fairly frequently, and that an experiment works provided that they don't affect it every time.

The sort of external factor effect I am referring to is the bedding or food used for mice interacting with the treatment (actual cases), or the gene for colour being close to that for cold tolerance and there being a cold year.

I also thought about "Black Easter" almost as soon as this thread started, and the line where the magician points out that some of the most important forces in the universe are persons.

Mikko @172: So magic is like social sciences or economics? Instead of physics? It's not like we haven't tried to deal with this kind of stuff already, with various levels of success.

I kind of disagree. "We" (well, those who tried to analyze such things dispassionately) had no success whatsoever -- although they often THOUGHT they had success, -- until scientific method was developed, and then not until they recognized why exactly scientific method does not fully apply.

If you have to invent scientific method from scratch on nothing but social and economical interactions... good luck with that.

Haven't the ignorant idiots writing this trash ever heard of Isaac Asimov & the Three laws of Robotics ??

Greg, you can make a trivial end run around the three laws of robotics by redefining "human". As real world humans do that all the time ...

(More seriously, I hate Asimov's three laws with a burning fiery passion, and not just because they're trivially easy to mangle into non-workitude: they're actually what you get when a bright but probably ASD Jewish kid from New York in the 1920s tries to wrap his head around White Southern racism and chattel slavery and comes up with a flawless system for enforcing it. Thought experiment: consider a brain implant for humans that can enforce compliance with specified behaviour, such as Asimov's laws. Now replace "robot" with "slave", "human" with "slaveowner", and implant said implants into anyone designated as a slave. In what way is this not an ethical abomination, and what sort of abusive behaviours does this facilitate above and beyond simple enslavement?)

Your definition of "magic" gives you basically two stories. "I wanted this, so magic happened and I got it"; or as tragedy, "I wanted this, but magic happened and every I did was pointless". Once you've disconnected cause and effect, either you succeed for no reason or you fail for no reason.

That really limits your storytelling, because nothing beyond the most basic folk tale is possible. To quote Pratchett, "there's no such thing as a free goblin". If you want a story with more depth, you need some cause and effect. And the moment you have cause and effect, magic becomes science as per the inverse of Clarke's Third Law.

Re: '... consider a brain implant for humans that can enforce compliance with specified behaviour, such as Asimov's laws'

I wonder whether that specified behavior or reaction (feeling of reward/punishment) would eventually migrate to some other not zappable part of the surveiled brain region.

My impression is that Asimov came up with these laws to show via his robot short stories that relying solely on the literal interpretation of a small bunch of laws/commandments doesn't work. I like that he later had R.Daneel come up with the Zeroth Law which explained why R.Daneel left humanity to sort out its ethics/future on its own: don't dump your responsibilities (humanity) on someone/something else.

Troutwaxer@168: What happens to all your data-collection when the wizard next door decides to turn all the local peas purple because that's her favorite color?

So one of the basic precautions in any serious experiment is setting up the appropriate magic exclusion spell.

You might just as well ask in our world "what happens when the virulent purple pea plague hits your experiment?" Answer: biosecurity.

Charlie @ 179: you can make a trivial end run around the three laws of robotics by redefining "human".

Freefall has a lot of fun with this one. See for instance the sequence starting here: http://freefall.purrsia.com/ff1600/fc01581.htm

Correct.

Readers need internal consistency to make fiction plausible -- suspension of disbelief is contingent on the fictional world making sense (in a way that the real world is under no compulsion to follow).

You can't use magic in fiction unless you have an underlying rule system. It might be deep -- magic makes no superficial sense (but works allegorically or by providing a metaphor for some human experience), or follows dream-logic (because it's a dream, and dreams are just a Markov chain of internal references), or it has a complex systematisation of cause-and-effect like D&D or similar TTRPG magic systems -- but there has to be something for the reader to be able to lock onto and say "yes, this is internally self-consistent".

Which is why we get rigid magic systems in fiction. It's not a failure of the author's imagination, it's a necessary precondition for storytelling that the reader can follow.

To be fair to Asimov, what he hated was the old moralist formula of "Scientist builds robot, who then kills scientist (in response to his/her overweening pride.)" Asimov stated very clearly that he couldn't imagine general-purpose robots being sold without safeguards, and the Three Laws were the safeguards he imagined. This was his effort to bring the idea of robotics into the real world - you wouldn't build an appliance that might deliberately kill it's owner - and the laws were intended as pithy mathematical formulations that could be formulated into computer programs. (He can probably be forgiven for this assumption as the Three Laws were formulated well before anyone had done the work necessary to understand how difficult true AI might be.)

But you're right about the potential implications if you installed those laws into people, and once you have true AI (not the hideously flawed project that's currently polluting our intellectual universe) those Three Laws would become very oppressive.

There's still a problem, however. Science is ultimately more powerful than magic, as it would eventually extend itself to understanding magic. However, if magic can interfere with science, but science can't interfere with magic, then it's obvious to the ordinary eye that magic is "superior." So why bother with science when you can learn magic? The interference of magic against science isn't just practical, it also determines where to put resources.

Most people live in a magical world today, anyhow: it's how computers, the internet, the cloud, AI, and semiconductors work.

You say the magic incantation "Hey Siri ..." or equivalent, or draw a glyph on a magic slate, or tap buttons on a "keyboard" in accordance with instructions in a grimoire, and the magical boxes make stuff happen for you -- groceries turn up, the slate shows moving porn pictures, spirits come out of the walls to shout at you, and so on.

Oh, and if you poke the wires into the hole in the wall the wrong way up the magic smoke that makes stuff work comes out of the box and it stops working.

Push the wires in wrongly and you might let the magic smoke out of you. Almost managed that one myself as a kid.

• Most people live in a magical world today, anyhow: it's how computers, the internet, the cloud, AI, and semiconductors work.*

It’s worth considering the possibility that much of the modern tech world was designed by guys who grew up reading LOTR and playing D&D. The arrow of tech causality gets interesting at times.

…

Getting back to rules, Lewis Carroll just made a pink bunny in a Victorian girl’s frock pop out of my watch to remind everyone that good fantasy literature doesn’t need magic rules. Go ask Alice. Or deduce the laws of magic underlying LOTR without Tolkien’s notes.

Similarly, fantasy rules are more about expectations and prejudices than reality, whether it’s about Thud and Blunder, or about literary gatekeepers thinking that things like skin color, age, gender, or membership in clade Homo are what must determine agency in a story.

»More seriously, I hate Asimov's three laws with a burning fiery passion[…]«

But dont you pretty much end up the same place with UN's Human Rights Declaration ?

There are LOTs of animals who have a better case for being included, than a lot of humans, who through their actions reasonably could and should be excluded ?

In the end it always come crashing down when somebody tried to bright-line "Us" from "Them"

To me the ability and willingness to feel guilt seems to be much more important than DNA.

Lewis Carroll just made a pink bunny in a Victorian girl’s frock pop out of my watch to remind everyone that good fantasy literature doesn’t need magic rules.

Eh, if you haven't already I strongly recommend reading "The Annotated Alice" -- the edition with footnotes and explanations by Martin Gardner. Carroll's fantasy-land had an underlying logic to it which may not be immediately obvious on first sight but nevertheless informs it consistently. (It was also a satire on the "new mathematics" of the 30s-50s -- the 1830s-1850s, that is, new-fangled stuff like Boolean algebra and set theory instead of good old fashioned trigonometry and logarithms.)

Most people live in a magical world today, anyhow: it's how computers, the internet, the cloud, AI, and semiconductors work.

Totally. And since the real rules are so complicated and messy most people (tribes?) make up their own rules for the magic. That seem to be valid but fail outside of the simple cases.

Amulets to prevent EMF from making folks ill comes to mind.

My mother, born in 1932 and dead now for 9 years, made up all kinds of rules as she didn't like the real world rules we'd tell her.

One brother and his clan are big into magic (nonsense) rules about medicine and technology.

Yes. It's not perfect, but explains almost all of the apparent irrelevances, numerous satires of 'improving' works and sayings, mathematical / logical jokes and many of the puns.

"Or deduce the laws of magic underlying LOTR without Tolkien's notes."

Sufficiently advanced technology.

You've got a problem there: unless you're living, like Dan'l Boone, who moved every time he could see smoke from a neighbor's chimney half a mile or more away... you live in a society.

Judge: Defendent, you're charged by the monk from the abbey with changing the color on his peas. What is your defense? Defendent: I like purple, and I thought purple peas would be pretty. Judge: you will turn all of them back to their original color, and if you do it again, it'll be a month in jail, and no grimoires for a year.

"Too much trouble"? And it would take days? How big was this organization? If it was that big, it would get you fired, at least in the US. It should take an hour or two (depending on lunch, or when they come in in the morning.

And I hate that, with git and the others. IMO, if you need two people working on one file, than a) they should be sitting together working on it; b) that one file has too many functions in it, or it's one function that does far, far too much, and should have been broken up into multiple separate functions.

In other words, bad design. And I'm getting flashes of spaghetti code that I had to fix. A co-worker, around '92 or so, and I were in complete agreement: unless it's a zillion moves, no function should be more than two screens - that's 50 or so lines - long.

I'm astounded that no one here has come up with the instant response: Jack Williamson's The Humanoids, which he wrote explicitly to respond to the three laws.

Sorry, can't let you use that soldering iron, you might burn yourself. No, you can't use a saw....

Like the quote I gave that I think is from Lord Dunsany. We're talking about fiction, where things must make sense, at least within the context of the story.

Elderly Cynic brought it up.

192 - I feel your pain, My mother is the same, like "off" means "standby" on the landline, but "standby" means "off" on the washing machine. Or some vegetables have "tails"...

196(b) - And if the one subprogram does "far too much" but proves impossible to simplify?

"Impossible to simplify"? Um, right. Redesign, and in the meantime, I'm betting on spaghetti code. I, in my career, have taken a number of programs from 1000-2500 lines to about 300-400.

suspension of disbelief is contingent on the fictional world making sense (in a way that the real world is under no compulsion to follow).

Looks at Boris Johnson and ... what's the term for a whole sack of sock puppets? ... one of those of the Republican Party... and thinks about suspension of disbelief. "real world" my foot... I want a refund.

I kind of like authors who do realistically unreal things occasionally, but it does get difficult when they go too far. Like writing about an English Papist defending his support for paedophiles on the basis of superstition. Still, at least that one has gone off to help the English do something, probably involving pointing a laser at their remaining eye. Ooops, sorry, that was so-called "non-fiction" by alleged "journalists" writing about whatever shithole country I live in.

Reading "Alligators in the Artic" is definitely not bringing out the smiley happy people in me.

I have one notorious block of about 200 lines that has ~800 lines of comment above it starting with "these are business rules decided by management". It's utterly fucking bonkers, by careful design, and has an unnecessarily large number of inputs because "but what if the customer makes a wifi hotspot from their phone and uses a VPN...." times about 300 emails and many hours of meetings.

The comment is mostly references to and tiny excerpts from emails.

The final result of all this code? "the device is Online/Offline" (pick one)

"off" means "standby" on the landline, but "standby" means "off" on the washing machine

I saw a discussion online recently about machines that beep when offended. Many people are unhappy that, for example, washing machines beep forever if their door is not closed properly when they are not in use.

My main reaction was... you leave those things turned on when they're not in use? Mine don't have mains power available in that situation so I have no idea whether they would like to beep, and no interest in finding out.

"Impossible to simplify"? Um, right. Redesign, and in the meantime, I'm betting on spaghetti code. I, in my career, have taken a number of programs from 1000-2500 lines to about 300-400.

The one I work on is about 110K lines of code, and it is only one (fairly small) part of the over-all project.

And while parts of it could be better, improvements are not easy or trivial (well, usually).

Some programs are just going to have overlaps, no matter how you factor it. Think of it like a wind-up clock that only shows the hour, one developer is adding a second hand, another the minute hand. Obviously they are going to be playing with the same "guts" in the clock, and needing to run off the same master gearing. No matter how you built the original clock the problem will exist (WARNING, do not stretch the analogy to far, it will break the clock :) ).

For any who are interested, There is No Antimemetics Division, by qntm (a guest poster a while ago) is on sale in Canada and the USA. Thought I'd pass on the info.

201 - I'd really like to know how to design block structured Ada 83 and somehow finish up with spaghetti code at the same time. "How not to do do something" can be really instructive. BTW 3 separate programmers have already tried this and reached the conclusion that sub-program divisions move blocks that you pass through once for an index value and a reference value for no actual gains.

204 - Exactly. This machine only flashes an LED when in standby but you've not allowed to actually turn it off...

Like the quote I gave that I think is from Lord Dunsany. We're talking about fiction, where things must make sense, at least within the context of the story.

I’d suggest that at most they need to feel like they make sense. Things like the Rule Of Cool fit under this category, as does narrativium. And being told to turn off your brain and use the force just go with it? That’s also a valid storytelling technique.

I’d also point out that Alice In Wonderland was based on 19th Century math jokes, but probably 99% of the audience didn’t get the jokes and still thought it was a good story. High Nonsense, in the Wayward Children sense, is a perfectly valid story setting.

but you've not allowed to actually turn it off...

The daemon/demon inside it gets grumpy/escapes/starves?

I mean, it's one thing if the magic uses electricity and lack of electricity lets the magic escape. In that case it should come with a battery and even a knife to allow emergency tentacles or engrams or whatever to be drawn round the outside in fresh blood.

Moz
"real world" my foot... I want a refund.
Yes, well ....
Our misgovernment are currently trying to suppress their own enquiry, presumably because it shows the current "leadership" { DON'T LAUGH } up very badly.
Elsewhere the war in Ukraine widens, Fucker Carlson in the US almost-quotes Adolf about "Rat-like Jewish leadership", it looks as though RU blew that dam up { Under the cui bono rules } & a Russian cyber group { "Clop" } appear to be behind a vast date breach - as usual in cases like this, are they "just" criminals, or is it war by other means?
At the same time certain utterly - quote- "leftwing" Trade Union Tankies are denouncing Ukraine, simply because they hate the USA, or are they just stupid tankies?
You tell me.

Surely you have heard this one? (Ada-like) language design is a race between producing a more and more idiot-proof language and the universe producing bigger and better idiots. So far, it seems that the universe is winning.

I prefer John Sladek's counterpoints. Broot Force (by Iclick As-I-Move) from 'The Steam-Driven Boy and other Strangers' comes to mind, along with Tik-Tok.

I swear I've read at least one story where there were demons who will do a task but will take your ability to feel certain sensations/emotions as payment, but I can't remember the title or any additional details.

In one of Paul Cornell's London Falling novels, one of the characters sells the ability for her to experience joy. No demons involved as such, though

These nano-imps presumably have perfect recall for the instructions given, so each one is just a simple TM manipulating data. They are also not limited to a tape: you could give one a 3d crystal and tell it to move in 3 dimensions to manipulate the atoms.

I like the idea of magical spells being written in a funge. If the grid is three dimensional documentation could get very challenging indeed.

Although if someone was inventing code from scratch, sending tiny imps crawling across a surface reading mystic runes would make a lot of sense. The pathways wouldn't even have to conform to right angles, and spaghetti code would be all too literal.

The pathways wouldn't even have to conform to right angles, and spaghetti code would be all too literal.

Congratulations, you've just described genomics!

(Seriously, it turns out that DNA doesn't encode any kind of "blueprint" for a human being; first it was a bunch of junk surrounding "introns" that mapped 1:1 to protein peptide sequences -- basically, static strings rather than instructions -- then it turned out that the "exons" control whether the protein sequences are transcribed or not, then it turns out they're all coiled and supercoiled weirdly and external short interfering RNA sequences that aren't even in the DNA handle flow of control, and the whole lot is mediated by transcription factors which are themselves among the peptides which may or may not be expressed by the genome, and it's herpy-derpy turtles all the way down to the atomic level, just about.)

And, if we believe some of the more, er, imaginative pundits, it goes down even below the atomic level - though most people think those pundits are quarkers :-)

I can claim to be one of the first people who said that the real problem would be interpretation not decoding the genome and that we would never be able to predict everything, but, even in my wildest dreams, I never imagined the complexities that have been discovered since then!

As with AI, it makes it difficult for people like us, because it's irrational to take a position absolutely for or against allowing genetic modification, and that's the debate that is going on :-( Yes, it has great potential, but it also has great dangers; it is simply not possible to predict what any particular change will do.

The other way around - exons encode peptides, introns do not.

Actually, it's more complicated, exons are the sequences that form mRNA after splicing of introns, and mRNA consists of the coding region (CDS) that actually encodes the protein, and the 5'UTR and 3`UTR regions (upstream and downstream of CDS) that don't encode proteins, but control the protein expression via other proteins binding to them.

Actually, it's even more complicated, because 5'UTRs can contain upstream open reading frames (uORFs) and produce different, usually small peptides under certain conditions.

And of course the majority of proteins then undergo post-translational modifications, expanding the space of all possible proteins into infinity.

And some proteins (inteins) can actually splice themselves.

As someone who does VR/AR/XR development every day I can 100% get behind this. Watching people to cut the world into pieces and apply a first directive of the speed of sound in clay from the inside looks normal, from the outside it is a weirdly precise interpretive dance.

language design is a race between producing a more and more idiot-proof language and the universe producing bigger and better idiots

You can make it foolproof; you can't make it damnfoolproof.

(Lesson from engineering design class.)

Yeah, I was going by ~30 year old memories rather than looking things up when I wrote that comment (hence the intron/exon blooper).

But I stand behind my core assertion which is that eukaryote genomics is utterly crazycakes, even before you start looking at stem cell differentiation and epigenetic modulation and other external influences on gene expression.

https://www.oxfordreference.com/display/10.1093/acref/9780191826719.001.0001/q-oro-ed4-00005122;jsessionid=703C77D42496DE6A2C262FF75C3888EE

"Congratulations, you've just described genomics!"

Well... you did :) Paul's comment @ 91 was basically setting out the details of one of the things I had stuffed, unlabelled, into the jar of "very generous assumptions about the operating environment" when making an estimate of the kind of energy levels a single nano-demon moving one atom at a time would have to be operating at in order to achieve a worthwhile result on human scales of size and duration. I was basically saying that from the starting point you describe, the development into a lump of multiple nano-demons coordinated in a manner which I suppose I could have likened to Befunge with more than 2 dimensions if I had heard of Befunge, and with overall behaviour more recognisable as and better described in (broadly) terms of biology than computer science, is sufficiently natural that it is likely to happen without regard for anything the initiating magician might have intended or not intended.

Getting back to rules, Lewis Carroll just made a pink bunny in a Victorian girl’s frock pop out of my watch to remind everyone that good fantasy literature doesn’t need magic rules. Go ask Alice.

"Wonderland" is actually a fixup novel, linking together various short sketches invented to keep the Liddell kids occupied, and with no plot at all beyond "and then another thing happens". As shown by the various stage/film adaptations, you can rearrange the scenes in pretty much any order you like. The only parts required to stay in the same order are sleeping and waking.

If anything, this proves my point perfectly. "Wonderland" has no cause and effect; and therefore Carroll was unable to tell a story. All he could do was come up with a lot of examples of "magic happens, and nothing anyone does has any effect". Each individual scene is striking, in the same sense as each individual image in "Un Chien Andalou", but they're equally disconnected.

Conversely, "Looking glass" has a plan behind it (the chess game). Whilst each scene has its own things going on which aren't explained, there is definitely cause and effect happening.

Graham @ 223
Now, where have I seen that before?
Oh, yessss ....
Pterry's first two "Discworlds": Colour of Magic & Light Fantastic, where he openly admitted stringing as many SFF gags together as he possibly could, & labelled them "novels" ...
Then it all escaped & took over, much to everyone's delight.

"Impossible to simplify"? Um, right. Redesign,

Maybe not impossible. But not practical in terms of time and cost. If you've ever worked with code that had to implement state laws and court cases the code quickly becomes a mess. I worked in the US Property & Casualty insurance agency industry years ago. Forms and rules on policy coverages varied (wildly at times) by state and every court case tended to create new rules that had to be layered on top of existing rules. Often by state. Mostly without a 1:1 mapping to existing rules.

Back in the 80s as a software vendor we were looking at switching to laser printing (in offices of under 10 people) various standard insurance forms. These forms were color coded to make them easier for clerks to find and sort. We learned that a judge in Georgia had recently ruled that because the claim form used didn't have the standard ACORD yellow border it wasn't a valid form.

"Sorry, can't let you use that soldering iron, you might burn yourself. No, you can't use a saw...."

Oh, you mean health & safety regulations?

I'm sure I have read some story about health and safety ninnyism being unavoidably enforced in all aspects of life by omnipresent three-laws robots going overboard on the second part of the First Law. I think the viewpoint character was a visitor of some kind, who hadn't grown up being used to it, so that he could repeatedly be upset when he tried to have a hot drink or cut himself a slice of bread or something and the robots wouldn't let him do it, and start going nuts about what a ridiculous pain in the arse it all was.

But I don't have any mental record to contradict my assumption that the author was Asimov himself. After all, that's what pretty well all of Asimov's three-laws stories are about: the three laws don't work. They're enough of a vast improvement on what would happen if you simply didn't have them that they are an inviolable minimum requirement, but they are far too rigid to cope with the fuzziness of real-life situations, so when rigidly implemented they keep on buggering things up.

The overall theme of the series is really basically the extreme difficulty of imbuing artificial intelligence with common sense. The originators of the three laws were just getting into a totally new field, so common sense was simply a normal unquestioned background assumption for them, and it didn't occur to them that it could be a matter that would need to be specifically addressed. And indeed nobody ever really does address it; instead as the robots themselves are developed to be more intelligent as time goes on, they begin to work out their own version of "common sense from a robot's point of view"; the complexity of the situations which throw them off becomes correspondingly greater, and their ability to come up with workarounds off their own bat comes to take precedence over external correction by humans as a method of getting back on track again - but it's still a robot track, rather than a human one. So the point at which "three laws plus robot common sense" runs out advances from collecting a bit of selenium to the moral condition of the whole of humanity, as the series goes on.

I'm sure I have read some story about health and safety ninnyism being unavoidably enforced in all aspects of life by omnipresent three-laws robots going overboard on the second part of the First Law. I think the viewpoint character was a visitor of some kind, who hadn't grown up being used to it, so that he could repeatedly be upset when he tried to have a hot drink or cut himself a slice of bread or something and the robots wouldn't let him do it, and start going nuts about what a ridiculous pain in the arse it all was.

Sounds like "With Folded Hands" by Jack Williamson, Golden Era stuff. I think there were several stories eventually collected into a single volume.

And it couldn't be rewritten to read a table, fed from a d/b, a (let me be politically correct here) tuple with the relevant data?

Please clarify: are you saying it's one function that has zero called functions that's 110k lines long? That there's zero duplication of code, and that it could not be broken up into something a couple hundred lines long, with a few hundred function calls?

At a con last year, I mentioned genengineering, and this guy went off on a minute or two long rant, no pauses, and walked out of the con suite.

So I didn't have a chance to say, "oh, looking at how many folks in the room were heavy, so you don't want any genengineering to prevent diabetes? Or arthritis, asks the guy (me) with both knees partially replaced, and issues with a hip. Or a vulnerability to MS (says the guy who lost one friend to it, and has two more with it)."

I could go on, but why bother? Yes, I want genengineering, done RIGHT, with adequate testing per currently existing rules, plus probably new ones.

Yes, I want genengineering, done RIGHT,

Two questions.

• Can it even be done right?

• Who defines "right"?

Not trying to beat you up but these are the basic questions about much of "modern" life today. IMHO.

I would mention some medical things that looked great but went bad (a year or 10 later) but I'm not sure how UK laws operate on these subjects these days.

Can it even be done right?

Unknown as of yet.

Who defines "right"?

Difficult question. (Majority vote, perhaps?) But it is very easy to come with the answer to "What is 'right'?", to which vast majority of human beings would agree. Physical condition of a mentally and physically fit 25 year old, with no recognized genetic abnormalities. In other word, no diabetes, no MS, no macular degeneration, no arthritis, no arterial plaque. Whether it can be done is unknown (see your first question), but few would claim this is an undesirable goal.

Where it gets iffy, is improving on the "mentally and physically fit 25 year old". Nobody but a Sherpa or a Peruvian Indian can function on top of Everest without supplemental oxygen. So I claim that Sherpas and Peruvian Indians are the only people with healthy lungs, and the remaining 99.95% of human population have congenitally sick lungs. Why shouldn't my grandchildren have lungs of a Sherpa? And while we are at it, the spleen of a Bajau?

https://www.npr.org/sections/goatsandsoda/2018/04/24/604059598/the-secret-to-deep-diving-may-lie-in-the-spleen

Physical condition of a mentally and physically fit 25 year old, with no recognized genetic abnormalities

And - 20 years ago, or now if you're incautious about who you ask - no people with ADHD, no autistic people... you begin to grasp the problem, I hope.

My point is that LeBron James is certainly physically fit. Do we all want to have his body?

What is a "good" height? 4', 5', 6', etc...

Do we go for physical attributes that are great for survival in the wild, in a modern world, space travel, etc....

As to mental. Well. As some of the recent debates here have shown, many of the variations in how people think and process are considered abnormal but many and normal by many others.

For myself. Many consider me smart. (A non trivial number do not.) But I do score high on the standard tests. I have issues with visual colors and patterns that most people do not. But is it this "abnormal" wiring of my brain give me the smarts I have in other areas.

To me this is a slippery slope.

And let's not go into Charismatic religious types. Is this most of the time due to brain wiring or environment? Or both.

Leaving aside whether it's possible in practice, is it even viable in theory? Am I misremembering or weren't the majority of the Robots short stories illustrations of the Three Laws failing due to various corner cases? "Bicenntennial Man" is the counterexample that jumps to mind.

There are so many comments on here to which "you should read the Commonweal series!" is a reasonable response to. :-)

One thing I always found puzzling about the opponents of human genetic engineering: they seem to take it as a given that it would lead to a uniform "human monoculture", with all the attendant dangers of genetic uniformity.

Even if you assume that genetic diversity is important[1], it should be blatantly obvious that genetic engineering can create far greater diversity than would ever happen "naturally". Why do they not see it?

[1] I would argue that genetic diversity is far less important for a technological species than it is for a non-technological one, but that's a separate topic.

My point is that LeBron James is certainly physically fit. Do we all want to have his body?

"All of us" almost certainly do not want it. But if someone does want it, why not?

I have issues with visual colors and patterns that most people do not. But is it this "abnormal" wiring of my brain give me the smarts I have in other areas.

I met some people who said they hoped for autistic children, because they wanted their child to be next Einstein or Bill Gates. Pretty naive if you ask me, but it shows that not everyone wants to be "normal".

Charlie dissed "Asimov's three laws":

First, it's worth noting that they're actually John Campbell's laws, at least according to Asimov's story of their origin. And Campbell was the only game in town for a certain type of fiction, and was infamous for insisting on rewrites that met his somewhat narrow personal criteria, so I suspect Asimov accepted the laws on the rationale that at least this way he'd get published. Which leads us to:

Second, Asimov spent a lot of time writing about exceptions to the laws. He rarely (possibly never) treated them as perfect laws. Rather, he did what good SF writers do: take an external constraint (here, Campbell's laws) and explore it. As Wikipedia notes (confirming my memory): "Many of Asimov's robot-focused stories involve robots behaving in unusual and counter-intuitive ways as an unintended consequence of how the robot applies the Three Laws to the situation in which it finds itself."

Third, you're writing something like 70 years post-Asimov. You have a lot of hindsight he lacked.

Given the way software and AI is being designed now, and will probably be designed in the future, it seems highly probable that developers will follow the Asimov-Campbell suggestion of what our future holds: developers will implement simplistic, ill-considered measures that won't actually work in practice. Sometimes the golden age authors actually got it right.

Pretty naive if you ask me, but it shows that not everyone wants to be "normal".

Way too many of us (and many I know personally) have been considered NOT normal by some group or the other.

Normal is a bad target.

Normal is a bad target.

I am autistic. You don't have to tell me that. My point is, a lot of people, even some who are arguably "normal", do not consider it a target.

I have seen ones of thousands of lines that could not be split up in most languages without severely complicating the code. They were advanced scientific calculations, and had many dozens of local variables. They could be split in languages with multi-level hierarchical scoping (think: Algol 68) or in ones with suitable module features (e.g. modern Fortran) by using a module used only in that function.

However, to what purpose? It wouldn't actually have clarified anything, because the basic structure of the functions was a simple sequence of complicated operations each using almost all of the variables.

I'm getting annoyed. Where did I say gemengineering to make people "normal"? Please reference the post.

I said I wanted genengineering to cure/prevent disease.

Would I object, if someone wanted a tail? No, but that should be their choice.

And by the way, in my novels, I've got the mesh - genengineered cells that reproduce (no, anyone getting them is not "base human" anymore... that, to start, give early warning of disease developing before there are symptoms. And as an assistant: my partner has a language processing disorder. Some sounds she literally can't hear. She would LOVE to have the mesh, to help her speak more clearly, more "normal". And no one, much less anyone here, has the right to argue against that desire.

I can answer that one, though most of the people who say that probably don't understand the issues.

A good many genetic disorders are due to recessive genes where the heterozygote is beneficial, but the homozygote is severely harmful (e.g. sickle cell anaemia).

A good many others are beneficial under some circumstances, but not under others (e.g. skin colour). The most obvious form is where the disadvantageous form is 'modern living', so they are always harmful in (say) the UK of today but are pro-survival in others (e.g. (e.g. many tendencies to overweight and famine).

And a large number of others are where there may be such factors, but where we haven't identified them.

So it's a real issue.

I can answer that one, though most of the people who say that probably don't understand the issues.

I had to check which of my posts you were replying to, and it is #237. Sorry, but I find your answer a non-sequitor. How do you get from "A good many genetic disorders are due to recessive genes where the heterozygote is beneficial, but the homozygote is severely harmful" (yes, I know that) to "genetic engineering will make everyone the same"? If a single copy of Hb S gene is beneficial (which I agree it is), genetic engineering can ensure that every child has exactly one copy, and nobody should roll dice on getting two copies and sickle-cell anemia.

That's not so much genetic engineering, as zygote selection. Yes, it could be done by genetic engineering on the zygote, but that's using a sledgehammer to crack a nut. Let's ignore the fact that zygote selection is heresy to the 'fundamentalists' - there are no genetic problems with it.

What most people mean is to engineer their genome, and the genome that they will pass on to their offspring. And THAT's where the problems arise.

I want genengineering, done RIGHT

So therapeutic (to prevent diagnosed illnesses) rather than eugenicist (to prevent traits considered undesirable by eugenicists)?

Well yes, except prior to DSMIV, being gay was classified as a disease by US medical opinion, so germ-line gene manipulation to prevent homosexuality[*] would have been seen as therapy, not eugenics.

Do you see the problem here?

[*] Which is a really terrible idea. (For the record: all humans share the majority of the human genome: who you are attracted to is a function of how your nervous system develops, not your gonads. I suspect that if you could flick a genetic switch to stop men being attracted to men, a side-effect would be that you'd end up with females who are not attracted to men either. What you get from an off-switch for homosexuality is not heterosexuality, it's asexuality. Homosexual behaviour may be a conserved trait simply because it emerges from sexual receptiveness in members of the other sex.)

First, it's worth noting that they're actually John Campbell's laws, at least according to Asimov's story of their origin.

That suddenly makes a lot more sense. (And also explains why Asimov spent so much time trying to break them!)

Campbell was a horrible piece of shit, ideologically speaking.

(PS: can I suggest a re-read of Saturn's Children for my deeply cynical angle on Asimov's Laws?)

What that function really needs isn't turning into a module, it's a really good syntax-aware folding text editor.

Therapeutic? Yep. However, per your note, it would appear to me, at least, that the DSMIV is psychological, and so until it's demonstrable and repeatable, I think that falls under engenics.

Actually, I should have put it that the physical issues are clear. I suspect, however, that issues like guys attracted to guys are far more complicated. And genengineering is going to come in slow - there will be a huge number of court cases over many of the steps.

Meanwhile, I dunno 'bout eugenics, but my genics is just fine.... (Resist? Moi?)

Please clarify: are you saying it's one function that has zero called functions that's 110k lines long?

You said:

taken a number of programs from 1000-2500 lines

So, I was talking about a program, not a function (it has LOTS of functions).

That said, yes, there are some long functions in it, some things are just too complex to make short. Like EC said, there is no real reason to make them shorter, it wouldn't help anything. You end up having a bunch of non-sensical one use functions and/or functions that you are passing a zillion parameters to. It makes things harder to understand rather than easier.

Charlie @ 179: you can make a trivial end run around the three laws of robotics by redefining "human".

Freefall has a lot of fun with this one. See for instance the sequence starting here: http://freefall.purrsia.com/ff1600/fc01581.htm

It was pretty comprehensible with just appropriate block comments and good layout. Insofar as any code of that complexity (or, indeed, the mathematics it described) is comprehensible.

I once needed to work on a program that had followed the dogma that no function should be more than 5 lines long (yes, really - that was a thing, once), though it had quite a few longer ones. That was in Algol 68, so had hierarchical scoping. It wasn't large (10,000 lines or so), but it was gruesomely incomprehensible. There were argument-free functions whose sole purpose was to add one to a particular variable, and so on.

Forms and rules on policy coverages varied (wildly at times) by state and every court case tended to create new rules that had to be layered on top of existing rules. Often by state. Mostly without a 1:1 mapping to existing rules.

It's a rules-based AI where the rules are written by legislators whop have no training in formal logic.

This seems to apply generally to legislation, I've worked on medical stuff where the "implements legislation" side is just a complete mess. The saving roll is that the people verifying that we comply with the legislation have no better idea than we do.

»Nobody but a Sherpa or a Peruvian Indian can function on top of Everest without supplemental oxygen. So I claim that Sherpas and Peruvian Indians are the only people with healthy lungs, and the remaining 99.95% of human population have congenitally sick lungs.«

Have you ever asked an altitude adapted Peruvian Indian or Sherpa how they like their lungs at sea-level ?

The lungs absorb O2 and dispose of CO2, in a roughly constant ratio, and getting that O2/CO2 ratio wrong is almost as bad as getting the absolute amounts wrong.

Within a fairly large range of normal atmospheric conditions, the partial pressures work out so that both the absolute and relative numbers can be modulated by the amount of air transport, but there are a lot of corner-cases, which your body will make sure you perceive them as very uncomfortable.

The way the regulation is implemented is by monitoring the acidity of the bloodstream: Too much dissolved CO2 turns lowers the pH value. That makes a lot of sense because measuring dissolved oxygen would be a lot harder and would have to happen distant from the breathing muscles etc.

The adaptation you unwisely desire, amount to making it relatively /harder/ to dispose of CO2, because acquiring the necessary amount of O2 is thermodynamically more difficult at altitude.

Disposing of CO2 on the other hand is almost the same effort at all altitudes, until you start diving.

So bringing altitude adapted lungs to sea-level, the CO2-based breathing reflex will keep going as it usually does, but you will absorb a hell of a lot more O2, which, counter-intutively, is really bad for your health, because of oxidative stresses.

Likewise, bringing sea-level lungs to altitude will starve you of O2, because your CO2 disposal runs the show.

It's not that your sea-level lungs could not absorb more O2, but they dont get asked to. Dumping too much CO2 also has bad health-effects, but mostly instantaneous rather than long-term, this also counts against your survivability at altitude.

If you really want to improve that system, make it regulate on both O2 and CO2, for instance by tilting the curve of O2/CO2 ration vs. depth of breath much further than it is today.

»It's a rules-based AI where the rules are written by legislators whop have no training in formal logic.«

I talked to a couple of the brains behind Gov.UK once, (heavy Varnish Cache users :-) They told me they found that UK laws are a true mess in really strange ways.

For instance one law, prescribing veterans benefits, was only specified for the bits of UK's geography where veterans lived when the law was enacted. There were many places where, if veterans moved there, the law did not authorize any benefits for them.

He said they found so many of that kind of problems, that it nearly cancelled the project, because other departments insisted on charging gov.uk for "fixing the problems they caused".

In Denmark they tried to build a "unified debt-collection system" for the tax-authority, and it imploded because there were many thousands of different ways people could owe money to the public purse, from parking tickets over day-care lunches to criminal sanctions, and they are all legislated differently.

Three points about human lung function:

1) You've got to include Ethiopians and other African highlanders in the list. And possibly some Papuans???

B) Every group of highland humans does it differently. The Tibetan/Himalayans, for instance, appear to have inherited their adaptations from Denisovans, while Andean natives independently evolved tolerance ca. 10-15,000 years ago, or more recently.

III) With respect to highlanders living in lowlands, it's a truism that they suffer disproportionately from lung infections (like tuberculosis).

The take home is that humans are fairly confined by altitude. There are populations adapted to sea level, high elevation, diving (Bajau), and each adaptation is suboptimal for other elevations. But we're all confined within a band a few kilometers above, and a few hundred meters below, mean sea level. Earth is vertically zoned with a vengeance, from the exosphere down to the deepest trenches, and human really can only live freely in one layer.

It's a rules-based AI where the rules are written by legislators whop have no training in formal logic.

You seem to think this is a bug. It's a feature. If the bigger government can't fold the local system into itself, the locals perforce maintain control of it. I'd gently suggest this is yet another battlefield in the ubiquitous struggle between local and national control.

Some random thoughts re: magical fantasy versus scientific realism discussed earlier in this thread. A recent Nova program on PBS (or maybe it was Nature) explored how the human brain assembles a composite ersatz "reality" from various sensory inputs. Most information gathered is left unused and the remainder blended with automatically synthesized filler material to compensate for info discarded.

The example shown was how eyesight focuses on only a small fraction of the visual field and leaves the brain to fill in the rest. This seemed to me somewhat like dvd file compression leaving the values unchanged which represent background pixels staying the same from one frame to the next, varying only those values for image components that actually do change. Another similar example is how the blind spot from an optic nerve connecting with the retina gets seamlessly filled in by brain algorithms, almost like a cloning tool in a photo-editing program would do.

Similar types of brain processing create an almost first person shooter videogame-like image for the individual organism to use as ongoing reference in understanding its relationship with the surroundings. Accuracy and fidelity to source material would vary from time to time and from brain to brain, but at no point would synthesized parts be missing. So the difference between fantasy and reality in terms of mental states would be differences only of degree, not of kind.

Chomsky's idea of language acquisition as a genetically determined predilection has been largely accepted, so why couldn't genetic tendencies also influence one's general world view? Magical fantasy could be the result of projecting inborn mental shortcuts onto the external environment, for better or worse as far as survival results, probably better in the past but less so in the modern world due to, you know, civilization.

Anyway I got thinking about it this morning over coffee when my wife criticized a quarrelsome neighbor and I replied yeah the guy lives in his own fantasy world, and then had to add, don't we all.

Well, the air quality reached 7 today, expected to hit 9 tomorrow. (10 is the maximum the scale goes to.) Have been able to smell smoke outside all week. Hopefully we'll get cleaner air Friday or Saturday, but that's uncertain right now.

Worst fire season on record. Unless you're a Republican politician, apparently, in which case it's simultaneously fake news and the fault of the woke diversity-obsessed lamestream media somehow preventing the forestry companies from raking the forests. How the f%^! do these people keep getting elected?

You say the scale only goes to 10... Australia has to "make it go to 11" with fire danger, because we don't have meaningful air quality reporting. There were quite a few older signs with "catastrophic" manually added after "extreme" but I think they're mostly been replaced now.

https://afdrs.com.au/

My particulate meters only go to ~1000 micrograms per cubic metre while the official ones go to 2000 (transmissive IR vs paper filter based). Given the official guidance to "stay inside, avoid exercise, wear a mask if outdoors" past 500 I'm not sure there's much benefit measuring past 1000... mind you, with the shitty houses we have "stay indoors" is weird advice since so many will have worse air quality inside.

If the bigger government can't fold the local system into itself, the locals perforce maintain control of it.

My experience is that more often the system remains uncontrolled. Which is great when you actively want the tragedy of the commons (complete withdrawal of insurance companies from areas affected by climate change risks*), but all too often it's "X is not regulated, national government doesn't care and local government doesn't have the power".

Australia is internally notorious for unfunded mandates from national government. The latter raises 80% of total taxes, but is only responsible for 50% of government spending. They make rules for stuff like national education standards, and they subsidise private schools, but otherwise compulsory education is a state responsibility. And so on, resulting in large, essential transfers from national to state governments. The pattern repeats between state and local government.

Having read that Australia is better than most federal systems I fear this pattern repeats worldwide. "local control, federal funding"... as with the recent hostage negotiation it leads to "we're cutting funding by 10%, but we're not responsible for any particular cut you choose to make".

(* you may not want people to be unable to get insurance, but it appears to be a necessary step along the way to realising that the climate catastrophe may have costs in this quarter's earning report)

Um. I think you've actually reversed the causality arrow without meaning to.

What you're describing is how real-world magic actually works. Magicians have learned empirically how to hack the way our brains process reality. Teller (of Penn and Teller fame) coauthored a paper in Nature on the subject in 2008 or so, and in 2011, Nova had an episode called "Magic and the Brain" (e.g. https://www.pbs.org/wgbh/nova/video/magic-and-the-brain/ ). There's a lot of this online already.

What Nova showed last week is neuroscience after researchers have spent a decade hobnobbing with actual magicians. If it looks like the basis for magic, it's entirely possible that this is because they studied with magicians.

One key point is that magic systems in fantasy have nothing to do with this. Designing new magic systems is part of the (sub)genre, and there's a lot of ad hoc and post hoc rulemaking around this aspect of fantasy.

These rules and their importance are a pub-level conversation topic on par with building closed ecosystems in space, and they're a strange attractor around here. My person meta-rule for determining the rules of fantasy magic is simply Teller's "Nothing fools you better than the lie you tell yourself." ( https://www.smithsonianmag.com/arts-culture/teller-reveals-his-secrets-100744801/ ). To put it in this context, the most important rule you can derive for why and how magic should work in stories is what feels right to you. If it feels wrong, why argue for it?

Now, going back to Charlie's original question ("Now, taking a D&D-ish generic high fantasy setting as a given, what can this lead to?"), I'll suggest the following:

European medieval math and geometry were largely devoid of elements such as recursion, scaling, and even infinity, ideas which are fundamental to the notion of programming-as-magic.

So if you want medieval-level technology with recursion, scaling, and infinity, you've got to go outside Europe. Specifically, you've got to go about 3000 km south to West Africa, where these ideas crop up repeatedly in art, urban design, and spirituality. Check out Eglash's African Fractals: Modern Computing and Indigenous Design.

[[ Fixed broken link - mod ]]

My experience is that more often the system remains uncontrolled.

That's often my experience too. I agree with you that this kind of shit often results in tragedies. However, as with so many things (cf decarbonization), simply pointing out that there's a way to rationally fix the problem normally doesn't get it solved, because the people proposing the rational solution don't fully understand the human dimensions of the problem.

At least we're getting better on this blog. Decades ago when I started here, the common line (usually delivered with great arrogance) was that physics was the only important limit. If something was physically possible, it would get done, and politics was irrelevant. We've had this innocence beaten out of us by the last few decades of politics, and that was something of a good thing. Maybe.

The rest is stamp collecting? 😋

Yup.

And now cosmology is trying to sell us on the notion that the multiverse is just an immense stamp collection of universes, all basically randomly-designed stamps made out of superstrings. But they need a bigger accelerator to probe this idea, of course.

I wonder if part of it is that software is much more obviously an ecosystem now? The typical geek has had their nose rubbed in the "it's unknowably complex interactions all the way down" for a couple of decades at this point. Even the really smart ones have had to admit that they aren't capable of writing their own smartphone software stack from scratch even if they wanted to. Let alone something like a motor vehicle.

But also, I thought ending papers with "more research is needed" was like Christians ending prayers with "amen"? It's part of the ritual...

What you get from an off-switch for homosexuality is not heterosexuality, it's asexuality.

So, there's this world where people have done the eradication of homosexuality, wanting to go to that (fantasy) 1950s (American) suburbia 'utopia'. It works! No annoying homosexual deviants!

Only now there really is a good reason for separate bedrooms - nobody wants to to have sex.

Of course the problems become apparent when there are no children, either...

(Yeah, yeah, asexual does not mean aromantic nor even not wanting to have any physical contact, but could be the basis of a story...)

Even the really smart ones have had to admit that they aren't capable of writing their own smartphone software stack from scratch even if they wanted to. Let alone something like a motor vehicle.

Yeah, I could see myself capable of doing something like late 1980s computer software stack from scratch. Maybe early Amigas, maybe, at the highest level of complexity, or a bloated terminal emulator on a 386 machine. Modern systems? Too many turtles on the way down.

I've worked with many kinds of systems, though, from kernel level to UI, but there are large gaps for modern systems which I really have little idea of.

At one point I felt comfortable that I perhaps could design an 8-bit processor to go with it, but that was probably the hubris of youth. (Though a simple one shouldn't be that hard.)

»Yeah, I could see myself capable of doing something like late 1980s computer software stack from scratch«

Only if you get to hand-wave away all the necessary DSP stuff required to actually talk to base-stations over the air...

In the first generations of GSM phones, 90%-ish of the LOC were DSP related, and while that percentage has dropped, the difficulty of the subject domain certainly has not, to the point where it is now baked into silicon to prevent programmers from mucking about with it.

In many modern smartphones, the entire RF-part is just a USB-modem - in some cases still using "AT" commands...

But if that is what we're talking about, sticking a UNIXoid kernel on the so-called "main-CPU", and write some code to talk to a USB modem doing all the hard work ?

Of course a single person can do it and several have done so over the years.

Will it be "a Smartphone" ?

Probably not in most peoples eyes, but by the time you have dumped Wayland, KDE/Gnome/Firefox on it, it will have at least all the same fundamental flaws, but probably not as much spying/surveillance.

Only if you get to hand-wave away all the necessary DSP stuff required to actually talk to base-stations over the air...

Yeah, I was thinking only of desktop computers, not mobile phones. DSP stuff is probably doable, but, uh, wouldn't want to (I have a degree in electrical engineering and signal processing, so I have some knowledge of the field).

Radio frequency stuff... uh, yeah, I passed the (introductory level) courses, yes.

Yeah, I could see myself capable of doing something like late 1980s computer software stack from scratch. Maybe early Amigas, maybe, at the highest level of complexity, or a bloated terminal emulator on a 386 machine. Modern systems? Too many turtles on the way down. ... At one point I felt comfortable that I perhaps could design an 8-bit processor to go with it, but that was probably the hubris of youth. (Though a simple one shouldn't be that hard.)

A group of us teens discovered computer logic around 1971 or so, read everything we could find and even designed a very simple 7400 logic based computers. I/O was basically a few leds and switches. We didn't build it, no way we had the money for it. Especially any amount of memory that would be useful. But it would have likely worked at a slow clock speed.

But the 7400 logic based "mini-computers" of the late 70s / early 80s were just a more complicated iteration of the kind of thing we did.

I and the other primary quickly decided to step up a level and learn to program hardware other built. Much more satisfying. In the later 70s we had a complete OS source for a mini-computer brand and I likely had read all of it. Ditto the microcode for a different vendor in the early 80s.

I also had a very deep understanding of an insurance agent office automation package in the 80s. You could ask about one of the modules and I could tell you the logic and the main variable flows without looking at the code. But compared to today it was very simple. But at the time we were the biggest vendor in the market.

But that was then. And my timing light and dwell meter haven't been used in over 30 years either. Ditto my drum brake tool.

I read over 5 years ago that the OS inside a Samsung EVO (3?) SSD was over 300 meg.

I was literally at the Wright Brothers memorial a weekend ago. It is unreal to consider what they first flew compared to where we are 120 years later. Or evern 70 years after they first flew. To the point, two guys designed (very nearly in isolation) a powered airplane. Built it and flew it.

Will it be "a Smartphone" ?

My "smartphone" will take high resolution photos, 4K video, and panoramic scenes stitched together in real time. The effort to just do this ignoring the phone aspects are unreal for a single or even a few people.

Well, the air quality reached 7 today, expected to hit 9 tomorrow. (10 is the maximum the scale goes to.)

I think you're near Toronto. This was Wednesday and New York City supposedly had the worst air quality of any major city on the planet. Maybe a bit of hyperbole but you can easily find images of the city during mid afternoon that were downright scary.

And the stay indoors if possible warnings extended all the way south to a few miles north of me. 400 miles give or take from the Canadian border.

What did I ever do to you? [/sarcasm off but only slightly]

Er, no. Ethiopia is lower than much of Colorado. There is a HELL of a difference between 2,000-3,000 metres and 4,000-5,000. Yes, the inhabitants have some adaptation, but not to anything like the same extent as the Tibetans or even the Andeans.

What you may be missing is that oxygen absorption is not proportional to oxygen concentration, but is offset by an amount equivalent to about 8,000 metres (in most people). I really noticed this when we went up El Teide, and it made me cross Machu Pichu off my bucket list :-(

Unfortunately, too many people have adopted the political viewpoint - that the only 'solutions' that are possible are the politically acceptable ones. Well, if that is so, we are SHAFTED. Actually, we probably are :-( Most problems (including social) need addressing as engineering ones - i.e. what is the simplest solution that will actually work, and how do we get there.

This is why I am an anti-EV person - not that they aren't a 'good idea' but that simply replacing current gas-guzzlers with even larger EVs is going to make things worse, not better.

.. maybe ..

Flooding downstream of the dam and disrupting water supply to agricultural land is bad for both Ukraine and Russian occupied areas, Including Crimea. Arguably, Russian forces had easier access to plant explosives, and immediate tactical advantage by flooding the river.

It's also possible the dam gave way after months of high water levels and overspill.

Zelenskyy said as much:

"The fact that Russia deliberately destroyed the Kakhovka reservoir, which is critically important, in particular, for providing water to Crimea, indicates that the Russian occupiers have already realised that they will have to flee Crimea as well."

Well, maybe, but all the evidence is that Russia will defend Crimea pretty well at all costs.

Given that both sides have out-of-control 'irregulars' operating freely with access to large amounts of weaponry etc., it could very well be one of them. We don't know, and probably never shall.

https://www.aljazeera.com/news/2023/6/7/ukraines-zelenskyy-says-dam-blast-will-not-stop-military-plans

I was literally at the Wright Brothers memorial a weekend ago. It is unreal to consider what they first flew compared to where we are 120 years later.

And for a practical comparison:

The Wright Flyer's first flight was shorter than the length of the cargo bay of the An-225 Mriya (before Russian paratroops wrecked it in the first week of the Ukraine war).

This is why I am an anti-EV person - not that they aren't a 'good idea' but that simply replacing current gas-guzzlers with even larger EVs is going to make things worse, not better.

The problem is a more general one with automobile marketing trends over the past 20 years -- partly due to weirdness in the US tax code, "light trucks" can be written off against tax in 12 months if used as personal transport, building in a tax incentive for driving a tricked-out pick-up truck. Which in turn drives non-truck-owners to want bigger cars -- in this case, SUVs -- purely for survival purposes on a road dominated by thuggish behemoths.

A 2CV should be a decent size/form-factor for a family EV: holds four adults plus luggage, can drive at up to 100km/h, has a sun roof, even has removable seats you can pull out if you want to have a picnic or use the rear for cargo. Make it out of modern materials and with better aerodynamics and to a higher safety standard and it ought to be achievable at the original's kerb weight of 600kg, which for transporting four adults is a lot more reasonable than the Tesla Model 3's 1700kg, of which 500kg is battery. Allow maybe 125kg for the new 2CV's battery and it's still less than half the weight of the Model 3.

Nor does a car really need to go faster than 100km/h -- that's close to the universal maximum road speed limits in force everywhere because human reflexes aren't up to consistently going much faster and the kinetic energy involved in any collision rapidly becomes unsurvivable above that speed no matter how good your crumple zones and air bags are.

Alas, on today's roads a modernized electric 2CV would feel like a death trap. And it's entirely due to marketing bullshit to ramp sales on the basis of an American income tax loophole.

As you know, that is my position, only I go rather further!

It's less than clear why the USA lunacy should affect Europe so much, or even whether the idiocies here are actually due to a knock-on effect from the USA. Looking at the prices(*), gimmicky and regulations, I am pretty certain that there is a tacit conspiracy between the manufacturers and some influential governments.

(*) Such as exactly WHY should prices have gone up so much faster than inflation in the past decade? No other consumer products have.

Having read that Australia is better than most federal systems I fear this pattern repeats worldwide. "local control, federal funding"... as with the recent hostage negotiation it leads to "we're cutting funding by 10%, but we're not responsible for any particular cut you choose to make".

In Canada we've got the same funding with the reverse problem: provincial governments taking federal money and spending it on other things, then blaming the federal government for funding shortfalls. For example, Ontario getting money to improve ventilation in hospitals and schools, then not doing that and giving a tax cut to motorists. Or the federal government giving taxation power to the provinces to fund health care, and the provinces promptly cutting those taxes and demanding more money for health care/*.

I'm currently serving on a federal committee on health care, and one of the big issues we deal with time and again is how to create and fund a program in such a way that the money will actually be spent on the people it is intended for, rather than 'repurposed' by the province.

Oddly, this is only a problem with right-wing provincial governments.

*Health care in Canada is under provincial control.

I forgot to say this earlier. There is actually only a very small proportion of our roads that even feel like a death trap to small vehicles - and you don't get much smaller than a recumbent trike! Excluding motorways, they are essentially only the trunk roads, similar main A-roads and a comparable number of lesser, high-speed A- and B-roads (mainly rural, but a few suburban and a very few urban ones).

The problem is that there are a fair number of routes that are ONLY viable by using one of those roads. That wouldn't be infeasible to solve (in a variety of ways), but needs the political will and a relatively small amount of compulsory purchase and road-building. It has been solved for motorways, after all, except in a few places like the Dartford Crossing. Many cities are trying that, but generally without considering the whole problem.

My guess is that it would be considerably easier, cheaper and quicker than upgrading our generation and distribution infrastructure to support all-electric power. They're not alternatives, but the sane approach would be to prioritise the road restructuring.

You might almost be describing the Citroen Oli concept, which might be less of a deathtrap than a 2CV or VW type 1. And US pick-up trucks have gotten immense in the last half century, as if light truck R & D was building towards "Your very own monster truck!".

partly due to weirdness in the US tax code, "light trucks" can be written off against tax in 12 months if used as personal transport,

While the use of pickup trucks for personal transportation is a bit crazy in my (and other people in the US's) opinions. I don't think you understand the US tax code here at all.

I own a pickup truck. It gets used when needed. It has sat unused at times for over a month. But I'd really like to see notes on the US tax code to back up your comments.

Such as exactly WHY should prices have gone up so much faster than inflation in the past decade? No other consumer products have.

It seems your talking autos.

BECAUSE the people who are buying new cars want nicer cars than in the past. And the people buying new cars can afford them. So the auto manufacturers are building what people want.

The Wright Flyer's first flight was shorter than the length of the cargo bay of the An-225 Mriya (before Russian paratroops wrecked it in the first week of the Ukraine war).

Standing where the flights took off it was impressive. There are 5 big stone markers. One at the take off point. 4 at the landing point for each flight. The first 3 were from around 140 to 200 feet in distance. Standing there you can imagine that detractors would claim it was the wind that carried the plane.[1] But the last marker at about 800 feet out, you know they had to know themselves they had it. They could fly.

[1] They planned to fly some more flights but while taking a break after the 4th one the wind flipped the plane and wreaked it more than they could repair it. While I was there it was to me a crazy windy day. I'm thinking 20mph or so steady with gusts of more. It felt like I could got 100' or so through the air with an oversized wind breaker and arms spread wide. The Outer Banks ARE windy.

"Standing there you can imagine that detractors would claim it was the wind that carried the plane."

And would miss the point. All the wind's doing is making it easier (or harder) to achieve takeoff speed relative to the ground. Once they're off the ground the ground speed and ground distance become irrelevant; the achievement is all about how long a time they can keep it up.

The people writing newspaper articles and reading them in 1904 would not include or care about such. As it was many of the news articles at the time exaggerated what they did to the point they had a PR problem with much of the public.

But I'd really like to see notes on the US tax code to back up your comments.

I don't have access to a copy of the US tax code (and life's too short, anyway). AIUI the issue is that if you own a small business you can get a "light truck" and write it off as capital expenditure in one year. Meanwhile it has a tricked out leather and walnut interior with a surround sound system and you use it for commuting between home and your dental practice rather than for hauling lumber ...

There are a lot of small businesses out there.

"I'd really like to see notes on the US tax code to back up your comments" - ISTR that you're a USian, so perhaps this query is better directed to a USian tax accountant than to a UK writer?

BECAUSE the people who are buying new cars want nicer cars than in the past. And the people buying new cars can afford them. So the auto manufacturers are building what people want.

That hasn't been true for decades. Vehicles take years to design, because all the supply chains also have to be built or upgraded, factories have to be tooled, people or robots skilled up, labor contracts done, politicians greased, etc.

If you listen to the news, especially NPR, occasionally the real stories slip out. Charlie's point about SUVs and trucks being favored by the tax code was a big one. A smaller one was about how, back in the 1990s, the car companies were already planning for increasingly bloated SUVs, going out to 2005 or so, with the names of the then-unreleased vehicles in the episode.

They weren't responding to demand for urban assault vehicles, they were creating a market and juicing the demand for a profitable product they were already planning to make. Charlie's unthinking repetition of the "you've got to have a big SUV to defend yourself" is a marketing line, not reality, as the Ford "Exploder" proved.

The assault rifle makers in the US are currently using the same tactics to sell their guns, despite our dropping crime rates.

Now, the car companies are reportedly building mostly trucks, SUVs and high end sedans. Why? Computer chip supply shortages dating back to the pandemic. They made the mistake of computerizing functions (seat warmers?), and they screwed up their chip orders in 2020. So now that they're short on chips, they're apparently prioritizing them for the most profitable vehicles. These aren't what most people want to buy, which are affordable commuter cars, it's most of what they can get at the moment.

But the bottom line is that, when you have a complex product, like an AI, SUV, AR-15, drug, or whatever your preferred shiny is, if you're selling it commercially, you can't spend years getting the supply lines ready, then hope there's a demand for it. You've got to create that demand through marketing. The market doesn't follow consumer's demands, it works hard to coerce us into buying what they're selling.

OK. You're talking Section 179 capital asset write offs. It is not just trucks. It is almost any kind of asset used for business more than 50% of the time. Computers count. Almost any auto. Instead of depreciating the asset (up to about $1mil) and deducting the depreciation amount from your income as an expense you can take it all in the first year as an expense.

This just gets back to using a a truck instead of a sedan/coupe (I forget what it is called in the UK/EU) for business is a bit crazy in my mind but it is NOT aimed at trucks.

And if you don't keep (or invent) decent records is one of the top things that will bite you in an audit. Which is why the Apple and Google App stores are full of mileage tracking apps.

Oh, you can't just do this indefinitely as a way to not pay taxes. There are rules about showing a profit (or at least getting close). Tax accountants get rich assisting big businesses putting side companies into insolvency as a tax dodge but it's much harder for a sole proprietor to do such.

I believe John Oliver did a segment on this issue in early April. My understanding was that it wasnt a tax incentive but an tariff/ import incentive on auto manufacturers, as well as the USA auto companies inability to develop cars to compete with import vehicles, so they focused on light trucks, the classification that includes SUV's and pick-up trucks.

That is the manufacturers'/petrolhead polemic, to be sure, but all the assertions are demonstrably false in the EU and UK.

As you saqy, but I think there's more to it than that. Even the price of the low-end models has gone up far more than inflation. I smell a rat - a rotten one.

In addition to the fact that GDP/capita has been almost flat for more than the period mentioned, the price and (lack of) availability of spare parts is one of the reasons that the second-hand car market has boomed in the past decade or two. The problem in the link predates Brexit, incidentally. People have been reported as swapping their new cars for older models, for this reason; I have met some, so suspect that it's fairly common. And, unlike days of yesteryear, even seriously old low-end cars (like my 2011 Skoda Fabia) sell for fairly large amounts (say, 40% of what I paid for it), and ones with 150,000 on the clock still sell well.

https://www.theguardian.com/money/2023/jun/08/mercedes-says-it-cant-find-a-vital-part-for-my-2018-car

I own a pickup truck. It gets used when needed. It has sat unused at times for over a month.

Never mind the tax code -- why not rent a pickup truck when you need one, instead of owning one and have it sit unused?

"Alas, on today's roads a modernized electric 2CV would feel like a death trap."

Disagree. I would be perfectly happy to drive an original unmodified 2CV on today's roads, apart from the sole point that the top speed is inadequate. A vehicle needs to have a top speed reasonably in excess of the legal limit, so that at the legal limit it still has plenty of spare capacity in hand to cope with adverse gradients, headwinds and so on, without struggling to keep up with the traffic. Although that is probably less of a problem for a 2CV now than it was when they were being advertised with the headline "At 71.5mph you'll hear a funny noise" and a picture of a flashing blue light reflected in the rear view mirror. The retrospective joke being that they would have had far more vehicles exceeding 85mph to be chasing then than they do now.

I despise the way people have been conditioned into thinking the ever-increasing load of safety bollocks on cars is actually necessary and they daren't risk driving a car that doesn't have it. The vast majority of people simply never experience a situation where it does anything useful at all, and even fewer one where it actually makes a critical difference. But they all pay the multiple penalties for its existence and the manner of its implementation, in the increased cost of the fuel needed to drag around the extra weight, much of the massive increase in insurance costs, the increased cost of the car itself, and the greater likelihood of the necessity of its premature replacement. And also the indirect penalties from the additional carbon dioxide emissions from burning the extra fuel, and the amount of unnecessary additional production and unnecessary activity in general.

The SafeTiPed comes irresistibly to mind, but at least that was a "proper" design, that did actually concentrate on its ostensible aim and so did sensible things like reducing size and mass. What we have now is worse, because it is not inspired by the intention of increasing safety, it just says it is. It's a commercial inspiration with the intention of increasing profit. The manufacturers can charge disproportionately more for the greater mass of materials in the car by using "safety" as an excuse. They can sell new cars more often by inventing extra features that reduce the chance of some particular type of injury from 0.000001 to 0.00000099 and convincing people it's suicidal not to have them. They can surround the front ends of cars with "energy absorbing" bits of tinfoil and plastic which interlock in such a way that even a trivial ding at walking pace will destroy several of them, and then charge outrageous prices for the replacement parts. They can also incorporate tinfoil elements in the structure, so that a ding at slightly above walking pace that a car fifty years ago would still have shrugged off will now irreparably ruin the suspension alignment, and they get to sell a whole new car.

And it's easy to get the legislators in on the racket, because so many of them think the unnecessary extra production and activity is actually desirable, and the magic word "safety" is great for propaganda and for making sure nobody argues against it because they look like a murderer if they do.

Compare the absence of commercial inspiration and paucity of legislative support for such notions as, for instance, that relying on the surrounding armoured cage for protection may not be the best way of going about things...

See #286.

There is also evidence that (at least in the UK) 'safety' features have probably killed more people than they have saved, even ignoring the pollution aspects, because of the way they discourage walking and cycling, with the consequent reduction of the population's health. Apparently, the average speed went up by 10% following the introduction of seat belts.

Naomi Klein rooted out the source of the SUV/Truck inanity in North America, in that the US had a minor trade dispute with whatever the EU was called in the 70s over import tariffs on chickens. In retaliation they placed a tariff on imports of light trucks.

Building on that, US automakers successfully had SUVs defined as 'light trucks', which provided them with a built-in market advantage on trucks and SUVs. They then started marketing heavily - watch any sporting event on television and you see some of the ~10 Billion/annum spent on marketing trucks and SUVs.

A trivial retaliatory measure in a trivial trade dispute about chickens leading to catastrophic climate change. Somehow it might be a fitting epitaph for our species.

Some of the knock-on effects have been that the US automakers have had no incentive to make 'better' trucks or SUVs because they have the bonus profit margin. Those few 'foreign' automakers who sell trucks into the US market have had to compete on quality, which is why the Toyota trucks (among others) are vastly superior in durability, safety, fuel efficiency etc.

ROTFLMAO!

Feel free to look up General Semantics, and the core of it, the Structural Differential, that Korzybski developed almost a century ago... and tell me the difference.

European Common Market? I remember a "Chicken war" in the sixties, much attention to a pick-up variant of the VW type 2, Volkswagens being a far more common import back then.

Disagree. I would be perfectly happy to drive an original unmodified 2CV on today's roads, apart from the sole point that the top speed is inadequate.

Since I drive an EV, I'll add two points of clarity, from a US perspective.

One is that primary safety isn't about top speed or mass, it's about moving at the speed of the surrounding cars. Going too slow is as problematic as going too fast.

This, incidentally, is why safety features only go so far. If a car wipes out going, say, 130 kph, the people inside are in danger no matter what size the car is.

Small EVs have one safety feature that got missed: superior acceleration to most ICE vehicles. Being able to simply scoot out of the way of an idiot in a bloatmobile is the biggest safety feature that modern EVs, and old subcompacts, have.

This is what got missed in the idea of a 2CV-equivalent EV. Such cars don't accelerate rapidly, and that's a safety handicap.

In '08, my late ex and I were in Colorado for Worldcon, and visited friends of mine before the con. They took us up I think it was up Mount Evans (a bit taller than Pike's Peak). We got hailed on in early August, looking out at a glacial cirque*. At the top, over 14,000', we didn't stay that long, because I actually started having issues breathing.

* It took me a while, looking at the cirque, until I finally realized, and told the others it reminded me of my youth, but there weren't any mammoths there....

Magic is Money and Money is Magic.

Missed commenting on this when you wrote it. Sorry!

This is a really good idea, whether on its own or paired with the idea of "mani" in 28 (mani is like both mana and money).

The one thing that would make it even twitchier is to add the real world problem of intergenerational wealth transfer, which is the major existential threat that wealthy families see. It's a major threat for talented families too, for much the same reason.

bing Charlie has it dead right.

In '04, I wound up working on the Kerry campaign (after Faux Noise destroyed Dean). I was talking to a woman also working it, and she was a dentist, and told me how her accountant had tried to get her to buy an SUV for tax purposes, since they were listed as "light trucks", and so...

This was in Brevard Co, Florida, where the highest point is 25 meters.

why not rent a pickup truck when you need one, instead of owning one and have it sit unused?

All kinds of reasons. That I'll admit do not apply to everyone.

While I might not drive it for a month, I might also drive it 6 out of 7 days one week.

When I NEED it I need it NOW. Not after 3 hours of acquiring it for a 1 hour need. Or waiting till the next day. Or Monday.

Small truck / cargo van rentals come in mostly 3 flavors. Rental car companies that treat them as passenger cars. Scratch it and you get to pay for a body shop to fix it. Mostly Nope. Short term places like U-Haul. $20 a day but $.50/mile. A very specific use case. Long term business rentals for weeks or months. Nope. Also the hours of WHEN you can rent or if they have one on a lot near you also make it harder.

Most consumers who rent for a day or few are doing things like taking home an IKEA load, moving to a new apartment, getting a load of mulch for the yard. Things that might need a truck one to a few times per year. I carry things in mine that will not fit into most any passenger cars (even if you didn't mind the load in the car) typically every week or few.

As I said, I'm not a typical use case.

The Youtube channel Not Just Bikes released a video about the truck/SUV-insanity a few months ago. It covers the history, the chicken tax debacle, how the trucks are growing like tumors etc.

Fair warning: Jason Slaughter can be quite opinionated, and in this case he was furious when he made the video. He's still right though, IMO.

These Stupid Trucks are Literally Killing Us: https://www.youtube.com/watch?v=jN7mSXMruEo

Because their marketing dept is pushing bigger (ROI). I have 100% confidence that most of the idiots who can't drive their massively oversized pickups do not ever use it for work, given the pristine condition of most.

And because the US auto industry is staffed by former jocks, and bigger is better, and....

“There is also evidence that (at least in the UK) 'safety' features have probably killed more people than they have saved, even ignoring the pollution aspects, because of the way they discourage walking and cycling, with the consequent reduction of the population's health.”

What is your evidence for this? Current road safety features have pretty definitely reduced direct road deaths. ( https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/556406/rrcgb2015-02.pdf “In the last 30 years Great Britain’s population has grown by 15 per cent (8.1 million people). Despite this growth in population, road fatalities have fallen by 68 per cent...”)

What is the evidence that the safety measures driving this decrease are outweighed by health reductions caused by those measures.

That hasn't been true for decades. Vehicles take years to design, because all the supply chains also have to be built or upgraded, factories have to be tooled, people or robots skilled up, labor contracts done, politicians greased, etc.

Yes. But. Only for new models with new options.

My 2016 Civic came out of an assembly factory in Indiana that made Civics at the time with a list price of $20K to $28K. Maybe more as I wasn't looking at the sport models. It doesn't take years to switch the mix to more at or near $28K from $20K. Days to months depending. And the Civic is a low end car. Teslas at times have had a much wider spread. And to bring this full circle back to pickup trucks I think the low to high end spread on a Ford F-150 and similar trucks from others is maybe 1 to 3.

Another reason was supply chains. In the US if you were buying a car in 2020/2021 you might find a new dealer with 20 cars on a lot set up to hold 200. And most or all of those 20 were waiting for the buyer to show up in a day or two. My son had to buy one during this time. The sticker price was what you paid. Period.

In the US in the past the sticker price for most cars was a starting point for negotiations. The supply chain induced shortages ended that. And if you can only make and sell a very limited number of cars, why not make the most expensive ones that you could sell?

And in general for popular models the dealers would add various local options. In Dallas a Honda dealer put pin stripes on all their cars and marked them up $500. The Honda Civic I bought had anti-theft wheel locks installed. And no option to remove them. I suspect that $175 option cost the dealer $30. And back in the day of restrictions on Japanese cars in the 80s / 90s a local Honda dealer had ADP $500 on the sticker. ADP = Additional Dealer Profit.

it wasnt a tax incentive but an tariff/ import incentive on auto manufacturers, as well as the USA auto companies inability to develop cars to compete with import vehicles, so they focused on light trucks, the classification that includes SUV's and pick-up trucks.

I'm sure the details of the rule making are buried deep but SUVs and light trucks are NOT a part of the CAFE fleet standards for MPG that manufacturers have to meet. So the car manufacturers now market such vehicles to people who really don't need them. I think of it as putting the ruler on the edge of the table and asking all the guys to unzip type of marketing.

You're talking about SUVs and trucks. I was referring to the low end getting more expensive in passenger cars.

Not saying you're wrong about trucks.

ISTR that you're a USian, so perhaps this query is better directed to a USian tax accountant than to a UK writer?

It was a UK writer making a statement about the US tax code that didn't make sense as stated.

While not an accountant, I'm a US sole proprietor who does his own taxes with a daughter in the accounting/audit industry. If it didn't make sense to me then why should I not ask the person making the statement?

This is what got missed in the idea of a 2CV-equivalent EV. Such cars don't accelerate rapidly, and that's a safety handicap.

The original 2CV didn't accelerate fast because it had a 29 hp 600cc engine. Given its weight, that's equivalent to a Tesla Model 3 with 90 hp (instead of the 480 hp it actually has). Oh, and that 29hp/600cc engine in the 2CV was the biggest version -- it originally came with a 9hp, 375cc engine.

I think if you built a modern 2CV with better suspension and tires, and a bit more power, it would be entirely tolerable.

Pigeon is dead wrong, by the way, about small cars needing to be capable of considerably exceeding the speed limit. What they need is to be able to accelerate throughout the legal speed range. And for all vehicles to be fitted with a mandatory GPS-controlled speed governor to prevent speeding so the small cars don't have to exceed the speed limit to avoid idiots in Land Rovers.

Cars of all kinds are in short supply, as several comments on this thread point out. I'm not sure "low supply + demand = prices go up" needs a conspiracy theory?

Have you heard of the term "obesity epidemic" and seen the figures for both that and type II diabetes? A major factor is lack of daily exercise, and both walking and cycling have changed from being main modes of transport to minor ones over the past 70 years. The evidence of causation is from people who have studied the epidemiology of both areas; it's not easy to find, largely because it makes the government look bad, and that means that research into it and publication is definitely not encouraged. There is AMPLE evidence that daily exercise extends life expectancy considerably, that walking and cycling as modes of transport USED TO provide that, and that most people don't do anything like enough.

As a cyclist, I utterly LOATHE drivers who treat the speed limit (and white lines) as absolute. The safe way to pass a cyclist is to wait until there is a suitable gap, floor the pedal, pull over onto the other side of the road, overtake, pull back, and slow down again. The rule fetishists are the ones that pass far too closely, and for far too long.

So it's not QUITE that simple.

And it's easy to get the legislators in on the racket, because so many of them think the unnecessary extra production and activity is actually desirable....

Read "The Midas Plague" by Frederik Pohl. One collection I read it in had a forward where Mr. Pohl was amazed that it was being in University economics classes. It explains sooooo much about modern business.

Eh? The price increases were fixed WELL BEFORE the chip shortage.

There is AMPLE evidence that daily exercise extends life expectancy considerably, that walking and cycling as modes of transport USED TO provide that, and that most people don't do anything like enough.”

But what does this have to do with making vehicles safer? You said safety features killed more people.

Rise of motoring=rise of obesity, yes. Investment in road safety = rise in obesity??

And it’s not just car occupants who have got safer (at least in UK, think rise of mega-trucks in US may lead to different outcomes). See for example: https://www.gov.uk/government/statistics/reported-road-casualties-great-britain-pedestrian-factsheet-2021/reported-road-casualties-great-britain-pedestrian-factsheet-2021#casualty-rates-per-mile-travelled

Table two shows that per mile walked pedestrian injuries and deaths massively reduce over time. How does making walking twice as safe in two decades discourage exercise?

Re. Charlie's reply in comment 248:

Not sure I'll be rereading "Saturn's Children" any time soon. That's not a critique of the writing or storytelling... just that if I want a depressing read, I'll read the local newspapers. Instead, I'll go read something optimistic like "Season of Skulls".

You are not wrong in linking at least part of the obesity epidemic to the increase in driving over human powered transport. There is also a link between reduced traffic fatalities and improved vehicle safety.

Your mistake lies in forging your very own additional link between the two correlations to state that increased traffic safety causes increased obesity. They correlate, they do not have a causal link. You have taken two pieces of steel chain and drawn an additional link between them in crayon.

"And for all vehicles to be fitted with a mandatory GPS-controlled speed governor to prevent speeding so the small cars don't have to exceed the speed limit to avoid idiots in Land Rovers."

Cue the resurgence of off-market GPS spoofing devices... ;) Joke aside, I agree in principle, but wonder whether the Finish/Swedish approach might not be easier to implement: https://robbreport.com/lifestyle/news/finnish-multi-millionaire-nearly-130k-speeding-ticket-1234852538/

Essentially (as little as I understand it) this approach scales the fine associated with a speeding ticket with the speeder's wealth so that the impact of the ticket becomes more equally "painful" to all offenders...

They need to get better at it.

I have a 2020 vintage car that displays the current speed limit on the dash, highlighted if exceeded. It uses a mix of GPS maps and sign reading.

It works pretty well except for when it's hilariously wrong. I'd say it's correct about 95% of the time.

Most of the errors are missing the end of a temporary speed limit or similar, but from time to time it will suddenly decide I'm allowed to go at 180mph.

Please do not misrepresent me. I did not say that. I said that there was evidence for a causal link between vehicular safety features and the reduction in walking and cycling. The relevant period is before the one quoted by Johnny99.2 because there have been almost no significant improvements in vehicular safety features in the past 20 years, and the earlier period is when the demise of walking and cycling occurred.

The NTS is one of the most misleading sources of data imaginable, and has got worse since it was privatised; if it is to believed, both the (car) distance driven and number of trips are dropping steadily. Well, why do we need more road capacity, then?

If you look at its summaries, you will see that walking is not a significant mode of transport where it might come into conflict with cars (e.g. the ,ean trip distance is 0.7 miles), and cycling is little better. Since the government privatised it, it hasn't been possible to see the data but, in the last year it was available, it was clear that most of the distance and trips was either uses like taking small children round a park, or a small number of dedicated cyclists (15+ miles a trip). The mean distance and number of trips is misleading, small enough though they are.

No, I can't face the thought of reverse engineering better data from their summaries, though I have done it before :-(

Strongly disagree. The effect of making the "speed limit" a hard cap will be to make overtaking anyone moving at 75% of the extra-urban limit on dual track roads impossible unless you get mile long straights. Now, if instead we make the speed limiter allow full throttle acceleration for, say, 30 seconds after which it becomes a soft limiter and will not allow reapplication of power until the vehicle speed drops below said limit, when we may, repeat may, be getting somewhere with still allowing drivers who make progress at the speed limit to do so.

EC on temporarily exceeding the limit to get past cyclists - as a fellow-cyclist - SPOT ON. Recently, in the L-R I was approaching a long hill which has a bend half-way up - speed limit 60 mph ... the bimbler in front of me was halfway out, doing 35 - I flashed them, they SLOWED DOWN, so I floored the pedal, hit about 55 & was past them.
In a car, you are safer in front of idiots like that, compared to behind them.
OTOH, last week, I cycled more-or-less the full length of the main part of Epping Forest at dusk - as soon as I could, I got off the main road & went down the back parallel road & then the gravel-tracks ... much safer.
AND I saw 5 foxes & a Muntjac

I wasn't referencing anything by NTS (assuming we refer to the same acronym) but rather my decades of marriage to a transportation, cycling and health activist. Some of the details of her work manages to stick to my faulty brain.

Vancouver has, over the course of decades and a tremendous amount of foot stomping by the car lobby, managed to build out a fair amount of quite decent cycling infrastructure. Aside from the dry policy side of it, I experienced the improvement directly over 20 years of cycle commuting - augmented by transit where necessary and the use of the local car co-op.

There is strong evidence that increasing the safety - real and perceived - of cycling increases the use of cycling. Decreasing it increases car use. None of that is related to the safety for occupants of a car - improving that safety is a good thing.

There is strong evidence that increasing the safety - real and perceived - of cycling increases the use of cycling. Decreasing it increases car use.

Anecdotally, I haven't cycled in decades — ever since the 407 was built. Too bloody dangerous crossing the exit ramps, because drivers don't slow down and are coming over a blind hill — and they hit the bike path before they hit the traffic lanes. And I know from decades of car commuting experience that 407 drivers are more careless/entitled than ordinary drivers…

Note to non-Ontarians: the 407 is a toll highway bypassing Toronto. Originally conceived as a bypass to relieve traffic in the city. Built as a public-private partnership, sold off by Harris' conservative government to balance one budget. Highest tolls in North America. Could have enforced speed limits because entry/exit points are time-stamped and linked to transponder or license plate, but that idea was scotched by the company that runs it on the grounds that sending speeding fines to their patrons would deter use and hurt their business…

Sompletely off topic, but we're past 300 so: the kiwis have just had another review into the voting system and it's reached much the same conclusions as the last few with a minor tweak around the recent court decision that barring under-18's from voting is not compliant with NZ's human rights legislation. There's some good points made in the review but the ruling class are already making use of their fainting couches.

They recommend 16 as the voting age. The 1986 Royal Commission... strong case for lowering the voting age to 16 (also 2011, 2014, 2017, 2020 all suggested "debate" and 2022 said 16 for local government elections)

Changes to donations and disclosure of donors:
* Permitting only registered electors to make donations and loans
* Requiring the disclosure of all donors and lenders who give more than $1,000
* Expanding the definition of donation to include a range of fundraising activities.

https://electoralreview.govt.nz/

I'm mostly amazed that one of my strong preferences has finally made it onto the list (only voters can donate). It's upset a lot of rich people, especially non-resident ones.

The next step is to tie voting rights to tax resident status. If you're not a kiwi for tax purposes should you really be voting?

You mean sort of like how Scotland has 16 as the voting age for all elections except Westminster ones?

you are safer in front of idiots like that, compared to behind them

I'm pretty sure they felt safer with you ahead of them, too.

drivers who treat the speed limit (and white lines) as absolute

I can't agree about the speed thing, but in terms of white lines it's interesting. In the state of Queensland it's forbidden to cross a double white line with one exception, and that's when it's necessary to provide the mandatory 1.5m clearance when overtaking a cyclist (more is encouraged). I believe the exception was written into the highway code at the same time as the rules about 1.5m clearance were brought in, though it's not such a burning issue that I am inclined to go check. Queensland is relatively cycle friendly from the point of view of the law (and there's a lot of investment in cycle infrastructure, it seems to get bipartisan support though that could be because it's largely the wealthy, conservative-leaning electorates who lobby for bike paths and lanes), but the motorist culture in Queensland leans hard in the opposite direction, with a lot of people who seem to see speed limits as a guaranteed minimum you should be able to go anytime anywhere and anyone in your way is (clearly, obviously) in the wrong.

Too bloody dangerous crossing the exit ramps

I find this really interesting. In Queensland cyclists are forbidden from motorways, but not open highways out in between places, though I'm really not sure where the transitions between these are, or even how you'd get out to the open highway on a bike, since the motorways are often the only road. I have noticed in NSW on our recent road trip there are cycle lanes most of the way along the Pacific Highway (M1/A1/M1), with the entry and exit ramps having signs warning motorists of cyclists crossing. Some even have dedicated cyclist crossing zones.

Completely off topic but time to look up Scalzi's Schadenfreude Pie recipe. Gotta celebrate Ol' Agent Orange being told he has to be in court next Tuesday at 3 pm for a very special date with a judge. And with an indictment or three under the US Espionage Act.*

Couldn't happen to a more deserving guy, IMHO. Now off to find that recipe...

*Bets that he'll seek political asylum in Moscow over the weekend? Or Riyadh?

Very similar. But with the amusing effect that those dang youf took what was offered and promptly demanded more. If 'they' don't want youf thinking about politics maybe don't fuck youf over politically? Just an idea.

The "road user suggestions" are much the same in NSW as far as bicycles go. But as with all suggestions few motorists know what they are or care to follow them if they do. I occasionally wonder whether motorists using that clause know they're doing so or just DGAF about the suggestions painted on the road either.

Pacific Highway (M1/A1/M1), with the entry and exit ramps having signs warning motorists of cyclists crossing. Some even have dedicated cyclist crossing zones.

Usually those are designed so they're out of the sight line of motorists using the exit for as long as possible. Which seems backwards to me, but that's how they're drawn. Motorists and busy exits are a bad combo at the best of times, just starting to slow down and suddenly there's a cyclist crossing slowly at right angles to their path just seems like a recipe for disaster.

We won't discuss the "suggestion" that cyclists ride part-way down the exit, perform a right angle turn, then ride back to the high speed route they were on to continue their travels. It's not relevant to motor transport engineers, it's not relevant to us.

In completely unrelated news, on busy roads I like to have a really bright constant rear light plus a really bright flashing one. As well as a rear-facing camera, reflective tape on the and reflective gear. I want the coroners report to fall back to "well he shouldn't have ridden on the road".

Shortly after I finished grad school and we moved back to Vancouver I was unemployed and trying to fix a shelf we had in our house. All I needed was a 25 cent piece from Ikea, which was just on the other side of the Knight Street Bridge. If you don't know, said bridge is a major thoroughfare for heavy trucks, commuters and all other traffic through the city, which terminates/originates at the Port of Vancouver.

Being as I was unemployed and apparently had no real sense of risk/reward, I decided that I might as well just ride my bicycle over the bridge, buy the piece and ride back. I did not anticipate the multiple blind off and on ramps that I would have to cross. There was a separated sidewalk that was about 15mm wider than the handlebars of my bike.

I distinctly remember crossing the second blind, high speed offramp and thinking 'I'm going to die for a 25 cent piece of shelf'. Easily one of the most dangerous hours of my life, and I've ridden out a week long hurricane force storm in a fish boat off the West coast of Haida Gwaii.

RocketJPS
Agree 150% - particularly since I got Wiley E Bicycle & now always wear a helmet & an (Rail-Orange) high-vis.
My greatest fear is of E-scooter fuckwits - no lights, no helmets & do not appear to look where they are going .... colliding with one of them & being thrown ito the path of something heavier isn't a funny prospect.
...
Damian
Yes, they could safely bimble on, paying no attention AT ALL to the traffic conditions, until they drove right off the road!
{ Not only were they going ridiculously slowly, they were "wandering" - i.e. not following a consistent straight(ish) path. }

The figures quoted by Johnny99.2 were. But, to address your specific points, yes, you are right, BUT:

As I said, there is evidence that seat belts increased driver speeds by 10%.

The improved handling of modern cars has led to drivers taking corners faster. That's a particularly bad one, in the UK.

The 'safer' cars are also wider, and mean closer passing of cyclists and pedestrians.

And so on.

As I said, there is evidence that such changes discouraged walking and cycling. Note that I am NOT saying that the same is true for other safety 'improvements' though, as I and others have said, many of them almost certainly increase the danger to vulnerable road users and, in many cases discourage them. The term 'psychle farcility' is all too apt.

The reduction in cyclist and pedestrian deaths in the UK is primarily due to to discouraging them on most transport routes. A walking distance of 0.5 miles (the median in the UK) is a typical distance from a car park to office or shops, for example.

Quite so. You're always better off with the idiots in front of you, where you can see them. If they really want to go faster, they'll disappear ahead. If they just wanted to go first, well there will be an area where it's genuinely safe to overtake again eventually. For the ones who come up behind you flashing their lights when it isn't safe to overtake, you always let them overtake anyway (and slow down to make it easier) because otherwise they will do something really dangerous.

Now, if instead we make the speed limiter allow full throttle acceleration for, say, 30 seconds after which it becomes a soft limiter

That's how GPS-controlled mandatory speed limiters work in the real world. (They make allowances for emergency acceleration, they just prevent continuous driving above the speed limit for long periods.)

I have had such a system nearly create a collision - admittedly, because it was designed by a Complete Idiot. 30 seconds would usually be enough, but not always, and then you get stuck in a dangerous position.

I was overtaking a slow cyclist, floored the pedal, got half way past, and then the engine dropped to an idle. It turned out that it limits me to 2/3 of the maximum RPM in first gear (but not in second or third - I can't speak for fourth and fifth because I don't drive that fast). Luckily, I managed to complete the manoeuvre (the alternatives were an emergency stop in the face of incoming traffic or to run into the cyclist). The emergency stop wouldn't have been catastrophic, provided the oncoming driver was awake, but I try not to rely on such things!

No. You are only better off if they are relatively fast idiots. Many drivers mishandle narrow, twisting roads with hedges, which I learnt to drive in. While they are on average slow, they take the corners too fast and brake hard unpredictably. Also, they block the vision of the car behind.

However, whether or not I would be safer in front, overtaking is rarely possible on such roads unless they cooperate or I know the road very well. 'Eventually' can be half an hour's drive further on. The only viable strategy is to hang back and curse the idiot in front.

This is why lane narrowing as a safety measure is a danger to cyclists - it creates a serious conflict between drivers and them, because it can double the trip time for the drivers. That was one of the reasons I had to give up cycling to work, after doing it for 25 years.

You are only better off if they are relatively fast idiots.

You're saying I shouldn't let Greg overtake then?

This whole conversation just tells me that we need tunneling cars. Overtaking isn't an issue if you can exploit the 3 dimensional freedom of movement available within the earths crust.

Of course some sort of self driving will be required as most people aren't mentally equipped to interpret sonar data but given the likely speeds involved it shouldn't be too difficult.

There may be minor issues related to buried infrastructure, spoil heaps and inconvenient releases of pterodactyls but nothing that can't be dealt with by regulation.

Not only were they going ridiculously slowly, they were "wandering" - i.e. not following a consistent straight(ish) path.

That's classic drunk driver behaviour. Are you sure they weren't over the limit?

Why not harness your inner Fred Flintstone and ride a pterodactyl?

Because if I could do that then it would be a convenient release of pterodactyls rather than an inconvenient one, thus outside the scope of discussion.

Greg Tingey @ 331:

AND I saw 5 foxes & a Muntjac

Are there jackalopes in the UK? (I know Muntjacs exist. To my Canadian sensibilities, it sounds like a fake animal, though).

Heteromeles @339:

Bets that he'll seek political asylum in Moscow over the weekend? Or Riyadh?

Maybe R'lyeh? (we can hope).

Can DJT run for election while resident in a foreign country?

No, but you have to watch out for wild haggis in the Highlands.

https://www.thehaggis.com/wild-haggis-all-about-haggis/

Unfortunately, if he took asylum in R'lyeh, OGH might respond as Lehrer did to Kissinger's Nobel Prize for Peace.

In Queensland cyclists are forbidden from motorways, but not open highways out in between places, though I'm really not sure where the transitions between these are, or even how you'd get out to the open highway on a bike, since the motorways are often the only road.

Here too — bicycles (and pedestrians) forbidden from dual carriageways (which the 407 is).

The problem is the design of the exit ramps. The private part of the public-private partnership had decision-making power and used it to take as many shortcuts as they could get away with. One shortcut was shorter ramps, and the street I'd need to cycle along had a particularly short ramp exiting onto it, which also had a hill so anyone leaving the 407 there had a sharper-than-usual turn over a blind hill before they got to the street.

So you have drivers travelling (on average) faster than usual, and also averaging more entitled behaviour than usual, having to react faster than expected. After too many times witnessing them pulling to a halt barely before traffic (and well into the crosswalk) or not even bothering to stop at the red light because there were no cars coming I decided I'd rather not risk it.

I commuted along the 404 for years, and if there was an accident it was almost invariably at the 407 interchange, either someone getting impatient merging or trying to cut across four lanes of traffic to get to the HOV lane*. Even the 404/401 interchange, which was much busier, had fewer collisions.

I hypothesize that the psychology is a combination of "I've paid a lot of money to use this road rather than wait in traffic, so I should get priority" combined with the well-known tendency of richer people to be less empathetic (and anyone who can afford to use the 407 daily is wealthier than average).

*HOV = High Occupancy Vehicle. Ie. a lane reserved for vehicles carrying multiple ordinary people, or one rich git who's too important to wait with the plebs.

It turned out that it limits me to 2/3 of the maximum RPM in first gear

What is this "gear" you speak of, old-timer?

(Seriously, EVs generally have two gears: forward and reverse. A handful of supercars have two forward gears, but in general EVs don't have 'em unless they're designed to lay down serious rubber tracks at speeds above 200km/h.)

This whole conversation just tells me that we need tunneling cars. Overtaking isn't an issue if you can exploit the 3 dimensional freedom of movement available within the earths crust.

As I keep SHOUTING, all discussions of methods of personal transport are tap-dancing around the essential fact that they all entail sub-optimal trade-offs until we get teleport booths.

Fricken' teleport booths, people. That's the only really conclusive answer.

At least I don't still use a crash box - and, yes, I almost certainly still could do so - but I do regret not being able to rear-wheel drift :-)

Teleport booths work for me as a means of commuting, getting to and from Trout and/or SF cons...

What if it turns out that the only way to make them work is to implement the scanning and reassembly mechanism using a computational assembly of quantum pico-demons?

Mikko Parviainen @ 271:

So, there's this world where people have done the eradication of homosexuality, wanting to go to that (fantasy) 1950s (American) suburbia 'utopia'. It works! No annoying homosexual deviants!

I've sometimes wondered about doing a satirical story (working title "The Cure") where someone discovers how to precisely disable any given unwanted sexual impulse (mumble precision MRI mumble proton beam). The obvious goal of the researchers is to cure (for want of a better word) paedophilia. But then the anti-gay lobby start arguing that even if being gay wasn't a lifestyle choice before, it is now. Nothing about this resolves transexuals because that (AFAIK, being cis myself) isn't a matter of sexual attraction. But that doesn't stop people believing that it does.

Cue lots of messed up thinking about various paraphilias, people who don't want to be heterosexual either, people who have this forced on them and find themselves now asexual, and so on.

(For added yucks, the technique only works on someone who is receiving the target sexual stimulus while in the MRI scanner. So the original researchers had to figure out how they could legally use CSAM in their experiments, on sex offenders who might or might not be entirely willing volunteers. And then it turns out the use of CSAM wasn't actually that legal. And looking at porn turns out to activate different bits of sexuality than actually having sex, so its not quite effective. And so we're in an ethical minefield on a pogo stick, as someone once put it)

Teleport booths would be delightful. Personally I'd love someone to discover personal teleportation a la 'The Stars My Destination'. The knock-on effects of removing travel time for everything would be really incredible (and barely touched on by Bester).

I've sometimes wondered about doing a satirical story (working title "The Cure") where someone discovers how to precisely disable any given unwanted sexual impulse (mumble precision MRI mumble proton beam).

Peter Watts wrote a non-satirical story using that technology. "Eyes of God" asks, among other things, are you a criminal/sinner of you feel the urge to commit an act you know is wrong, but have the self-control not to do it? What if your urges could be made visible? What if they could be eliminated? Would choosing not to have them eliminated become a crime/sin?

Not easy, but a good story. As is usual for Watts.

https://talesofmytery.blogspot.com/2013/09/peter-watts-eyes-of-god.html

Charlie @ 353 - re "drunk-driving"
I have absolutely zero idea & I didn't want to find out the hard way (!)

EC 346
The reduction in cyclist and pedestrian deaths in the UK is primarily due to to discouraging them on most transport routes. A walking distance of 0.5 miles (the median in the UK) is a typical distance from a car park to office or shops, for example. - Which I will NEVER do in the L-R - unless I have to transport something really bulky & awkward.
For short journeys & small loads, I have the aforementioned Wiley E + a load-carrying small trailer if I can't get "the load" in my rucksack, or on the back rack..
Like this ....
Wiley E Bicycle - and - Trailer

First thought: someone hacks the machines, to turn CSI gay, or gender fluid.

I can see a lot of suicides and lawsuits.

whitroth
Talking of lawsuits ... how's the Rump shitshow going?
I assume his side will try to stretch it out as long as possible?

This is two separate indictments. We're expecting more, including from Georgia. And some of that may be state, which a President can't pardon.

Short term he will do what he always does and use it to wind the grift machine. For reasons of public insanity he will receive a lot of additional donations etc.

Long-term he is likely to be convicted. Very rarely do federal prosecutors in the US charge someone unless they are almost 100% certain they can get a conviction. Triply so in this case, because they will be under a microscope. I suspect they have irrefutable evidence of blatant lawbreaking (no doubt he bragged about it on camera/sold some of it and kept the receipts). He will also not stop talking, which helps prosecutors immensely.

He is also likely to continue his electoral campaign. Given that the Republican party has become entirely bereft of quality candidates, he may well end up the nominee, and possibly even the president.

In the event he is convicted he will probably just die rather than spend 1 minute in prison. Alternatively he will suddenly manifest a debilitating ailment that means he cannot possibly go to prison (but then brag about it shortly afterwards).

Alternatively he will suddenly manifest a debilitating ailment that means he cannot possibly go to prison

He's got form for using convenient medical excuses…

On EV speed and acceleration & limiters. I have a fairly middle of the road small family car style EV. Range is fine for local stuff, annoying but manageable for long distance.

It will out accelerate petrol cars 3 times the price, while looking like a toy.

It clearly has a limiter as it goes from good acceleration to none at an exact multiple of 10mph.

On teleport booths: yes so long as there is still a way to go the last mile (or ten), and if you can convinced me I haven't just violently wasted someone very similar to myself.

Seriously, EVs generally have two gears: forward and reverse.

I don't think there is even a gear for that, they just run the motor in reverse.

So your hypothesis is:

A) increased car safety measures (seat belts + wider vehicles) lead to a more hostile and dangerous environment for cyclist and pedestrians (faster cars taking up more road)

B) a decrease in walking and cycling results from this

C) the resultant health impacts from this decrease (people taking less exercise) outweigh the safety benefits meaning a net negative effect.

Meaning we’d have better overall outcomes if we all drove 2CVs with no seatbelts? More dead motorists, lots more healthy pedestrians and cyclists

(Only pursuing this as we over 300 comments, obviously happy for OGH to tell me to shush )

Alternatively he will suddenly manifest a debilitating ailment that means he cannot possibly go to prison (but then brag about it shortly afterwards).

Nope. Doesn't work that way. He would likely wind up about 30 miles from here at Butner. It's where Bernie Madoff spent the last years of his life. Low security but still behind razor wire. You can drive past it. I have.

If someone there needs serious medical care they equip a room or two and bring in docs from the area. Duke and UNC are world class in terms of medical care if you want to spend the $$$. (US total medical system not with standing.) And if needed both have the facilities to house someone at their hospitals under tight guard. Duke is where the uber rich show up for medical car with an entourage of 100 or more at times. The bills for their care and feeding supports all kinds of research. Anyway Duke was have the facilities for someone like Trump and a team of Secret Service agents if needed.

Sorry, I could have sworn he made it clear that was NOT what he was saying. Rather that a lot of the "safety measures" made in the last 20 years (come on, seat belts and laws about them are so 20th century).

I think a lot of his thoughts could be summed up in the half-hour video.... https://www.youtube.com/watch?v=jN7mSXMruEo

Please drop this line of discussion. It's annoyingly confrontational and won't produce anything of interest to anyone else.

Boris Johnson is to step down as an MP with immediate effect after receiving the Partygate report