« That old-time new-time religion | Main | Excuses, part 673051 »

The future, today (maybe)

Some news items from the future:

Synthetic oil fermented from crop waste using genetically modified e. coli — looks like someone in Silicon Valley's decided that $140/barrel crude is a good enough reason to fund a dash for Oil 2.0. (Let's hope it's not snake oil or another long scam like the car that runs on water: carbon-neutral synthetic oil is exactly what we need right now.)

Meanwhile, here's a new spin on an old material: paper is the new cast iron, or so it seems; cellulose molecules extracted from wood pulp can be processed into a material (dubbed "nanopaper" by its inventors) with a tensile strength of 214 gPa (compared to 250 gPa for steel and 130 for cast iron). More here.

(Random thought: right now, most 3D printers, including reprap, manufacture items by extruding layers of plastics such as polycaprolactone or polylactic acid. These substances are useful but they're soft and they lack tensile strength. If it turns out to be possible to deposit nanopaper in layers, the future may turn out to be made of papier maché.)

Mind you, not everything that comes out of a rapid prototyper is good. Here's the Magpul FMG-9 prototype: and here's some more. Is it a flashlight? Is it a submachine gun? Who knows? Here's another baroque weapon that probably started life on a rapid prototyping machine. If reprap-like machines with strong materials turn out to be cheap and easy, then never mind licensing handguns — we're going to have a problem with home-made crew-served weapons. (Reminder: yr. hmbl. crspndnt. lives in a country where, for better or worse, possession of a pistol by anyone who's not in the police or military carries a mandatory 5-year minimum prison sentence. The implications of rapid prototyping machines for this sort of legislative environment probably parallel the effects of peer to peer networking on music industry cartels.)

What kind of society are we likely to get if it turns out that yes, we're hitting peak oil round about now, but that it's possible to process random junk biomass into crude oil for $100 a barrel, and $1000 will buy you a machine that you plug into your laptop and that can make, well, just about any small macroscopic structure you can design, out of feedstock derived from biosynthetic crude oil or woodchips, or paper?




Available to download already: RepRappable solar collector system. Just add stirling engine.

On the broader point, "anarchy" might cover it - either in the sense most people use the word (DR Congo, but everyone's got manpads and basic drones), or in the sense *anarchists* use it.

Completely off topic, here's some World of JENNIFER MORGUE watch. Inexplicably tiny deep-sea subsurface organisms that may form a community of "metacells", each one made up of several individuals.


Or, relating to the AI topic last post:

Feed: KurzweilAI.net
Title: Roadrunner supercomputer puts research at a new scale

Los Alamos National Laboratory researchers are using the lab's new Roadrunner petaflop supercomputer to model the human visual system, as a test. The "PetaVision" simulaton models more than a billion visual neurons, reaching a new computing performance record of 1.144 petaflop/s. Based on the results of these trials, Los Alamos researchers believe they can study the entire human visual cortex in real time. Applications could include "smart" cameras that can recognize danger, or an autopilot system for automobiles that could take over for incapacitated drivers in complex situations such as navigating dense urban traffic. See also: Military Supercomputer Sets Record (Source: http://www.physorg.com/news132497981.html)


Speaking as a Materials Scientist, properties like yield strength and elastic modulus are quite important... and to say that it has the "toughness of cast iron" is right out.


3D printers can print metals too; the raw material is metal dust. A thin layer is put down and a laser scans over it. turning on over the places where the part is to exist. The laser sinters the metal, effectively welding the grains together into a structure that has the metal's electrical and heat conductivity, but is lighter because of the voids between grains. It's not as strong as solid metal, but stronger then the plastics that are used in 3D printers. Currenty laser sintering machines are more expensive than plastic extruders, but that's mostly the cost of the laser; the rest of the machine is very similar.


>> "What kind of society...?"

One that un-learns human factors. Have you seen some of the crap people knit?


The takeaway should be that the disruptions caused by the Computer (and Internet) Age to non-scarce resources (media, software, etc.) are coming to the World of Atoms. Computers plus networks plus reprap makes weird things happen real fast.

We can fret over the details, but the big picture is pretty clear IMO. Absent super heavy-handed legislation Real Soon Now, this horse will be out of the barn (for better AND worse) way before anyone "in charge" even starts to grok the consequences.


"Let's hope it's not snake oil or another long scam like the car that runs on water..."

And sodium and/or lithium borohydride (probably), which admittedly sounds way less sexy. Not sure I'd classify it as a "scam" though, just overhyped.


Oil 2.0 beats boiled Whale as a fuel source but does paper maché furniture mean we are heading back to the Victorian era.


It's nanotech, Jim, but not as we know it.

I can barely wait for the new online iKea, where every wardrobe, chair and vase costs $0.99/download.

As for Oil 2.0 - weren't we already producing methane from biomass, like, last century or so? Petrol to LPG engine conversion is trivial, and you can easily adapt the old infrastructure to support the new fuel - most countries in Europe already did it, to my knowledge. The problem was (and remains) the scalability of the whole process.


For one thing, the Open Source movement becomes 3D in a big way, and a lot more demographics become interested. This may be the death-blow to closed source culture, unless something else comes along to put it out of its misery first.

And it's like nanoassembler training wheels.


Also: the copyright laws? You ain't seen nothing yet. As a trainee IP barrister of my acquaintance put it, this stuff is going to do for patent, trade secret, and IP law in general what p2p file-sharing did for copyright law.


I'm a little concerned about what happens when some Brazilians gets their hands on oil-shitting bacteria and turn all of Amazonia into feedstock for it.


This (minus the fuel bit) reminds me of Cory Doctorow's story Printcrime.


Bespokery and customization dislodge brand-loyalty for all extrudable commodities, and all the imaginary value imparted by brand evaporates (good riddance). Hopefully this catches on in areas that remain unaffected by extrusion.


Kids get (and get to design) totally wikked toys.


On the other hand, I wonder what happens to nanopaper when it gets wet? And, so far, rapid prototyping in hard materials is a rare, expensive technology. It's relatively easy to make castings from soft materials through a lost wax process, but that does take some skill; it's still easier to machine most parts.


How about something simple?

Fitting shoes and clothes.


tp1024: if you think clothing and shoes are simple you obviously don't share accomodation with someone who makes her own.

(I suspect those teraherz wavelength body scanners that are being rolled out at airports are going to start showing up on the high street in the next couple of years, so you can have off-the-rail clothes tailored to your precise body shape in a factory somewhere, but despite the availability of knitting machines, we're a long way from having home fabricators that fab fabric.)


Ooh, Santa Claus machine. Another SF concept after the invention of which the future ceases to have any reliable predictabilty.

Is it Singularity Month or something?



What kind of society? The right one. I guess it's my redneck roots showing, but I really like the idea of self-reliance and decentralization. I think the world's going to get a lot bigger, and frankly, it's about damn time.


Alex@1: Regarding the nanobe stuff, look at the date. Nanobes are... equivocal at best and most people don't think they actually exist.

(For a nice horrible example of what might happen if they did exist and we acted stupid enough around them, see Peter Watts's _Starfish_ trilogy.)


Charlie @18 The fashion industry has been toying with 3d scans for years. There is a problem with the armpit/ groin areas in fitting this way. Are you scanned doing star jumps or standing to attention? Okay there are algorithms dating back to the victorians to solve this, but they have to be applied .
The problem is the bespoke-ness of clothes that fit. Why do you think a hand made suit costs? The apparel industry makes its money on the fitting as many pattern pieces on to a length of cloth as possible and cutting out a six inch pile with a laser. The more banana/star shaped the pattern pieces the more waste. That's why you can't walk in cheap trousers or lift up your arms in cheap suits. Its also why you find clothes made up of hundreds of little bits, it reduces the banana problem (all polygons can be made from triangles). And conversely why NEW LOOk suits featured 10 yrds of fabric in the late '40s.
Since the invention of the sewing machine * the cost of construction is negligible compared to the setting/cutting out. Which can't be easily automated as outlined above for bespoke-ness.

As to general lack of fit. Apart from meanness of cloth or 'there's no call for it'. The industry is still using 1950's measurements of the population and we all know how much we have changed shape since then.

*The invention of the sewing machine was stalled for a long time, because people were trying to reproduce the action of hand sewing. The final suite of designs involves processes that have little relation to hand sewing. The first machines produced chain stitch, and were used to make uniforms for the French navy, unfortunately if the seam was damaged it completely unraveled…


Yes, I know it is not that simple.

Well, this was the same kind of "simple" as in finding a cure for the common cold. It's not that you couldn't live without it, or that it would make such a huge difference that spending a few billion Euros on it was a no brainer, but it is so convenient that it MUST show up as soon as the technology is available. So it should be a good indicator of progress.

But then again, I think that this time the problem really is a lack of trying. Except for some clothes and items that are liable to be out of fashion after the 6 weeks it takes to ship them from [insert current developing country here] to the rest of the world, almost all clothes are made in remote places, in fordistic low wage environments that offer zero customization and almost zero automation - as that would only drive up initial investment costs that have to be paid for with money that just isn't there. (And no, almost no matter how high oil prices will go, ships are reliable and cheap means of transport, and nothing short of a 100 fold increase in marine piracy will change that.)

I'm pretty sure that if there was something like a textile industry left in Europe, clothes on demand would be here rather sooner than later. You don't have to make them at home, some things just don't make sense in terms of economies of scale anyway. I don't know which ones though. Anything that is small enough and needed in small quantities should work out.


The implications for the third world are staggering. Especially considering your previous post on flash mobs/modular housing. Fabbable playpumps alone could save millions of lives. If the fabs are fed by junk biomass and solar power they're almost totally self-sufficient. Farming and irrigation implements. Man/animal-powered farming machinery. Bikes. PC/laptop cases. As far as the OLPC vision, unfabbable componentry is still the bottleneck, but it becomes a helluva lot cheaper (sub $100?).


tp1024 @ 23 I wouldn't count several hundred industrial sewing machines zero automation.
Fashion is one of difficult to define ones. It takes two years to produce 'something that will be out of fashion within 6 weeks.' If your a teenager who cares about what their peers think your going to by clothes for pennies and customise them. That's one of the founding definitions of teenage.
The textile industry in Europe concentrates on high value, low unit number, because that's the only way they can afford to pay their work force is with lots of shiny value added.
About ten years ago Sultex developed a brand new way of weaving x3 faster than standard, hoping to preserve the weaving industry in Europe (more out put for the same staff). I guess china is their biggest customer now

I guess a five bed knitting machine and a cunning programmer would be the best way forward for bespoke soft 3D things, both cost money, and as Charlie said knit isn't for every thing.

Unless some AI guy in is black t-shirt and jeans is going to spend a few years learning an expert system in sewing the best way to get bespoke clothing at affordable prices is to re-negotiate the cost of living.


On clothing:

tp1024: You don't have to make them at home.

What if:

* You are a non-standard size? (As Maggie commented, current clothing sizes on the UK high streets are calculated to fit 80% of the population who were measured for the sizing charts ... in the early 1950s!)

* You want something that's properly made out of good quality fabric but don't want to pay someone else to do the work at developed-world labour prices? (Again: much high street fashion clothing is abominably badly sewn and finished precisely because it's designed to be discarded and replaced after 6 weeks.)

* The particular item you want is out of fashion and not likely to come back for 5-10 years?

Last week I went shopping for a black velvet jacket to replace the one I'd been using for two years. "Oh, they went out about nine months ago," said the nice man at the suit shop. Feh. I don't care about what the rest of the population want to be seen in, I care about my appearance, and insofar as I am not a fashion herd follower, this offers endless opportunities for annoyance.

Anyway. The market is under no obligation to supply me -- or you -- with precisely what I -- or you -- need. Periodically the fabric manufacturers decide to change colour on us, and the clothing manufacturers decide to switch us to a new style; at which point, everything turns beige, or straight-leg jeans become unavailable because everybody is expected to wear flares.

And thereon probably hangs a thread.


Nifty (in a scary sort of way) though that FMG-9 may be, it still uses a fairly conventional automatic pistol for the working parts, and doesn't seem to be anything that would be beyond a skilled machinist with a half-decent workshop. I'm not sure the ability of people to make guns in their home office would change the law much mind, although I would imagine it would increase the number of armed police officers.

As for textiles and the fashion cycle, ISTR Alex talking about companies re-opening Yorkshire textile mills to be able to get designs to market at Topshop and H&M and the like in about 6 weeks without having to wait the same time again for the clothes to arrive from China.


Tremble for the future day of Peak Woodchip! It will be at least as terrible as the dark day of Peak Whale!


> The textile industry in Europe concentrates on high value, low unit number, because that the only way they can afford to pay their work force is with lots of shiny value added.

Exactly, but that is mostly because there are cheap alternatives in developing countries in Southeast Asia and Africa that can undercut European prices any time - without prior investment. Without cheap imports they could probably produce clothes locally at reasonable prices (probably much less than a 50% mark up, due to higher productivity - with adequate investment) and even provide jobs to mostly unemployed people without adequate education for other jobs (part of today's reality). Higher prices would also mean that fewer used clothes are exported to the least developed countries - that have destroyed the local textile industry there. (In about the same way that imports of cheap subsidized food did to their farming. That said, I should add that I'm all in favour of globalisation, I just don't make a religion out of it.)


I've done some thinking about the future of rapid prototyping and the "Boutique Economy". The IP problems are a nightmare, though there would be an interesting new niche for open source designs.

But it would basically make all the problems with have with software and media piracy, and extend them to physical objects too. Anything that's designed for a rapid prototype machine can be copied and recreated.

I hadn't thought about the application of the technology for criminal weapons production though. Probably because legal and illegal weapons are already cheaply available here. It won't be the end of the world, but it may give police a lot of trouble... Though I can't wait to see some of the insane hillbilly guns rednecks design...

If the technology becomes advanced enough, there are some serious military applications. What happens if Country A steals the fabrication files for Country B's new tank? Or how about on-site arming of troops? Fly in your soldiers and manufacture their gear on the battlefield...


People, come back to the real world, please. No way home manufacturing is going to cost less than a mass-production in a foreseeable future. No way you could build any sophisticated weapons like a tank with this stuff.



> You are a non-standard size?

Check. (Just a tat too small to be considered a grown-up male by the industry.)

> Good quality at developed-world labour prices

Would be much lower with economies of scale.

> is out of fashion and not likely to come back for 5-10 years?

All the time.

But what if we were talking about books? If a book is out of print, has a readership of barely above one, thus very unlikely to get back into print, and you don't quite feel like going to a copy shop and get a shoddy bound block of paper that defies the notion of "book" in all but the letters-printed-on-paper sense at developed world prices - what do you do?

Well, if you happen to want to have the right book, just go and order it at books on demand - who clearly do have to deal with developed world wages and still manage to be not that much more expensive than others. Having the same for clothes - clothes on demand if you will - would be a start. (I want that pant[shoe] design, just in size S[40].)


At my house we recently acquired a state of the art device, I'm sure many of you will be jealous.

The system can generate a myriad of clothing designs, has an intelligent interface capable of taking basic orders in and can process raw food supplies into comfort food.

We can power our new device on tea, short bread and fifteen to one.

We have named it Grandma. During the purchase phase we managed to recycle a house for a new family and reduce our carbon footprint due to lack of travel on a Sunday.

I recommend a Grandma replicator, ours is the Val Doonican 5000i - it's the future.


You build DRM into the fabs.

Right from the start.

So ther's no equivalent to the bog standard MP3 player.

Guns, I expect revolvers. The springs in an auto pistol are critical, and need very specific characteristics.

And any sort of locked breech is going to need fairly specialised materials. A revolver, you can use a Whitworth barrel to get around the rifling problem, and you can make the chamber walls and barrel much thicker.

But where do you get the ammo?

And a harmonica gun would be another way to get around some of the limits.



Better plumbing/glue/duct-tape for interfacing preexisting but incompatible hardware. Like real-world Perl for DIY projects.

Items are extruded as a solid unit, so seams disappear.

With enough power and feedstock you can have (power of 2) fabs in logarithmic time, cranking out an ass ton of whatever. (struts and panels for geodesic domes, ball bin, instant Carnival, massive buoy platforms, mega-scale food processing)


You build DRM into the fabs.

Right from the start.

So ther's no equivalent to the bog standard MP3 player.

Guns, I expect revolvers. The springs in an auto pistol are critical, and need very specific characteristics.

And any sort of locked breech is going to need fairly specialised materials. A revolver, you can use a Whitworth barrel to get around the rifling problem, and you can make the chamber walls and barrel much thicker.

But where do you get the ammo?

And a harmonica gun would be another way to get around some of the limits.



I think you mean MPa, not gPa. Don't know what gPa is but it wouldn't be much.


It will be particularly interesting when these devices are starting to emerge, but still only offer marginal "print quality" compared to traditional products, but as such low cost that they really start to permeate. how many people will print themselves a "low resolution" but open source running shoe rather than going to foot locker and shelling out 100 dollars for shoes with springs built into the heel. We may temporarily end up with a generation of the worst dressed kids of all time. Maybe it will become fashionable.


The big problem is that we won't have enough "random junk biomass". As Adam rice points out, the technology would have rather bad results if used in the wrong way. A worst case scenario I can envisage is that oil prices keep climbing, and this technology becomes widespread just as a large number of countries/ societies realise they desperately need some oil like stuff, so ruin their local biosphere in order to stay alive another few years.

As for guns, you can have single shot ones firing black powder right now. So all you need to do is stick a bunch of tubes together and hey presto, a multiple fire gun.
What I'd like would be a replica of one from the late medieval period, I believe is tas 3 or 4 barrels, with holes drilled such that you load each barrel with multiple charges, aligned with the holes. Then you ignites one, it fires through the hole, and ignites the next one. So you get a barrage of shot over a few seconds.
In fact given what I know right now I could probably be arrested for thought crime, and I don't even know quite as much as I can learn. Plus I'm planning on making a Greek Fire engine for next year, got to cement my mad scientist reputation.


The FMG9 add is interesting insofar as it suggests a level of paranoia waaaayyy beyond that which I consider acceptable or necessary or desirable. Good thing I live in the UK.

One of my hobbies is medieval re-enacting, and I'm spending about 90 to have a seamstress make a knee length gown (1460's or so) in nice blue wool. It is made to measure, although is a comparatively trivial piece of work compared to my arming doublet, which is being made up after measurements have been taken and actual cloth patterns made of my body and torso.

I would imagine that this sort of technology will lead to some interesting control measures. What I have been finding out, whilst trying to re-create some medieval technologies in my back garden, and generally fiddle about with fun stuff. (Whats the point of being an adult in your own flat if you can't do things that you thought would be cool when you were a child?)

Basically, getting the correct material can be rather hard. Finding a catalogue that will sell you tin, or sulphur, or whatever, is difficult. I suppose if you wanted to make your own loudspeakers, you could rapid prototype the actual body and frame inside, but the speaker itself would have to come from a catalogue.

So what we might see is gvt and others attempting to control the technology by limiting access to the necessary parts and materials, which would work for a large chunck of the populace, and the growth of many little owners groups and clubs, with their own online forums where they pass tips and hints and designs. It'll be a bit like model railway enthusiasts for example. Certainly people will be trying to hack their prototyper.

Then what will happen is that some teenagers will prototype themselves something like the FMG9 and get shot by the police, and there will be the usual outcry. Add that to the copyright and patent laws, and you'll find that all these machines will be DRM'd to heck.


Anatoly @31: Of course rapid prototyping is going to cost more than mass production. That's not the point. The point is, it gives you a level of control and/or customization over the product that isn't available from the mass market suppliers; it fills the opposite side of the balance from economies-of-scale with tailored-to-meet-exact-needs.

Dave @34: you build DRM into the Fabs. Right from the start. Too late, they're already out there. (One of my mates is working on a startup with one right now.) As for your take on guns, you're thinking of sophisticated stuff; I'm thinking about Saturday night specials using unconventional components and materials, manufactured as a single lump, designed to be discarded after use (and ideally biodegradable -- to remove the evidence).

Ed @37: Typo: should be MPa.


So, instead of scaling up the fuel-producing vats, how about we scale down? Sure would be nice if the contents of my home septic tank could be routed to a bio-fuel fermenter. The only danger would be accidentally injesting the modified E. coli. If they established themselves among your gut flora, you'd end up farting diesel.


Guthrie @ 40: I live in the US, and also consider the FMG9 to be over the top paranoia...

As for police shooting people with home made guns, the more worrying thing would be people manufacturing guns that didn't look like guns. For the most part that's illegal now, but imagine people making guns that looked like ipods or packs of gum in their machines.

instead of scaling up the fuel-producing vats, how about we scale down? Sure would be nice if the contents of my home septic tank could be routed to a bio-fuel fermenter.

well, as it turns out, here in new zealand a company seems to have cracked just that.

these guys install these scalable filters in sewerage ponds for the natural algae (that the sewerage firms normally pay someone to clean away). the algae is high in lipid (?) oils, which makes biodiesel. the advertising says they could supply the country's entire diesel needs off the current supply of poo.

why spend money to harvest your biomass when humans are producing thousands of litres every day?


Andrew- I should have made it more clear- I was thinking of teenagers making gun look a likes similar to those in the film of the day, or TV series everyone watches, or suchlike. Add a cap firing mechanism or something which makes gun like noises, and some people will get confused.

I have a giles annual from the 1950's which is based on the idea of the USSR producing small spy guns hidden in packets of cigarettes...

I think it would be very easy to make such a hidden gun using home made explosives and a carefully manufactured cellulose tube, for one off use. And then if you use the right kind of ammunition...


Perhaps it's time to consider the "proposed spam solution" checklist to replicator technologies.

I'm still particularly fond of "( ) Armies of worm riddled broadband-connected Windows boxes". It would be fun (for some definition of "fun") to let loose something content-modifying like the Wazzu virus (sadly, google isn't finding as many documents infected these days). Maybe it prints IR-visible 2D barcodes---on parts designed to fall off!

We're pretty sure this future doesn't have an analog of the yellow dots (both scanning and printing)?


Making guns:

It's easy to make a gun. The tech and materials to make one at home can be bought on-line -- just look in any model engineering catalogue or on-line supplier for a small hobby lathe, suitable drills and cutters and some pieces of steel (I get a lot of my raw metal stock for home engineering projects off the streets as litter). There are a few obvious bits that need really good steel (the barrel and bolt) but they are easily sourced too (excuse me if I don't list URLs here) and there are ways around that if you know what you're doing. You don't need a "replicator" unless you want to make several units all the same.

What you'll end up with is primitive (single shot, inaccurate) but capable of delivering stored energy at a distance, just like a "real" gun. Then again you could deliver the same amount of energy to a human being with a good swing with a baseball bat, or even a sword (which can be and often are made in hobbyist workshops).

There are bows and crossbows, darts and lots of other non-chemical projectile weapons which can easily be made in home workshops, but there is no real need to do so since they can be bought in shops or over the Internet without many restrictions.


Math underlies software for the universal auto-tailor. See, for instance:

arXiv:0805.3791 [ps, pdf, other]
Title: Isonemal Prefabrics with Perpendicular Axes of Symmetry
Authors: R.S.D. Thomas
Comments: 38 pages, 17 figures, 3 tables
Subjects: Combinatorics (math.CO)

It has really pretty pictures, convinced me that there is valid mathematical definition of "Satin" and "twill." But it takes a lot of Group Theory to understand, which most Fashion Design faculty seem
(alas) to lack.

Open Source fab may free us from the alien culture so well portrayed in "The Devil Wears Prada."


Building guns: IIRC there are fairly complete instructions for a muzzle loading shotgun in Bevis, the Story of a Boy.


There'll be an awful lot of guys w/ the nickname 'Lefty' if they use fabbers to make guns with as little understanding of the underlying science as the average meth-cooker. And I kind of assume we're talking about a substantially overlapping set.

Also, why aren't we rendering dead people for their oil? Only kidding, unless there's money in it. The reverse Dune analogy makes me chuckle.


If they established themselves among your gut flora, you'd end up farting diesel.

I think that already happened to me back when I was a student - is anyone researching that at Royal Holloway?

Regarding European textiles, exhibit A is Zara, which makes all its clothes in Asturias and the Basque country. They have reliable long-term suppliers there, as the Toyota production system prescribes, and they can turn models around quicker than the ship from China gets there. The mill the Chinese textiles trade association bought is in Wigan, not in Yorkshire.

Mind you, I was shopping for a new pair of trainers the other day, and I would happily have paid more to specify the details I couldn't find because of some knobber's second-guessing of what ought to be in fashion. (Khaki suede Adidas Samba with white 3 stripe can't be impossible, right? I used to have a pair.)

Regarding firearms, quite advanced ones are already made by artisanal workshops without the benefit of reprapping, or indeed computers. The Viet Minh copied Thompson guns and M-1 carbines, although they replaced them with captured genuine stock as fast as they could bully the peasant militia into handing theirs over and reporting them as lost to the Americans.

Then, of course, they swapped them as fast as possible for Soviet kit. And Afghan/Pakistani gunsmiths are quite well known for copying AKs and Soviet machine guns with their own tools. Imagine what they could do with quick lightweight fabrication and....electronics. In fact, the "Afghan lathe" was a concept the RepRap team thought of using early on - a drill aimed at an angle to the work, the work being turned on a shaft. Vik Olliver, I think, had seen this in the NWFP.

Repraps, so far, could lay down the wiring, and the logic could be done with the guts of something else. Iraqi insurgents seem pretty good at firmware hacking.


"Add that to the copyright and patent laws, and you'll find that all these machines will be DRM'd to heck."

How exactly would the DRM-ing prevent instantiation of off-the-internet (with automatic design obfuscation to confound contraband recognizers) or custom designs? Is the idea that access to generic fabricators would be severely restricted, with access to replicators of approved designs more generally available? This is the common SF solution, but would it actually work? The continuing arms races we've seen with computer viruses and spam suggest that an absolute technical solution is unlikely.


Got a little ahead of myself in the latter part of 35. Scale isn't what this is good for (yet). Especially since fabs aren't completely bootstrappable.

I'm sure a number of Open Schematics repositories will pop up. Electronic gadget re-skinning and re-mixing; housewares; Green DIY smart-car projects with lots of fabbable components; the hobbyist robotics and RC airplane scene, ahem, taking off (you called that one in "Accelerando").

There will suddenly be a demand for 3d models of objects that already exist, so there will be a need for affordable 3d imagers (or at least good software that can take photos of an object from multiple angles and stitch them together into a hi-res vertex map). And the legal issues behind the resultant schematics? Ye gods...

Actually, the consumer market for all kinds of unfabbable components will skyrocket. Motors, servos, FPGAs, LEDs, switches, etc.

I imagine assembly-centers will spring up, similar to copy centers now, that have the fabs and imagers and many of the unfabbable components and knowledgeable staff (open 24 hours, free coffee), and so on. Hell, maybe Radio Shack has legs yet.


The FMG9 is cute, as others have said, but if you want an SMG that's easy to make nothing beats the Sten.


Can be made with pressed metal and spot welding in five hours. Or maybe quicker with cellulose and super glue. Ammunition remains the problem however. I can see a few problems with trying to "print" any kind of volatile material.

Future Watch 1: Darwin Award for first attempt at printing nitroglycerine.

Future Watch 2: Super glue banned by authorities "for its obvious terrorist applications"

Wandering off guns and humour and back onto SF ground: Harry Harrison did a nice time travel tale (A Rebel in Time) based on a Confederate stealing a Sten and its blueprints and going back to change the outcome of the Civil War, pursued by an African-American Yankee, natch. Written a full decade before Turtledove made the subject matter his own. Anyone else remember it?


AH @54:
I'm sure the authorities will look benignly on as home fertilizer stills for high grade nitrates spring up. I mean, what's the worst a guy could do with ammonium nitrate and (bio)diesel? Much less TNT or RDX.

And isn't super glue already banned by the authorities as part of the eternal War on Drugs?


- more textile industry rant -
The global industrial process was enabled from the late part of the 18th century in North East England. Cheap cotton from the southern states. (The cotton gin, eliminated the problem of pesky slaves not working hard.) Water power, the spinning mule and 'automatic' weaving machines (6 -12 machines:1 person) in one building. A captive market in Europe and India (which had its native textile tradition destroyed by the British.)
The British entered the market as pile them high, sell them cheap, but then they had an expanding population to clothe. They lost the edge to Germany etc at the end of the nineteenth century because they wern't hungry enough and failed to invest in new plant. (That's the short answer.) When the industry collapsed in the 1959 lots of plant was shipped out to the far east, I think.

Before this time a significant fraction of the population spent a significant fraction of their working lives making enough fabric for their own use.
-/ more textile rant -

As for Fashion… Traditional it was either what Grandmother would abhor or the last generations sporting togs. Consider that most of us wander around in what a generation ago considered underwear and work clothes that aphorism works. It's all about the yoof.
Today there are people paid to predict what people will be wearing in two years so material can be designed, woven and made up. (That's why Charlie's jacket was so last year.) Cunningly what ever is predicted is made, but not necessarily bought.


tp1024 @23, almost all my clothes are made in Cotati, CA, and my underclothes are all made in Seattle. Yes, I pay more. I wish I was still able to sew, since it's difficult to find nice clothes in large sizes.

Graham @33, LOL But what's fifteen to one?


You can already make yourself a lethal backyard weapon with bits and pieces from a hardware shop. Go read Ken Macleod's recent post about Roman catapults. Then I think "tension catapult" is what you need to ask Google about. Even that nasty yellow clothesline rope is stronger than the hemp the Romans had. If you splashed out on 10 metres of Kevlar from a chandlery, you'd have the stored-energy equivalent of a sniper rifle.

These things are controlled by social factors, not by technological limitations. Someone slightly organised could get something much more potent than stones and Molotov cocktails into the average angry mob, it just hasn't happened yet.

Re: Biomass to oil. Our civilisation needs high-energy nitrogen fertilisers MORE than it needs 747s and airconditioners. Would anyone care to do a back-of-the envelope calculation of the energetics of running the Haber process to grow enough crops to make all that "waster" biomass? Although the sewage angle is interesting...


Well, David Axe did report this last April:

You’re an Australian peacekeeper in East Timor. It’s 100 degrees out. Any clothes at all are too many, and here you are wearing a 25-pound armor vest. Why not go without?

I’ll give you one reason: sharpened rebar arrows. That’s right, improvised projectiles made of steels rods that are meant for use in construction. In East Timor, the local riffraff fashion bows out of heavy-duty rubber bands and shoot these rebar arrows with enough force to punch through your skull. A couple weeks back, thugs from rival political parties squared off in front of the Australian embassy in Dili. One side had a gun. The other had these bows and arrows. The gunman wound up with a length of rebar in his head. [...]


@ 56 And we haven't even mentioned the fibres the textiles are made out. Technical fabrics have come along way from the 100% Nylon drip dry days. You have to think very carefully about the design of the extrusion to make something bearable to wear, mimicking what's nice about natural fibres.
In these days of global-flimflam it's impossible to go back to 100% natural fibres, unless you radically cut demand. Not enough space in the world (even without the competition from bio-fuels) and cotton for a start is one of the most un-green crops in the world (spot the Aral Sea anyone?). Don't burn bacteria poo spin it!
Hemp might be okay - difficult to make smooth shiny white sheets though and you have to build a processing industry from scratch. (Hint the fibres are as tall as the plant, -- macro-nanopaper.)


maggie@60: Most of my clothes that are prone to wearing out (and therefore the ones I buy most often) are at least 50% cotton, and I don't think I'm unusual in that regard. The clothes I have that more mostly synthetic fibres are things like rain jackets and warm tops, which last for years.

Others: I dunno about the 1950s sizings. I'm taller than the average man of 1955, and I generally have to buy the smallest pair of trousers on the rack...


Chris L @ 61 It's really easy to forget all the other textile applications other than the clothes and bed linen we see every day. Once upon a time all thin flappy stuff was made of nature fibres: Sails, awnings, tents, belting, bags, upholstery, ribbon, nets, sieves and filters and an infinity variety of string. Without mentioning cunning things you get now, like knitted metal catalysts, braided cores for carbon composites.

General Comment -- Use life: higher quality/more expensive fabric -> more expensive clothes which last longer. Poly cotton mixes may contain low quality cotton. How often do you use your rain jacket / warm tops? How worn out are your t-shirts before you reuse/recycle?

I guess you don't eat many doughnuts if you have a small waist size. Leg length is easier to compensate for than the complex geometry of middle-age spread, (she says politely) which changes shape depending on whether you are standing up or not. Also biology means that profile doesn't scale as you go up a dress/pant size.


I'd be more impressed if the demo guy for the FMG-9 didn't look like a cross between a mall ninja and IT geek.

OK, as to fabbing, this could get interesting. High density energy storage (batteries, fuel and explosives) are going to be the big issue here because they would be dangerous to fabricate. And expensive.

Still, the idea of a batch of terrorists assembling a bunch of cheap automatic weapons (and scoring the ammo somewhere) is frightening. Easier would be some sort of flamethrower for a terror weapon because you can buy gasoline (or its future equivalent) easily enough.

For less lethal options, I could see the paper model hobby leaping on this in a big way and it putting a serious, serious hurt on the gaming industry that makes miniatures.


It's pretty easy to build a device to deliver violence. That's not that interesting. What's important is the ergonomics of the device.

Consider the phenomenon of impulse downloading....


#64 -- that looks like Herman Kahn and Anthony J. Wiener's The Year 2000, Table III, no.9:


Some Possible Causes of "Surprising" Changes in the Old Nations

1. Invasion and war

2. Civil strife and revolution

3. Famine

4. Pestilence

5. Despotism (persecution)

6. Natural disaster

7. Depression or economic stagnation

8. Development of "inexpensive" doomsday or near-dooms day machines

9. Development of nuclear "six-gun" weapons technology

10. Resurgence of Communism, or revival of Fascism

11. A racial, North-South, rich-poor, East-West, or other disruptive polarization

12. Economically dynamic China (~ 10 per cent per year growth)

13. Politically dynamic U.S., U.S.S.R., Japan, West Germany, Brazil, and other powers

14. New religious philosophies and/or other mass movements

15. Development of UN. or other worldwide organizations

16. Possible regional or other multinational organizations

17. Psychologically upsetting impact of new techniques, ideas, philosophies, and the like


The injection mold plastic toy industry will be in dire straights. Archives of schematics of, say, the entire run of original Star Wars action figures could be traded on p2p networks. Transformers. G.I. Joe. Voltron etc. Nostalgia extrusion.

With touchscreen lcd's, micro form factor motherboards, keypads, card readers, etc. available on the cheap from southeast Asia there could be an explosion of bespoke hardware hackery and gadgetry enclosures -- much of it completely assembleable at your local FedEd/Kinko/FabCos. Maybe the wearable computing revolution is driven by hobbyists? After all, that shit's gonna need to fit. Custom fabbed eye-glass frames.

And you'll be able to fix a lot of devices that used to require needlessly propriety parts/accouterments. Broken vacuum cleaner? Download the schematics to all the fabbable parts of your Hoover SuckMaster 9000. Snapped a lever off your expensive espresso maker? etc. Other things: Hygiene devices, lawn/garden, bed/bath, kitchen. "Just six easy installments of $333.33! And that's not all! You also get five -- count 'em, five! -- imposing IP goons beating down your door absolutely free!"


We've had electrochemical machining for a long time, for those that wanted to print their own machine guns. You plug in the right alloy steel and it is eroded to shape in minutes. Alloys have special characteristics so you better know what you are doing.


Have a gear-fetish that can only be sated by a wee desktop difference engine?
Need a Rubic's Cube now?
And, (oh man), have the hinges on your laptop gone to shit?
Stepped on your expensive headphones and snapped the frame? (done it a dozen times)
Custom Keycaps for your Model M?
Custom chassis for your Model M?
Have a clever solution to your gnarled power-strip/wall-wart/wiring underbrush?

I guess I'm thinking small here, but this stuff is fun. Plus, you GPL your schematics and the problems never need to be solved again. Plus plus, others can debug them.

Christ, and I'm supposed to be studying right now, too.


Speaking as someone who has done a bit sewing(costuming, children's theater), I'm finding it a little hard to believe these machines can 'make clothes'. Printing the pieces rather than cutting them out from a pattern, sure. But an awful lot of making clothes is in the stitching. I don't think these printers can manage that yet, and to glibly answer that sections could be glued together . . . well, I'm just a little skeptical.

And I think this is going to be the big problem with these fab schemes. No matter how you slice it, unless these machines are awfully clever, there's still going to have to be skilled assembly work. The hype at this point strikes me as similar to the notion that 'sewing machines will replace tailors.' They didn't, of course. But they did put more ambitious projects within the reach of amateurs.


Wait a minute ... desktop analytical engine was your idea. "Singularity Sky". Heh.


If reprap-like machines with strong materials turn out to be cheap and easy, then never mind licensing handguns — we're going to have a problem with home-made crew-served weapons.

And missiles. How much harder would it have been to invade and occupy Iraq if every backyard workshop in Baghdad could make surface-to-air missiles? Now that would be a Revolution in Military Affairs.


Ooh, I get to wear my "RepRap Developer Hat"!

Bruce Cohen, SpeakerToManagers @ 4. One can take a laser-sintered steel object and dip it in a crucible of liquid bronze. The bronze wicks into the voids between the steel particles, and you get a solid, albeit hetrogynous object. Bathsheba Grossman makes her mathematical sculptures this way.

(You can also do lost wax casting in broze with many thermoplastics. It's a slightly involved process. I'm holding out hope that we'll figure out how to print bronze clay and sinter (fuse) it in a furnace. Google "bronze clay" for more details.)

ScentOfViolets @ 69 is basically correct; we're not going to be printing clothing any time soon.

Regarding the subject of firearms, ... most of the people involved in RepRap aren't really keen on shooting anyone. Robert Sneddon @ 47 is correct that you can make a firearm using hobby engineering tools like a lathe and drill press. For perspective, I'm from the US, and we seem to have an excess of firearms, but they weren't made in home workshops. It may be that RepRaps will be able to make guns someday, although I'd rather not think about that. Perhaps we should make bullets illegal?

“Gun control? We need bullet control! I think every bullet should cost 5,000 dollars. Because if a bullet cost five thousand dollar, we wouldn't have any innocent bystander .? - Chris Rock

Regarding 3D printing and paper, the process is called "Laminated Object Manufacuring" (LOM). A company called Helisys used to sell LOM machines; they were a bit expensive, apparently, but the consumables were cheap. They would glue down down a sheet of paper, score the paper where the desired object isn't using a C02 laser, and then glue down another sheet. The finished product looked a bit like wood. (I'd like to see one someday.)

Regarding RepRap and 'intellectual property', my considered response is "Bleagh".

Feeding RepRaps - unfortunately, the materials we've been working with a fair bit, ABS and HDPE, are petrochemically sourced, and that stuff is getting scarcer. Polylactic acid (PLA) we get from corn, ... which is also petrochemically sourced. I like to daydream about algae that makes PLA or some similar polymer, and am curious to hear how people would go about developing such an algae. Any thoughts?


#27: I'm not sure the ability of people to make guns in their home office would change the law much

You're right, it won't, in the short term. And, in the short term, P2P hasn't changed copyright law. Yet. But it will eventually, because once a law becomes unenforceable, the law will eventually have to keep up with the changing social norms that technology makes possible.


Sebastien @ 72,

When you say 'Regarding RepRap and 'intellectual property', my considered response is "Bleagh"', do you mean that you don't see IP as an important consideration or one that should potentially limit the development of technology such as the RepRap? Because I'd respectfully note that there are vast numbers of people in the design, marketing and production industries who would beg to differ. You might not agree with them, but their concerns can't be armwaved away.

Disclosure: I'm the 'trainee IP barrister' Charlie mentions at comment 11 (I think 'aspiring' would be a better description right now!). The postgraduate dissertation I'm working on right now centres on this; I was amazed to find that a pretty comprehensive search of the legal literature makes almost no mention of the intellectual property implications of personal rapid fabrication. Yet every IP lawyer I've discussed this with has the same reaction:

1) I've never even heard of this.
2) That's going to turn the IP world upside down.

Now, plenty of people round here might say 'and a good thing too' re the second point. But when you turn things upside down without thinking ahead, all the bits tend to fall out; what I'm trying to do is to point that out in advance, not so that we can stop the box being turned over - that's futile and impossible - but so that we can at least work out where the bits are going to fall, and not stand underneath there.


Incidentally, on the guns/ammo front, you don't need ammo for a Saturday night special.

There's a 18th century military technology that was so scarily effective that Napoleon banned it as a weapon of mass destruction; the only reason the Austrian army (who used it) didn't clean his clock was that it required elaborate logistic support back in the day. The technology in question is the air gun.

Today we mostly think of air guns as BB pistols for plinking at targets rather than as serious weapons, but that's really a side-effect of the convenience of cartridge ammunition for military quartermasters. If what you want is a Saturday night special, there's a lot to commend some sort of gas powered weapon: for one thing, there are no powder residues to haunt you, and for another, air is free -- you don't give yourself away by looking for a dealer in (presumably illegal) ammunition or explosives.

Really stupid reprap idea: how about looking for a strain of algae that has lignified cell walls and tends to clump or form biofilms? Then you could spray it in aqueous solution to build up layers, and dry it out to come up with a substance approximating soft wood.


OK - so all these auto-lathes (term coined by John Brunner who wrote a lot about this sort of stuff - the whole thread is straight out of Shockwave Rider) are connected to computers that are connected to the interweb:

Nightmare Scenario: Zombie-net has the world's 'lathes produce ricin or explosives (with pre-set timer).

Probable (I'd say inevitable) Scenario: Age Of The Widget. My kids bring home enough plastic crap now to fill an entire room with plastic pigs, spacemen, pen-caps emblazoned with various corporate logos etc. The biggest use of repraps will be to produce these things even more massively, not always with the owners consent. Face it - this sort of rubbish is the stuff that will be free of copyright restrictions.......


On the military applications, it strikes me that although rolling your own rifle would be cool, nobody who wants war ever seems to be constrained by the availability of small arms, and as pointed out above, common metalworking tools will do it anyway.

RepRap in camouflage sounds more appropriate to things where the shape, design, or electronic content is decisive. Things like shaped charge EFPs, RPG warheads, and basic guided weapons. Oh dear.

Interestingly, if you reprapped a gun, one feature would be that the whole thing would be plastic...which is nice for certain applications.


I believe DARPA is pouring considerable amounts of cash into portable fabs, for long-term logistical support of US military forces in the field. The idea is that it'd be enormously useful to have a mobile field fab that can run up spare parts for vehicles and tents and other pieces of kit in the field. One of the big problems all armies face, when operating away from their home territory, is maintaining a supply chain that can provide replacement parts for things that break; if you've got fabs in the field with your troops, all you need to ship is raw feedstock plus design schematics, rather than needing a fulfillment chain that can deliver fifty million different orderable SKUs into a theatre.


At the RSC Summerscience exhibition last year, some researchers from University of London Queen Mary were showing off an acoustically-driven powder printer - sand paintings by music. It gets more complicated, but the fun part is that you can control the composition gradient throughout, and print with dry powder to a ten-microgram resolution.

Dr Shoufeng Yang's research.

So instead of spraying in solution, you can always print in a powder & matrix format, and then sinter. (Anything that comes in a relatively uniform fine powder.)

The droplet size is too large to build up proper functional films, but what you can do is get uniform small-grain solids - or alternate layers of two reagents and then get a shaped object of a third substance you couldn't normally work.


Simon @ 74

Two examples of the wheels falling off as a result of cheap replication are:

George O. Smith: the Venus Equilateral story in which their "matter transmitter" gets reworked into a matter _duplicator_.

Murray Leinster: The Duplicators (what happens when the tech breaks down and people have forgotten how to fix it).




AH @ 54 wrote:

> ...Confederate stealing a Sten and its blueprints
> and going back to change the outcome of the Civil War
> Anyone else remember it?

Yes, I have the hardcover somewhere. It's an interesting idea but the materials technology of the time was not nearly advanced enough to make it work. About the best you could hope for from a blowback submachine gun fed with blackpowder ammunition is a couple of shots. After that, the fouling is going to prevent the cartridge from going far enough into the chamber for support, and it will blow the end off, leaving you with a jam that will require disassembly to get the remains of the case out.

The Gatling Gun worked because the action was operated by muscle power and a dud cartridge would be ejected anyway. The Martini Henry rifle suffered extraction problems as fouling built up, to the point that the extractor could rip the head off the cartridge.

The Sten requires precision ammunition - it's got to have repeatable chamber pressures in order to cycle the action - jacketed bullets to prevent the barrel fouling with lead deposits, and smokeless powder that was not invented until some considerable time later.



Something to bear in mind: free software to maked 3D computer models is already out there. (For instance: Wings3D)

The software to produce pictures from these models is cheap. Example Image

You can feed that image into a fabricator and get a 3D object. There was an article about this in 3D World magazine, sometime in the last year or so. But that model wouldn't work. There are things about it which are fine for a picture, and a total disaster for a physical structure. Parts aren't combined into a single mesh. Some things are essentially cups or tubes with zero thickness walls.

Converting things is not easy. You really need to design from the start for fabrication.

Simon, if you haven't checked out the low-end CGI model market, it's worth a look. www.renderosity.com for instance. What you'll find are a lot of people making models, selling them, or giving them away. And there are trademark/design/copyright issues.

One of them is about control. You're selling the model so that the purchaser can make images, which they in turn can sell. But you don't want them to sell your model, or the surface textures. You're going to need something more subtle than Creative Commons--it looks to me as though there is a division between different kinds of Derivative Work.

Renderosity has a forum on Copyright issues, though uit has the usual US bias. Free registration required, and, in my experience, a pretty safe site. They don't spam folks.


Simon@74: I think of small extrudable macroscopic objects and mostly what comes to mind is "housewares" and "toys" (I'm not very interested in weaponry). While I know that includes some big corporations ready to file lawsuits, somehow it doesn't seem to be the big sexy IP issue you and Charlie are hinting at. Especially since electronics can't be fabbed. Where do you see the most friction between end-users with fabs/schematics and corporations with IP to protect?


#54 & #55

Ammonium Nitrate CO
BINED with the right materials (Aluminium Powder, Toluene, etc.) is a VERY effective miltary explosive, though out-of-date now, I still wouldn't want to be standing anywhere near it when it went off!
Look up: Ammonal, Amatol .....


Simon Bradshaw @74
When you say 'Regarding RepRap and 'intellectual property', my considered response is "Bleagh"', do you mean that you don't see IP as an important consideration or one that should potentially limit the development of technology such as the RepRap?

The latter. It's a can of worms. We've been in touch with some folk from the EFF and the FSF, to ask their advice.


Charlie@78: I remember seeing an article referenced by slashdot on topic almost a decade ago. I wonder if its seen the light of day yet...


The technology in question is the air gun.

as a paintballer, I'll point out that the technology for air-driven weaponry is already fairly advanced. carbon fiber HPA tanks, basically miniature scuba tanks, hold air at 4500 PSI. You can buy them for a hundred. Not sure if you could fab em with this paper mache thingy. Might be a bit heavy.

paintball HPA tanks are a bit large for a pocket weapon, but they're designed for holding enough air for a couple games of paintball, which might consume a thousand shots of paint. If you just need a few shots of lead, then you could probably make the tanks smaller.

Normally, paintball guns (some prefer to call them "markers" to be less scary to the nonplayers) regulate the pressure down to 800 psi or so. But no reason you couldn't build an entire paper mache gun to operate at a couple thousand PSI.

Not sure how much pressure it would take to move a lead bullet down a rifled barrel. Or maybe you don't need rifling. A lot of crime can be had at ranges under ten feet.

Oh, and plenty of paintball guns today are semiautomatic and completely mechanical. They use the same air that moves the paint down the barrel to recock the marker and load another paintball. springs wouldn't be too hard to figure out.

The main thing would be they would be big and clunky and hard to conceal for criminal uses. Not sure how small you could make one that did only a few shots.


Another gunpowder-free gun making the rounds in an online demo accelerates spherical slugs using what's basically a centrifuge, and ejects them out a hole that sends them on a tangent path. The rate of fire on the thing is ridiculous (thousands of rounds/sec).


It doesn't matter how cheap the prototyping machines can be manufactured and maintained. They'll require a government license to own and operate, and it will not be easy to get one. It'll probably be difficult as hell to figure out which arm of the state apparatus is the one that takes applications.


JH@89: It'd be virtually unenforceable, especially since they can replicate most of their own parts.


As far as air-powered weaponry, a plain old blowgun is pretty darn effective, and totally innocuous looking. Used well as a plot device in "The Sign of Four".


Yeah, the thing about replicators is you can NOT depend on DRM'ing them. if EVEN ONE replicator gets hacked to disable DRM, that one replicator will be used to make a million more replicators.

And if the only fuel is electricity and a big stack of old newspapers and scrap wood for fodder, it'll be impossible to track down.

We might not know what the world will look like when cheap replicators are available, but having them under DRM control (or any other technical control) is one option we know won't be there.


jh woodyat: So the government makes you get a license to buy a fabber? Fine, you just buy one from China, or build it yourself using the RepRap online instructions available on Pirate Bay. Is the government going to look in everyone's shed three times a week?


Dave@82: You're going to need something more subtle than Creative Commons--it looks to me as though there is a division between different kinds of Derivative Work.

There is currently no strong copyleft license that I know of that protects specifications of physical objects. GNU-GPL is commonly used to protect ee designs on the opencores website. The problem is that GNU-GPL's copyleft only attaches itself to derivatives of the original work that are also distributed. And a physical device isn't a derivative, and selling it isn't considered distribution by copyright law. So, GNU-GPL won't protect the Open Hardware community nearly as well as it protects the open software community.

I started looking into the various licenses, to the point that I ended up writing a paper about it, "Libre Labyrinth". And the kind of Open Source license you would need to protect physical designs doesn't exist.

The license that is closest to doing the job is the Apple Public Source License. It doesn't require distribution to activate copyleft. It attachs copyleft to all derivatives, even privately made derivatives. APSL was designed when websites started using modified webhosting software, but were not legally required to make those modifcations public because they didn't distribute teh executable/derivative. The derivative remained on the host computer, and users would surf to the site and the code would execute. Since the user only saw the output of the executable, rather than the executable itself, the work wasn't distributed, and the GNU-GPL didn't require that the source code fo the modifications be made public.

Turns out, this exact same license is almost exactly what an Open Hardware or Open Physical Design license would need. It is a bit old, though, so I'd prefer to add some anti-patent clauses and an anti-tivo clause.

There's a discussion over on the Open Hardware Foundation website trying to come up with a good, strong copyleft license for hardware and physical objects. It doesn't contain any IP lawyers on the list though.


I've known a lot of Americans underestimate airguns, despite the example of Lewis and Clark.

In the UK, because of the firearms laws we have been lumbered with since the 1920s, weapons are designed with muzzle-energy in mind. The two main calibres are .22 and .177 and they deliver the same energy. For many decades, spring-pistons were the usual power source. Because of a couple of quirks in the law, the CO2-cartridge-powered gun was illegal in the UK until relatively recently, despite it's wide use in the USA and in international target competition. Britain was the place where compressed air was developed as a power source.

I'm not at all sure about paintball in the UK. I know it happens, but the legal constraints are different. (And I follow The Whiteboard.)

Still, if you can get the Firearms Certificate you can get airguns with much higher muzzle energies. And Colonel Moran's famous airgun was close enough to real.

And on other issues:

I'd expect to see specialised fabricators. A clothing fabricator would be nothing like the machine which can make the decorative wall-plaque from a mega-catalogue of kawaii.

I wouldn't rule out glues and spot-welding in clothing fabricators.

Modern fabrics have more stretch. Look at reenactments, and compare 18th Century breeches to modern pants.

Machine sewing made modern trousers possible. There's a pretty huge amount of sewing in that fly, whether it uses a zipper or buttons.

Modern factories still need skilled labour, but who trains it? You pay to train people, treat them like shit, and they go to work for your competitors. Fabricators look a very tempting solution to that mindset. People have been weasel-wording the definition of slavery for a long time.


In addition to woods, they are being used in small scale jewelery manufacture using composite metal resins - weird stuff where the metal is bonded into a plastic as a putty. You can form it in a 3D printer and then fire it in a kiln and take out a piece of silver. Quite cool technology.

I've also read of specialist bike shops getting into this for components they have trouble sourcing for weird types of bike.

It is going to change a lot of stuff.


Adam @88: hadn't heard of that one, but a few minutes scratching on the BOTE shows there's no showstopper in the concept apart from physical size and electrical power supplies.

With an launch arm 1.25 metres long fitted to a 6000 rpm motor you get a "muzzle" velocity of about 600m/s (ca. 1900fps) at the end of the arm. Ammo can be fed down a hollow tube axial to the motor shaft and then it rolls out along the launch arm to the release unit at the end of the launch arm which times the projectile's launch to aim it. The "muzzle" can be "traversed" instantaneously simply by retiming the release unit so it releases at a different point in the rotation (elevating a gun like this would be a pain given the gyroscopic effects though). Thinking a bit more it's really a lever catapult where the lever is rotating continuously. With a bit more work on the feed system it's not impossible to launch non-spherical payloads using something like this; DU or tungsten penetrators for CIWS-type operations, or cannonshells or fragmentable buckshot containers for AP work.

Neat idea. Now what have I got in the way of decent-sized motors kicking around in the workshop? I can probably pick up a bucket of ball bearings from somewhere, and...


Am I the only one here who knows of CandyFab?
They're doing 'printing' with a heat gun and sugar. ('The Revolution will be caramelized!')

The other thing is, sewing at home is customizable but not necessarily fast or easy. It's actually skilled labor: you not only have to know how to use a sewing machine, but also how to do the necessary alterations (the easiest is length, and it can get, well, really interesting depending on shape). (To get some idea of what can be done at home, with time and skill, read the instructions for one like this or this.)


Some comments regarding firearms;

Typical rifle pressures are in the neighborhood for 50,000 psi, which is a considerable step up for 4.5kpsi that paintguns run as a source. Even handguns use pressures in the 15kpsi range.

Typical rifle velocities are on the order of 1000m/sec with a 4 gram projectile.

You could fab bullets and cartridge cases, but you still need a source for nitrocellulose gunpowder and the contact explosive that's used to ignite gunpowder. That's going to be hard to due without some chemical alchemy.

Living in a part of the world where getting a permit for carrying a concealed handgun is relatively (1/2 day class + background check) easy I'm not sure that the ability to fab weapons will result in the end of the world. Would be criminals have to wonder if their mugging victim has a weapon.

It's also arguable that it's much harder for a despotic government to run roughshod over the populace when the populace is armed with the same types of weapons as the police/armed forces.



briansmithwins- did you just mention chemical alchemy? As I keep mentioning, I'm experimenting with medieval technology and alchemy, and given time, I could make just about everything needed to umm, have some fun.

Dave Bell #94- what on earth are you on about regarding stretch in fabrics? My re-enactment woollen hose stretch very nicely- if they didn't, they would rip, and we know that medieval cloth also stretched very nicely.

Russell Dovey #92- they'll probably monitor your electricity bill. More meters are being set up now to monitor remotely, and I expect a proper working fab would be quite a power drain. Of course they'll be ways around it, but in the surveillance society they are setting up, more options will be closed off.


Totally on a tangent to this discussion, but may take the "what can you do when you have home ___" discussion in a different direction:

I got a glossy newsletter a couple weeks ago from the high school my daughter graduated from. It mentioned that this year the AP Biology class gene-spliced the jellyfish fluorescence gene into E. coli, as one of their class projects.

Think about where that home technology may take you.

To get you started off, why would an enterprising young person want to waste time making bacteria to produce fuel oil which sells at US $100/barrel, when they could be making bacteria which eat random biomass and produce various compounds which sell for US $100/gram?

If you don't care for the blatantly illegal trade, how about home labs for the hideously expensive prescription mood stabilizers and such that a lot of people now depend on? I found out recently that a typical prescription of Lamictal costs over $700 US a month. Concerta (fancy time-release methyl phenidate) costs about $250 US per month. etc.


One of the things that's made open source software so successful is the definition of open standard APIs. That allows you to take software components made by two different sources and connect them together, and expect them to work (well, usually). I expect that a similar thing will happen with open source objects: there'd be a bunch of standard ways to connect things together. A mass manufacturer could keep a design closed (make the key components non-standard) and let all the repairable parts be open-source, so they could scuttle their service businesses, which are not as profitable as they once were. And they could have the equivalent of extended warranties by providing a complete set of build files for the repairable components for a "nominal" additional fee.

Shoes, ah yes. Fabs are going to give my former employer, Nike, lobster thermidor grade nightmares. They prototype shoe designs on a rapid fab machine now; if their design files ever got out they'd never be able to chase them down. They'd be better off just scrubbing the model and quick designing something very different so the stolen design won't be in style anymore. Nike, and a lot of companies like it are going to be in an interesting bind: their entire business consists of intellectual property, brand, and supply chain. They can streamline their business drastically (and get rid of a huge overhead cost), if they dump the supply chain and email the designs to their retail stores, where they'd be fabbed. But if they do that, the risk of having fabbable designs stolen goes up. They'll desperately want a solution to that problem, because they really want to only spend money building their brand, not in actually making or shipping anything. There are a lot of companies with the same sort of ambitions; I expect a lot of pressure to develop some way to ship fab files to retail stores securely.

But if they succeed in that, the retail business will change almost beyond recognition for everything except clothes, food, and housing (and that will finally get the kind of standardization it should have had half a century ago). A retail store will have contracts with some set of "manufacturers", who would actually only do the R&D and the fab design for a product. What vendors a store would represent would depend entirely on contract negotiation, marketing, and store location. Initially there will probably be retail chains whose stores all handle the same merchandise, but as economies of scale become less important (as more and more parts of a fabber, and of the products it makes, can be made by fabbers), chains will start to fail and/or break apart, and the Mom&Pop fab store will become economically viable.

One thing that'll never happen again. I once spent a week and traveled about a thousand miles around the US Southwest trying to find anyone who had a part for the exhaust system of my Buick Opel (built in Germany). In a possible future I'd go to the nearest shop and wait a few hours while the file was ordered, downloaded, and the part built.


The example I like to give of one of the first areas where personal fabs are going to cause concern for existing manufacturers is the clip-on covers you can buy for mobile phones. These typically embody two forms of IP: the design right for the physical shape of the cover, and the copyright for the design on it.

A phone cover has no particularly demanding mechanical properties; it has no moving parts, and even the graphical design is simple, being either applied to the surface or running through it, depending on the manufacturing technique. In other words, it's likely to be within the capacity of an early-generation personal fab to make. But nonetheless, from what I've seen there's a large and fast-moving market for them.

What next? As fabber technology develops, the range of materials will expand, potentially including metals and ceramics. The resolution of the 3D printing will increase, as will the complexity of structures that can be laid down. There will still be huge limitations - for instance, I think it will be a long time before we can fab weaves and fabrics. (I can just imagine a fabber producing a running shoe, but I actually see the laces as being far harder to make in this way). Nonetheless, many relatively simple products will be reproducible to a very good level of fidelity, and this is where the IP issues arise.

I've already mentioned design rights, and copyright in surface decoration. (And this is a complex area of law - in the UK, at least, there way in which the two interact when relating to the same product can be quite involved.) Beyond that though, there is trade mark law, especially as (in the EU at least) shapes as well as words and logos can constitute trade marks. Then there's the English common-law tort of Passing Off, where there is imitation of unregistered marks or marketing 'get up'.

Oddly enough, patents might be less of a legal problem, as (in the UK at least) you do not infringe a patent by implementing it for personal use or for purposes of research. However, this might make patents much less easy to enforce for certain sorts of product. A good example might be the baby-feeder cup invented by Mandy Haberman, which was relatively simple but nonetheless deemed original enough to be patented; Haberman had to fight prolonged legal battles against manufacturers who copied her idea on the assumption that a housewife wouldn't sue. How would she have got on against having her patented design downloaded to a million fabs?

It's likely that the personal fab will be the end of the private inventor as a viable economic niche; anything simple enough to invent without vast R&D resources is likely to be simple enough to be produced by personal fabs, with patents only meaningfully protecting inventions and techniques that require advanced technology to implement. Of course, if amateur inventors are all happy to be rewarded purely in kudos then this is fine - but I doubt that all will. This doesn't mean that personal fabs are a Bad Thing, but it's important to recognise that they will be disruptive technology for a whole range of people, not just big business.


The Pandora's box of easy lethal weapons was opened long, long ago.

First, don't worry about the folding submachinegun; various folding SMGs of about that size have been around since the late 1940s, and the type that folds into a cryptic box like one that date back to Eugene Stoner's Ares in the early 1980s (typically disguised as one of those old bulky radiophones, and intended for executives worried about being kidnapped). There are also the criminally produced multi-shot derringers disguised as cellphones (that's why you have to turn the devices on for airport security), shotguns disguised as flashlights, and all manner of readily built, lethal crypto-weapons. Good ammunition is harder to synthesize, but good old black powder still works just fine, especially in non-repeating, low pressure designs.

If you're not concerned about crypto-weapons, but just want to have a good old mugging, assassination, or revolt, things are even easier. Rifling is hard to do, but as long as you're restricting yourself to low pressure, short range black powder designs anyway, the shotgun is the way to go. British research in the 1950s showed that the shotgun was significantly more effective in close quarters than an assault rifle or a submachinegun, in terms of hitting your target in a combat environment. Before you dismiss home-built guns as ineffective in modern warfare, look up the paliuntod shotguns. These were basically giant zip guns made out of plumbing fixtures, but they were a mainstay of the Philippines resistance during the Japanese occupation.

I always thought Stephenson missed the target, as it were, in Cryptonomicon, when he picked a submachinegun to use in the HEAP. It requires too much special preequipment--a precision lathe, the ability to precisely fold and weld sheet metal, and access to or the ability to make good springs. In addition, as any shooter will tell you, any self-loading firearm requires high quality ammunition to be reliable. On the other hand, if you combine the multi-barrel shotgun construction of the Hillberg insurgency weapons with the stacked charge, electric ignition concept of the Metal Storm guns (externally ignited stacked charge guns date back to flintlocks, if not further, and so will work with black powder) and you've got a high volume, close range weapon that can be built of hardware store parts and 15th century chemicals.


guthrie@99: they'll probably monitor your electricity bill.

First, I'll fab a windmill and solar collector that uses a stirling engine. Then I'll keep the fabber off the grid. and hey, windmills are in these days, and solar is hot.


Daveon @95, I've never heard of the resin/silver being used in a 3D printer. I have friends who sculpt it by hand.


Scot, I think the anti-terrorism police will be breaking your door down soon.
I hadnt thought of electrical ignition, makes sense, I was still thinking about easily detonated chemicals.

Simon Bradshaw #102- is the individual inventor currently a viable economic niche? I've heard of one or two examples, but the number of people actually making a go of it seems minimal.

Clifton #100 is right when it comes to drugs- it would be great to be able to synthesise them at home. However, even with modern synthesis methods it would be rather hard to manufacture a lot of drugs, especially the more complex ones, and quality control would be non-existent, unless the whole package contains a GC and computer to interpret it all. I don't see home synthesis of pharamceuticals really working except for some of the simple stuff like painkillers. Unless we reach the stage of all having our own pet AI's, but I'm sure they'll be obliged to rat on us should we try anything illegal.


Hey, not only does electrical ignition simplify the chemistry, it also simplifies the firing mechanism in a multibarrel gun; you aren't limited to a bulky circular arrangement, like a rotary firing pin needs; you just have a rotary switch that advances one increment per press, and wire the contacts wherever you need them. That's a lot easier to fab than the firing mechanism of Hillberg's guns (he also designed the COP 4 shot derringer--seen in Bad Boys and Blade Runner).

And if I'm just too paranoid, it's all the government's fault. Alas, Babylon was on one of the reading lists when I was in middle school (not in a class I took, I just found a copy someone had lost and read it) and that turned on the paranoid section of my brain, and it hasn't gone away yet. Nor is it likely too; after all, in the paranoid state of mind, you figure it's safer to be a bit too paranoid than not paranoid enough...


BSW@98: You are correct about firearms having higher chamber pressures than air-reservoir weapons but their bullet acceleration profiles are very different. The high pressures from nitropropellant combustion result in very high initial accelerations but also speed up the combustion process. As the bullet moves up the barrel the pressure drops off very quickly as the total chamber volume increases and the gases cool. After a millisecond or so most of the propellant has combusted meaning hardly any more gas generation hence reduced acceleration.

A given round (say, a 240g 44 Magnum bullet) fired from a 6" barrel pistol will have a muzzle velocity of perhaps 1200 fps. The same round fired in a 24" barrel carbine will have a muzzle velocity of about 1800 fps, a 50% increase in speed from a barrel which is 400% longer.

An air rifle has a much lower chamber pressure as the bullet starts moving up the barrel but the gas regulator continues to feed high-pressure air into the chamber from the reservoir, keeping the bullet accelerating. It is entirely feasible to build an air rifle that can match the carbine ballistics I described above -- a 240g 44-calibre bullet with a muzzle velocity of about 1800fps from a 24" long barrel (it would require an extreme-pressure reservoir and robust valve gear). The chamber would not need to be built to resist the sorts of instantaneous pressures a cartridge firearm experiences. What is very difficult, and probably impossible, is to make a 6" barrelled air pistol capable of 1200fps muzzle velocity.


Robert @ 108

Not to get into a firearm tech pissing match:

Selection of powder burning rates is critical to avoid exactly the problem you use as an example. For instance, loading a rifle cartridge with pistol powder (designed to burn completely in a 6" barrel) will result in a burst rifle. The pressure rise will just be too great, too fast. Rifle powders are designed to provide a fairly continuous 'push' to the bullet. For example, the pressure along the barrel of a M16 (graph about midway down http://www.frfrogspad.com/intballi.htm ) shows how the pressure slowly decreases as the powder burns.

I suspect the .44 Magnum you use as an example was optimized for a pistol, not to take advantage of a carbine's longer barrel.

I'm thinking the only way to get firearm type velocities out of a system is to work at typical firearm pressures. Given a fixed mass and velocity of projectile the work involved is the same.

One of the big advantages of firearms is that they are self contained. The entire energy supply is in the cartridge. Cartridges can be stored for decades without losing that energy. Cartridges are also flexible in application: A firearm will work with one cartridge or with a full magazine of them. I don't need a external source of energy such as an air compressor that will fill a air bottle to 4.5kpsi.



My assumption there was that you don't directly synthesize the drugs, which I assume to be as prohibitively complex as you say, you buy a blackmarket gene blueprint and stick it into a bread yeast, a lactobacillus, or something of the kind. Then you just need a way to filter out or extract the parts you want, or in the crudest case you just estimate the dose per unit of output and eat it directly.

This might or might not be feasible for the modern medical psychoactives - I'm making a wild guess that it is, because living creatures manage to synthesize some incredibly complex and toxic compounds. Most of the traditional illegal drugs, though, already come from plant sources, so you would "just" need to isolate the existing genes that create them, maybe tweak them a bit, and figure out how to patch them into other existing organisms.


This blog really is an interesting place to hang out. How many comment threads combine IP, firearm kinetics, and the structural implications of middle-age spread on garment design?

Bruce@101: As someone whose everyday car is older than I am, and made on the other side of the world to boot, I can tell you that your parts trek is already a thing of the past. Internet shops are where it's at. An hour of searching, get out the credit card and your parts are here within a week. Ebay is even better, if you're a bit lucky.

I think people underestimate the nature of some materials, as maggie pointed out. You can't build carbon fibre up out of sprayable powder at room temperature, you'd end up with something more like a pencil lead. There's a reason that stuff is expensive. Same with lots of fancy textiles. But you can already build things by hand out of carbon-fibre, with a little bit of knowledge and access to the raw material. Plenty of home boat-builders do it.

In fact, I'd stick my neck out and say that internet-based sources of knowledge are the essential synergism for a home fabricator. Look up a design, build up the really fiddly bits (or molds for the really fiddly bits) in your fab, and stick in together. Let's not underestimate the human capacity to learn new skills.


I think the short answer with regard to air-powered weapons is that they can kill a human being at least at some criminally useful range. Police paintball guns that use pepperballs and crank up the velocity, intended to be a nonlethal dispersal weapon, have killed innocent bystanders. They still use a gelatin ball, but its packed with pepper spray. velocities on a paintball field are limited to 290 fps or lower because the clear part of the facemasks are rated only to 300 fps. And I know BB guns with a hand-pumped air chamber can kill small animals.

I think its safe to say that 4500 psi, the right barrel, and a cone-shaped hollowpoint metal bullet, you could reliably kill a human at close combat ranges. At least as reliably as a 22lr, or .38, or similar could.

The reason such a weapon doesn't exist now is because they'd be bigger than a normal firearm, just as hard to acquire as a normal firearm, with no advantages. If you had replicators that were as common as inkjet printers, then weapons made wholly out of paper mache, with no special chemistry for propellants and such, would suddenly have a huge advantage of being as easily available as the replicator.

The question I think Charlie is asking is not firearm mechanics specifically, but whether the bonds that hold mankind together are based on an abundance of social advantages to play nice, or, a scarcity of trouble.

What you say happens when you add an abundance of trouble, in the form of a replicator that can make lethal weapons as easy as you print a bunch of high res color photos, basically reveals how you view society.

I think replicators would increase the level of problems we have to deal with, but I don't think it would throw the world into anarchy.

And my response is really nothing more than my fundamental view of mankind. We've really got very little to go on that would allow us to make a prediction with any sort of statistical accuracy. So we go with our worldview.

Progun people say lots of guns would reduce crime. Well, that's their worldview in the first place, isn't it.

gun control advocates say lots of uncontrolled guns would increase problems, which is nothing more than their already existing worldview.

The thing is, this is really a singularity, a boundary beyond which I don't think we can accurately predict what will happen.

If there has been one constant in our history, it is the scarcity of physical objects. the "zero sum game"-edness of property. For me to gain steel, someone has to lose steel. And steel is something that is hard to manufacture, hard to machine, and hard to work.

We have never seen physical objects operate in our world as a positive sum game. And to be able to make something as strong as steel out of old newspapers and grass clippings is about as close to positive sum as you can get. With a replicator, I could suddenly produce a physical object with the strength of steel, without anyone else losing anything.

Certainly mass is conserved, and all that, but this is really a fundamental shift that we've never seen before. The closest would be creative expressions on physical media being shifted into a wholly electronic media. The way way websites made it possible for stores to sell massive quantities of stuff without a brick and mortar location.

But to actually transition to the point where you don't even need to go to a website to buy the things that you used to buy at a brick and mortar store? You go to a store once to buy your replicator, and then dump your old newspapers into the bin to create everything else you need that's inorganic and chemically inert?

basically, it comes down to what systems in our current society function because the zero-sum-gamedness of physical property is an underlying, and probably unspoken, assumption?


What's the difference between a home fabber and a service or services that you can order specialty items from? As someone remarked up-thread, the idea of a universal machine is pretty ludicrous in an economic sense. Sure, it's possible in theory, but it would be, well, cumbersome might be an understatement. Better to have groups that that are optimized for one class of items. So does the difference boil down to speed of delivery? Degree of customization? What?

Robert S. - A question: to what use do you want to put your gun? Shouldn't you be answering that first before getting into details like powder vs air?


Other random thoughts: Latex/silicone facial prostheses (for white and black hat purposes). Amputees could run off one-off task-specific appendages. Skinnable bathroom fixtures. Disassembleable/portable fabs enable Dadaist camping trips ("I'm bored ... Rock 'em Sock 'em Robots?"). The explosion of kitchen paraphernalia boggles the mind. LARPing props. Modular body armor. Baroque aquariums and hamster habitats. Historic calculator recreations (Antikythera, Astrolabe, Pascaline, Leibniz calculator). Klein bottle shenanigans. Escher renderings. Mathematical art in general (plenty already being done, I'm sure). Toy player pianos. Running off house/car keys becomes trivial. Cellphone Beowulf cluster enclosures. Scale models of entire cities. Airsoft guns. And as much Lego as your heart desires.



Mathematical art in general (plenty already being done, I'm sure)

Indeed. Sebastian Ballard mentioned Bathsheba Grossman; check out the site; there's some really beautiful stuff there. If I had just a little more discretionary income ...


ScentOfViolets @ 113 What's the difference between a home fabber and a service or services that you can order specialty items from

Well, you can run a home fabber for the cost of the plastic, so it's a lot cheaper. For a service, they'll need to pay the rent/shipping, pay the technician, etc. This is not to say that a small-scale coop wouldn't make sense for a club or student group.

With a home fabber you get the instant gratification, and you can run off a couple iterations until you've got the design right. (Hopefully being able to recycle the duds.) It's like having a paper printer at home, as opposed to waiting for the delivery guy or going out to the print shop.

But you still might want to send orders to the folks who do lost wax casting, metal sintering, or ceramics. It depends on the object and on the materials.


The comparison with the dogmas of gun politics is a good one.

There was (is?) a roleplaying game called Cybergeneration (from R Talsorian Games) which depicts a world with fabs. And an oppressive corporate-controlled state. And kids infexted with a virus which gives them strange powers.

I think the designer laid it on a bit thick.

Anyway, if one of the teens in the player-character group could hold down a job in a store at the Mall, and had the skills or super-power needed to hack the Fabricator so that it didn't record the out-of-hours jobs, you were in clover. Your little gang were the source of all that was good.

At least until the cops, or a more traditional adult gang, heard about you.

Organised crime. The War on some Drugs. You know the arguments. A lot of violent crime and theft comes from the illegality of drugs. It creates an artificial scarcity (high prices) which drives the criminal activity.

Now consider what an RIAA-like war on some fabricators would lead to.


BSW @ 109: I chose the .44Mag as an example because it's a round that is in fact used in carbines as well as pistols. It's the same round for both types of weapons -- the carbine I'm thinking about (a Marlin?) is a lever-action used in brush hunting at short(ish) range. It's often fed with a hot load (Federal 240g JHP, frex) but that round will chamber and fire in a suitable pistol safely too (although it will kick a bit).

Pistol barrel lengths vary -- I've seen everything from a 2.5" snubby "bomb on a stick" to 13.5" silhouette revolvers chambered for .44Mag. The muzzle velocities are not proportional to the barrel lengths, showing that most of a bullet's velocity is picked up in the first few centimetres of travel down the barrel. An air rifle bullet's acceleration characteristic is different hence the requirement for a longer barrel to get significant muzzle velocity.

Fabbers are meant to do physics, not chemistry and the ability to fab firearms isn't much help when you also need complex chemistry to produce ammo to feed them with (and yes I do know about handloading -- I would never have been able to afford to feed my own Model 29 without the help of the pixies at Lee, CCI, Hercules and Federal).

That's a point -- could you class a Lee Loader or one of the Green Machines as an early rather specialised form of home fabber? Raw materials in one end, finished loads out the other?


Marilee back at @57, Fifteen to One was a british teatime gameshow featuring 15 contestants who were eventually whittled down to 1 winner. I hadn't realised it had finished; I was going to say that it was exciting and interesting back when it started, was fairly standard when I was at university in the mid 90s, and is now a bit old fashioned.

(Finally, someone here has a question I can not only understand, but actually answer!)



#115 the link to Bathsheba Grossman is broken on non-functioning.

instead ....


I suggest reading "The Diamond Age or, A Young Lady's Illustrated Primer". Granted these printers are a long way from nanoassemblers but they are a step in that direction and we really are getting close to a singularity in that it is very hard to predict what is coming next.


How come no-one has mentioned A.E.van Vogt and "The Weapon Shops" yet?


G. Tingey: perhaps the fact that Van Vogt is barely readable has something to do with it?


116: What, the cost of your time needed to turn out the correct version of your customized item costs nothing :-)

My point is that home fabbers vs a centralized manufacturing location sounds an awful lot like dumb tty terminals vs real home computers. We ended up with the latter option, but I'm not convinced that this was inevitable. Oh, sure, given the right mix of performance and cost, one is probably preferable over the other, but what about the range in between? And home fabbers - the real ones currently or soon available - strike me as being more of a dumb terminal, and less than an intelligent autonomous machine. Why not posit a better distribution system instead?

Something I've been thinking about lately along the lines of getting away from fossil fuels - don't laugh- how about real intertubes? Something that will literally deliver an item to your door with no physical human intervention at all? I know the idea of pneumatic tubes is somewhat out of fashion, but, well, _something_ has got to be better than the UPS van driving up to your door.


Jim@121: What is this ... Diamond Age ... you speak of? A book?


On the general topic of home fabricators and downloadable patterns, it occurs to me that the current state of tech is actually pretty close to this, at least as far as some technologies are concerned. One can buy miniature, high precision CNC lathes and milling machines for around US$2.5k each (http://www.sherline.com/8540pg.htm) and this is top-of-the-line equipment from what I've read. While I've not seen CNC plans for building items with these for download, it's obviously a possibility--if you can program the CNC machine, you can export the programming and send it elsewhere. Look online at the miniature machinists' groups and you'll see all sorts of tiny steam engines and the like out there, made from brass, aluminum, and steel stock; that's precision enough to build nearly any mechanical device with parts small enough to fit the machine. A bit of special equipment and you can print and etch your own circuit boards; that plus programmable logic gate arrays, or PROM chips and microcontrollers give you the ability to fabricate DIY electronics. Motors take a bit more work (need magnets and coils), as do sensors, but much of that sort of thing is available as standardized parts. This is all more labor intensive than having a machine do all the work, but for a sufficiently scarce and desirable item, it could certainly be worth the effort. For that matter, you can do a lot of this even without buying the machines; places like emachineshop.com will make components to spec for you. Granted, they may get suspicious if you send them the blueprints for, say, a complete Sten submachine gun, but would they recognize parts for a centrifuge or incubation chamber used for creating bioweapons?

Once the next round of DMCA copyright clones appear, it probably won't be too long before someone comes up with a hardware/software hack to replace the electronics of whatever next generation encrypted data drive with a version that lets you bypass even the hardware imposed protections, just like DeCSS did for software protections. And, just like DeCSS, once someone posts it on the 'net, anyone with the appropriate fabrication facilities can download the information and (admittedly, with some manual input) fabricate a copy. As others have pointed out, controlled items such as illegal drugs have an artificially high scarcity, and this makes them a high value added product for fabrication. This makes legally restricted items a likely candidate to see the first glimmerings of an emerging post-scarcity economy, as fabrication becomes cheaper and easier than other methods of acquiring the product. This has already happened in many cases, which is why "feedstock" items like pseudoephedrine hydrochloride, nitric acid, fissionable materials, etc. are increasingly controlled. The problem is, there are only so many levels back you can exert such control before the precursors become universal. The cornucopia fabricators in, say, Iron Sunrise, appear to use subatomic particles as the "feedstock", so any form of matter, plus sufficient time and energy, allows you to fabricate anything you need, thus eliminating any effectiveness of controlling precursors. The only route there is "copy protection" in the device (like a "don't make fissionable materials" rule), but since you can fabricate a fabricator, you can, with enough clever workarounds, also fabricate parts to make a fabricator that doesn't include this rule.


Violets@124: home fabbers vs a centralized manufacturing location sounds an awful lot like dumb tty terminals vs real home computers

think inkjet printers. they are the closest we have to something that economically similar to a fabber. and inkjet printers plus digital cameras put polaroid cameras out of business.

A home fabber unit will be similar to an inkjet. Most of the brains will be in your PC, and you'll have some USB connection or similar connecting the brains to the box. And it is likely that fabbers as common as an inkjet printer will put certain existing businesses out of business, just like inkjets did.

Why not posit a better distribution system instead? ... intertubes...

a home fabber IS a better distribution system. Just like buying music online is a better distribution system than buying music in a brick and mortar store on pressed CD media. Better for custumers, at least, more options, easier to browse. Even better for bands, since a band can make some money even if they don't have a massive audience to support a massive record deal. But the internet is BAD for middlemen who act as physical distributers of physical media. Record companies.

DRM and the DMCA are all attempts by the middlemen to remain in the middle in a world where the internet does not require them.

There are equivalent physical middlemen for functional objects that will be knocked out of their distribution monopoly by the creation of cheap, hi res, fabber units in everyone's home. They will, no doubt, be arguing for DRM and other controls when fabber functionality starts encroaching their turf. The question is, can we look into the future from here and now and make some reasonable guesses as to who those middlemen would be?


Look at the lawsuits around Amazon on state sales taxes as an example of old middleman business structures scrabbling for traction in a new world.


127: You might try to make an argument, rather than simply assert something. Why are fabbers more like inkjet printers 'economically'? And why would a home fabber be a better distribution system? I can order my medication online and have it delivered within 24 hours. I can order a precision item like a camera, or a phone, or a computer, and have it delivered in 24 hours. I don't see a home fabber as being able to construct any of those things. So you must mean something different.

And again - I can't make this point strongly enough - _fabbers_are_not_assemblers_!!!!! Heathcliff Huxtable or Tim the Toolman Taylor, and that's most of us, will not be able to much of anything with this machinery. As currently operating, they are more a tool of production, like, say, a computer controlled lathe, and would require someone who was already skilled in the industrial arts to do a lot of what's being described here. They are not sf 'replicators' in any but the most limited sense.


The more I think about it, the more it seems the real advantage is the open source development cycle as it applies to physical objects. You leverage a worldwide community of hardware hackers incrementally improving various projects; releasing early and often; incorporating submitted patches and bugfixes. Secretive R&D departments get blown wide open into worldwide development efforts. Good ideas are immediately incorporated. Branding and marketing costs are nil. Open transportation projects immediately come to mind. Maybe when you reach a major version number you farm mass production out to factories to drive down costs. Maybe not. Depends on relative costs and fab ownership levels. Besides, that would imply some kind financial entity, which, ick.

Apart from transportation (bikes, smartcars, boards, Segways, gliders(?), chevaliers(?)) and weaponry, though, I'm having a tough time thinking of macro hardware that could benefit from this development model. Any ideas? Automated urban farming apparatus, maybe?


inscet_hooves @130

"But apart from the sanitation, the medicine, education, wine, public order, irrigation, roads, the fresh-water system, and public health, what have the Romans ever done for us?"

Transportation IS a major part of the global economy. The whole reason I buy stuff made in China is that is cheaper to ship them the raw materials, have them make it, and ship the finished product back. Cheap transport makes that possible, but cheap labor makes it necessary.

Fabbers eliminate the whole 'making stuff in a factory' need. They also eliminate the need to ship anything except the raw material the fabbers use.



You say this, but I don't think 'making stuff in a factory' means what you think it means. Not if you think these fabbers can do it.


Robotics seems to me to be the macroscopic area with the most headroom. Since the potential for complexity is virtually unlimited (like the linux kernel), we could really leverage a worldwide development community. By keeping the designs modular we can maintain a functional whole while seeing incremental improvements in the subsystems.

This is where the previous posts on the textile industry really fit in -- while we can't fab the textiles, we can fab the robots that fab the textiles and do the tailoring. We can fab the robots that do the farming, collect the garbage, build the housing, etc.

And its open source at every level of abstraction: the schematics for the fab, the CAD software, the schematics for the robotics, the programs for the robots, the patterns for the cloth, etc.

The complexity is so vast in some cases, and the potential for disastrous bugs so great, that we kind of need a worldwide community of eyeballs scrutinizing this stuff.

And after everyone is fed, clothed and housed?
Simple: beer.


insect_hooves @114, and if all those things were done by fab, imagine how much people would pay for the same thing hand-made by an artist?

Neil @119, thanks!

insect_hooves @125, The Diamond Age is a book that has a lot of interesting economic ideas and a really wonderful fabber. However, Stephenson drops half the plot threads partway through, which made me angry.


Marilee@134: Absolutely! Rich people still need to spend their money in unique ways. Liberate the wage slaves and turn them into artists (which they already are =).

Oh, I know the Diamond Age. My facetiousness wasn't very clear (or funny =). Sometimes I wish we had a sarcasm mark, but then again, maybe that's making things too easy. Or maybe that's what smileys are for.

You're so right about how unsatisfying the book is, though. He maintains that that was intentional, but who knows? It's a classic nonetheless.


Even CNC machines generally require a human operator to get the best out of them, to compensate for the varying effects of feed speed and tool wear. Dave Bell's point about designing for fabbing rather than for 3D computer graphics is related. The fact that CAD operator is a skilled specialism is possibly germane, as is maybe the fact that production engineer is still a distinct job.

David Noble's 'Forces of Production' goes into great (if sometimes overly exhaustive) detail about the development of numerically controlled and automated machine tools. He makes the point that attempts to de-skill the human operators and replace them with automated tools were generally failures, because the tacit knowledge embodied in the operators could not easily be formalised and designed into a machine tool.

I suspect that fabs are going to remain in the domain of the specialised hobbyist and the professional for some time to come, as that for most people there will be easier ways to make the parts that they will have the skill to produce on the machines.


It seems to me that the easy ( read really obvious cost savings ) processes are automated first. And we have reached a stage where lots of 'thinking' has to go into the processes, still left un-automated re our recent conversation on AI. Earlier I was tempted to mention the domestic robot in 'Door into Summer' when suggesting that no one yet seems to have deemed it necessary to go into the textile industry and work out just what they do. The protagonist teaches his robot to wash the dishes. Then I realised we don't need the dish washing robot, we have solved the dirty dishes problem another way; with the dish washer which removes all the judgement from the process. These have been around in hotels for decades before they found their way in to domestic houses (Discuss.)
Solutions we are looking for while not under our noses, might be in the next place over.


Violets@129: I can't make this point strongly enough - _fabbers_are_not_assemblers_

But ease-of-assembly isn't currently a priority in the design of objects because mass produced objects are usually assembled in a factory by robots or skilled labor.

If you had a fab in every home, designs for easy-to-assemble things would be much more desirable. Maybe they wouldn't have all the bells and whistles as a factory-robot-built version, but if you could fab that thing at home and assemble it yourself for a tenth of the cost of the factory version, that'll be a huge draw for a lot of people.

A lot of tradeoffs go into the design of a complicated physical object. Many of those tradeoffs could have different answers with the existence of a fabber in every home, and the designs would probably be different.

The thing that comes to mind is automobile design and parts. If fabbers were common, a simple car design might be developed that has few parts, can be fabbed at home, and assembled with a minimum of parts, and is a common, shared, open source design.

And dirt cheap.

Automobile manufacturers don't have any incentive to come up with a universal car design. They tend to sell to one image or another. They tend to want to differentiate themselves from other companies.

Maybe an open source car wouldn't have all the bells in whistles a current day car has, but maybe if you could build/assemble a fully operational car for four thousand dollars, you wouldn't care.

In game theory terms, fabber technology would be like introducing a strategic move that changes the best solution of the game to some new outcome that didn't make sense before.

Whatever that post-fabber solution is, it won't make sense to us NOW.


GregLondon: No offense, but you appear to have almost no idea of what's currently involved in manufacturing.

Have you pulled apart a home stereo lately? The only fasteners in mine hold the outer cover on, everything inside is plastic moldings that just clip together. Assembled, no doubt, by a 12-year-old Malaysian making 15 cents a day. Skill not required, just small fingers. Computers are the same, that's why they're all made in 3rd-world countries. Sticking the bits together yourself isn't going to make up for the inefficiencies in your home supply chain for raw materials.

The reason modern cars have such ridiculously reliable machinery is that they're assembled to tolerances that would give a car manufacturer of the 1960s conniptions (most of them, anyway: Citroen could build stuff to micron tolerances, but then they went broke). To get that kind of accuracy out of any tool, if has to be massively stiff and free from vibration. Even a small lathe weighs about 500kg... that's not an inkjet. An open-source car would be perfectly possibly if you outsourced manufacturing of components. The megasquirt injection computer is an open-source kind of project, and seems to be going great guns.


129, 130, et seq.:

I think what the "desktop fab" will do is reopen the question of what things are better done where - at point of consumption, or locally, or in regional/national production centers, or in some remote company and shipped internationally. Then the market will start optimizing at some very different points.

Consider copying, or printing: I'll copy some stuff on my cheap-but-nice Brother multi-purpose machine, print some stuff (including limited color material) on my Oki laser. Other things, including higher volume output or printing on glossier media, I'd probably take to Kinko's. If I had to produce a glossy brochure in moderate quantities, I'd probably contact a local printing company. If I needed book-sized stuff in dozens, I'd go to a vanity press; in thousands, I'd go to a printing or publishing company, etc. A publisher may in turn contract stuff out differently according to various criteria. Forty years ago, most of those options didn't exist.

When high-speed copy machines became affordable to businesses, printing businesses were no longer the only game in town, and their economic model got shaken up quite a bit. The same happened to copy shops when copiers got cheap enough to go into small businesses, and then into the home; and concurrently, in different waves, with affordable laser printing for the office and home.

I expect that practical desktop fabs are going to shake up the available choices the same way. They will not replace factories, any more than you get most of your books laser-printed, but they will offer interesting options for items in quantity 1 and 2. I expect there will be an interesting niche for the city or neighborhood "fab shop" for a while, exactly as there were for copy shops, because the fab equipment which can work with better materials than Reprap is substantially more expensive, and would be substantially harder to build oneself.

maggie: Door into Summer is such an interesting book when viewed as futurism, isn't it? Heinlein got so much right - CAD, domestic robots (think Roomba) - and yet so much wrong at the same time.


Arrgh, messed up logic due to re-editing repeatedly. For "most of those options didn't exist", please read: "the first few options didn't exist".

The reason modern cars have such ridiculously reliable machinery is that they're assembled to tolerances that would give a car manufacturer of the 1960s conniptions (most of them, anyway: Citroen could build stuff to micron tolerances, but then they went broke). To get that kind of accuracy out of any tool, if has to be massively stiff and free from vibration. Even a small lathe weighs about 500kg... that's not an inkjet. An open-source car would be perfectly possibly if you outsourced manufacturing of components. The megasquirt injection computer is an open-source kind of project, and seems to be going great guns.

Posted by: Chris L

You just made the Brown Moties cry :-( Is there any good sf out there that realistically (yet entertainingly, one hopes) portrays the difficulty of industrial processes? I had decided by the age of seven, I think, that I wanted to be a scientist. Why? Because that's when I read Campbell's Arcot, Wade & Morrey stories. In my mind, a scientist was someone who figured stuff out and built inventions (Tom Swift was another early favorite.) The fact that they went from, 'Hmmmmm . . . this is a rather odd set of observations' to finished working prototype in about three days didn't make much of an impression on me at the time. That's why I think home delivery of specialty items will dominate over home fabbing for a long time. Well, many development cycles.


@last half-dozen posts or so:
We're also projecting based on hypothetical fabs, fed by cellulose nanofiber (and other) feedstock(s) with a tensile strength close to that of steel, and with dimensions between maybe .5 and 2 meters cubed, themselves made mostly of extruded cellulo-fiber parts, and with logic, servos and print-heads being the main unbootstrappables. So we're already way out there. Weights around 500kg and production to within micron tolerances? Ahh WTF, why not? ^_^. Makes great sno-cones, too, btw.


Chris@139: No offense, but

I think you get a bingo point for that.

you appear to have almost no idea of what's currently involved in manufacturing

Good grief. We're talking about hypothetical Santa Claus machines of the future, and you ding me for accuracy of today-tech?

There seems to be some minimum level of handwavium that is required to be involved in this conversation. Feel free to wave harder.


Now, what would be really nice is if we designed a fabbable general purpose assembly robot that could tag-team with a fab (you know, saying "fab" this much is going to get real old, real quick) to assemble new fabs, assembly-bots, and whatnot else. Just keep them in power, feedstock, servos, logic, print heads, optical gear, etc. and away they go. As far as the AI involved, this thing doesn't have to be terribly smart, as we're high-level coding its actions every step of the way. Still, it's no small feat. Which is why we need the open source development cycle.


Er, now that I think about it, this is really just a macro-scale universal assembler. It needs some special purpose parts, but that's kind of analogous to gold or platinum atoms at the nano-scale.


145: Would that be the "Ab fab"?


(/me hunts for bad Fab Four pun, comes up dry.)


GregLondon@144: :-D

I think the handwavium is being applied in the wrong places, is all. I don't think our civilisation is, by and large, "stuff" limited. Despite the best efforts of the advertising industry to persuade me otherwise. Now, if fabs could bootstrap their own local energy supply...


Once we've got the fab-plus-universal-assembly-bot-setup going things start moving faster. We iterate continuously, tweaking and bugfixing, improving the precision with every pass. At some point down the line, the fabs become able to do printed circuitry, and, eventually, microchips. Its inevitable, really, because it all already exists as expensive equipment sequestered in fabrication plants.

Of course, maybe the lead in macro-scale means of production which industry loses to the people it regains with oddly parallel developments at the nano-scale.

Regardless, tho, humanity's baseline quality of life continues ratcheting up, to which, thumbs up.


insect_hooves@148: I guess that explains where the Thunderbirds got their technology from. And we all thought "FAB" was just some groovy catchphrase.


Wait a minute! The Beatles were skinjobs! Witness Harrison's When We Was Fabbed.


I don't think our civilisation is, by and large, "stuff" limited.

Well, "stuff" is currently a zero sum game. But if fabs were built that used recycled newspapers to create a new kind of "stuff" that's as cheap a recycloed newspapers, but as strong as steel, then I think there's a lot fo stuff that starts tending towards positive sum games.

At which point, the world stops looking familiar.


Hang on, you've collectively gone a little bit over the edge here!

I'm speculating aloud about the state of 10-15 years hence garage tech. Not nano assemblers or Christmas tree machines.

Clifton @140: I think you nailed the tension between what you do at home and what you commission a factory in China to build.

Chris L @149: it seems to me that our civilization isn't "stuff" constrained so much as its constrained by its logistics chain. Right now, we have a model whereby production occurs at hubs and products are then shipped along spokes until they reach the consumers -- a star topology. Moving from that, to a network topology where some manufacturing nodes are huge factories and others are about the size of a big photocopier and sit in the equivalent of your high street Kinko's is going to be a significant change, but those high street fabs aren't by any means going to be universal, either in the scale and quality of what they can produce, or in the materials they can work in.

One thing that really interests me is disasters. Our current hub and spoke logistics models don't cope well with disasters that take out a hub, but systems that can produce essential replacement parts and equipment locally (without them needing to be shipped halfway around the world) could potentially speed disaster recovery programs. Or help deal with major transport interruptions. And so on.

As for shopping ...

The high street store prognosis in the UK is already grim; they're facing stiff competition from the internet, and a lot of chains are under pressure everywhere but in their online storefronts. There are many items that we can wait a couple of days for, and they're vulnerable to price competition in those areas. (It's a much smaller problem for the food retailers; if you're hungry you can't wait 48 hours for a delivery.) Fabs might be a partial solution in some sectors, offering shops the ability to do boutique customized products with a much faster turn-around than an internet mediated fulfilment house can manage. But if so, that's going to radically change the whole look of your local high street or shopping mall.

And so on.


On the military applications, it strikes me that although rolling your own rifle would be cool, nobody who wants war ever seems to be constrained by the availability of small arms, and as pointed out above, common metalworking tools will do it anyway.

I'm not sure -- wouldn't this eliminate the need to get a backer to supply you with arms?

And while I doubt it would foster many more wars, wouldn't it change the winners, as controlling supply chains becomes way less important?


insect_hooves@125: Yes, by I can't remember who. It plays out in a world where really smart AIs are impossible but everyone has a nano assember in their home that is connected to a "feed pipe" shoving neatly stacked building blocks into the assembler.

Charlie@154: You are probably right that we went a little nutty but that's what happens when you suggest something even remotely like nano assemblers to a group of sci-fi geeks. It's something that will change life so completly and has so much potential to change it for the better (not excluding changing it even more for the worse) that we can't stop ourselves from voicing our dreams/hopes/fears.


Keir, I think we'd need some fundamental changes in warfare before fabs would make a difference.

The overwhelming logistical bottleneck is energy, in several forms. You have to feed guns and vehicles and troops and electronics: POL and ammunition and batteries and rat-packs.

Small scale renewables, wind and solar, already seem to pay off for the Fort Zinderneuf equivalents of Iraq and Afghanistan. They give a little more resilience on night sights and radios and such stuff. Likewise LEDs instead of hot wire for illumination.

Fabs can help with making the living tolerable. They're a long way from helping with the fighting.


Apropos Dave @158: the figure I've heard is that every US soldier in Iraq consumes 15 US gallons of fuel per day. (On average, of course.) And that a squad of US army soldiers on exercise get through something like 150 batteries, in 7 different shapes and sizes, over 5 days.

Energy storage media aren't something you can fab your way around. (Hence DARPA's interest in orbital power-sats: energy on tap, beamed down to Army units in the field anywhere in the world, with a free death ray thrown in on top!)


Charlie@154: I just don't see fabs being terribly disruptive in the commodities market -- they can't make enough different stuff. Same with disaster recovery. What could we possibly fab that would help out with an earthquake/flood/terror-attack?

The rapid prototyping angle seems to be the disruptive bit. Give a bunch of computer and mechanical engineering nerds some engineering PlayDoh and watch what self-assembles. And robotics seems to be the only macroscopic area with enough headroom that we could see continued, disruptive growth.


Er, you mentioned vehicle replacement parts, which is a good point.


I guess we can sort of see the seeds of it already. There's a flyswatter, some shoes, a doorhandle, etc. But most of the photos contain RepRap's own parts (robotics), and a gear-train.


After re-reading 154, I completely agree with you. I managed to totally misunderstand what you were saying, even though you were perfectly clear. The means of production will probably morph from a star to a mesh topology, and the added redundancy and decentralization makes the system much more resilient. This will probably happen in parallel to the Open Hardware revolution. This relatively recent Wired article is sort of interesting, and mentions rapid prototyping.



Charlie, I don't think those orbital power sats are such a good idea. After all rocket science isn't exactly rocket science these days and those satellites would he a HUGE target. (360,000 sq m for about 100 MW of power is a 600m x 600m target.) (For literary reference, see "Green Mars".)



The short term is going to look like the recent past in one respect: any business that doesn't start to think ahead is going to be found in the tar pits in a few megayears. The ability to sell a very large diversity of retail items without a long supply chain will erode the sales of any company that tries to hang onto its supply chain and not use the new distribution methods where possible, if only because the increasing cost of fuel for distribution will destroy their profit margins. This will ripple back to the factories in the developing countries like Thailand and Vietnam, which get a lot of their international income from selling to distributors in the US and Europe (and soon to China). China, on the other hand, is already starting to push high-tech manufacturing, which would leave them in a good place to sell the unfabbable* components that fabbing retailers would need. Some of the industrializing nations would be forced into a search for another business model for their manufacturing capacity.

So the map of globalization will be changed significantly by even a rollout of fab capability based on what reprap can do now. At the point where the unfabbable components are somewhat standardized, the availability of fabs in a Kinko-like local setting makes home fabrication for criminal or terrorist purposes mostly unnecessary as long as one or more people on the staff are part of the gang and the quantities required aren't high.

I wonder what effect this would have on smuggling, especially small arms and drugs. You could fab a replacement for a car door with internal parts that would hold smuggled goods and be designed to spoof any x-ray or terahertz imager by incorporating the goods into the larger image of the door parts. Searchers would have to cut the door apart to find anything, so it's not perfect stealth for smugglers, but it would pass a screen that most inspectors would be inclined to trust.

One of the big changes with local fabbing would be a reduction in the time from working prototype of a product design to receipt of the first unit sold. For one-off parts that probably wouldn't be a big deal, and it wouldn't affect high-volume business, but in between, say for 5 to 100 or a 1000 units (per customized variant of product) it might change the way short-sales-life-time products like toys, car accessories, kitchen utensils and plates, hand tools, etc are marketed and sold. It would also have a large and probably unpredictable effect on the economies of the developed nations, especially ones like the US whose economies are heavily dependent on the sales of consumer products.

Lest you think I'm concentrating on frivolous things, look at the effect on military supply chains, especially for rapid-deployment expeditionary forces. Rapid response, highly-mobile forces, Marines or Army in the US, are largely composed of ground troopers and light, fast vehicles; trucks and humvee-equivalents. Aside from weapons, ammunition, and food, the largest supply problems are repair parts for vehicles and weapons, many of which could be fabbed locally, a trooper's personal equipment and replacements for same: packs, harnesses, helmets, boots, binoculars, entrenching tools, canteens, mess kits etc., and all the requirements of fast-setup command and bivouac posts: tents, folding furniture, etc. The less of this stuff has to be shipped in before or with the first batch of troops the faster the deployment can hit the ground and start moving towards its first objective. If a lot of the materiel, especially repair and replacement items don't have to be shipped in, and especially if fab feed stock can be obtained locally, this can cut days or even weeks off deployment time. This might be especially important in regional conflicts between non-superpowers, where the first side to mobilize effectively might be the winner almost at once.

* I may get tired of this word before I'm done.


Perhaps we should be asking different type of questions.

How large is the fab device? Can I actually fit it and it's equipment for producing the raw materials into anything smaller than a small store.

Will I be able to make everything with one device? It might be more effective to produce one item in the paper/steel fab device and another in the silver fab device.

What we may end up with is mini industrial centres that provide a neighborhood with a variety of different specialty products. Just as likely a super market style that provides everything, including the unfabbable items manufactured in China.

It might be that we end up with a Walmart buying up our waste cheap and producing the fabbed items for us because they can afford bulk purchase and operation of the machines.

Yes, we could produce the stuff at home, but would it be worth the effort?


Bruce@165: All good points, as usual. Now, from gonzo-speculation-mode, I wonder at the implications of stirring unmanned warfare into the mix. Airdrop a bunch of all-in-one fab-plants-in-a-box into strategic remote locations, with feedstock and all the necessary ammunition and componentry, and they start churning out death drones like an f-ing hornets' nest. The supply chain complexity drops multiple orders of magnitude. The drones are controlled by a generation of gamer kids on the opposite side of the globe. And for the 's twist, you don't even tell the kids they're fighting a war. Tell them its some gaming competition or something. High Score!


Whoops, that last sentence was supposed to read: "And for the (book title redacted to avoid spoiler)'s twist, you don't even tell the kids they're fighting a war. Tell them it's a gaming competition or something. High Score! (mother clutches baby, weeps at the sky)"

Silly angle brackets and their "mark-up". It'll never catch on, I tell you.


Bruce et al
All of the following are made from very different materials for a reason.
packs, harnesses, helmets, boots, binoculars, entrenching tools, canteens, mess kits etc., and all the requirements of fast-setup command and bivouac posts: tents, folding furniture, etc.
I'd like to see you get optical grade lenses out of the 3D printer Charlie proposed. Or if you are using handwavium, a significant part of that is learning how to describe tempered steel edges to the miracle box.
Tensile strength is also a minor quality, when your concerned with things like durability, ductility, flexibility and a range of material properties that one encounters in every day use.
Laying down layers of nanopaper --sounds what you need is programmable wasps.

Some one mentioned high street fab stores. Sounds like a typical nineteenth century high street to me. With a range of light industry and workshops behind shops which could cater for your orders. The Victorians had their own problems with energy, communication and logistics. They didn't have to worry about standardization because everything was made by hand down to the screws -okay I conflate my industrial history a bit. Sounds like a zoning problem.

Another thing talking about supply chain. Big stores subcontract their logistics to specialized companies. Some of these companies in the UK are moving back on to the railways because its cheaper and they don't have to compete with Continental firms buying petrol abroad and driving in the UK. What happens when the logistics firms throw in the towel?



Many components today are made, not from fabricated pieces of a single material, but from extruded or formed layers of different materials with different characteristics. The effect is to create what on a large scale (centimeters) looks like a continuous gradient of a material property. For instance, a shock-absorbing strut whose ends are made rigid and whose middle is elastic, but all made of a common material with the amount of a (e.g.) plasticizer varied between end and middle. Or a material which is opaque in one area and transparent in another. This is precisely the sort of thing a fabber is good for. What's needed is a wider range of materials and material components to put into the fabber.

True, I was reaching with the optical parts, though I wonder if molding Fresnel lenses or diffraction gratings would work (probably not without much tighter tolerances). Of course you could always create a photographic plate out of a piece of rigid plastic by mixing in some emulsion and develop a 3D holographic lens in it, though that might be harder than extruding the average piece of plastic.

All of the characteristics you mentioned, as well as electrical conductivity, thermal conductivity, rigidity, and others are variable over some range using materials that can be used in some 3D printer today. The big problem is that no one printer can handle the whole range, and the cheapest printers are the most limited. This is primarily an engineering problem (one that will entertain some really capable engineers for some time, but still, it's very likely to be solved to some level within the next few years).

As for supply chain, as I mentioned above, I worked up until 2 years ago for Nike, which has one of the most sophisticated supply chain systems in the world. Note that they're not bad at their business*, and they divested themselves of almost all their manufacturing capacity while spending a vast (some would say ridiculous) amount on perfecting the supply chain.

* There are things theat they're bad at, ethics being one of them, but I like to have a beer or two before getting into those.



Sorry, I left out an important point. The way to make a success of a new technology is never to try to make it as good as an old technology at what the old technology does, but to figure out how to get the same results with different methods.

Back when I was a kid, some very talented people figured out truly amazing ways to make vacuum tubes smaller and put more functionality into each one. It turned out not to be a good way to build large-scale digital systems*; integrated solid state devices work very differently but achieve the end goal better**.

* Vacuum tubes were used in digital systems as late as the late 1960s; I worked with some around then.
** Even though today there actually are technologies that could put millions of vacuum tubes in the volume of a microchip, I don't think anyone's seriously trying to commercialize such a thing.


It occurs to me that one little niche for near-current fab technology (needs a suitably durable fab-substrate, for one thing) is the market for garden statuary, whether it's a cute hedgehog with a wheelbarrow, a standard garden gnome, or classical-Greek repro.

The sort of casting process these are currently made by is a pre-computer fab process anyway.

Nobody can stock the full range. You browse through a catalogue, put in your order, and wait.

Now, imagine the same with a fab. You can display the digital model in a form that you can rotate and translate on the screen. Yes, there would be a photo as well. The fab can scale the product to suit your garden (there are limits--you can make the gnome half standard size, but the system had better know the minimum thickness for the fishing rod). And, unless a queue has built up, you can take it home with you.

The material doesn't have to be as durable as the fake stone, but the lifetime/cost balance needs to be right.

And how about garden-size versions of those erotic statues? Open-crotch open-source... If you fab a customer's design, who is liable under any obscenity laws?

I've seen some weird stuff in VR, and I don't think some of them were meant to support hanging-baskets. "Mellors, the Greenhouse!"


Nobody seems to be talking about the COST of these Fabbing/RepRap machines.
It would appear that a GOOD one costs about $50k, but the price is coming down.
Now we have an existing model for this, which we've all seen twice in the past 30 years....
Mobile Phones and "xerox" machines.
Specifically photocopiers are an interesting model to follow. As recently as about 1985 Xerox were still selling top-of-the range commercial high-speed photocopiers for business use only at silly prices.
Such machines DO still exist, especially for really high-quality, large-area multicolour work.
Meanwhile, everyone else has a perfectly adequate color-printer attached to their PC.

So, this revolution IS going to happen, but slowly, as prices fall, and versatility increases, so that the machines become more multi-capable in the materials they can handle.
This is (ahem) MERELY standard development-and-production engineering. But it does take time and costs money so to do.
However, within (I guess, and it IS a guess) a minimum of 5 and a maximum of 20 years we are THEN going to see machines that will be available for no more than the cost of a good camera today ( 1k-2k = $2-4k) which will be able to use the new paper-stock referred to, AND plastic feedstock, AND sintered metal, as referenced by Bathsheba Grossman in her pages.
THEN the change happens - very quickly.

IF by that time we DO have a strong AI up-and-running, that really would be the big S, would it not?


There's another historical model to follow.
The Steam Locomotive.

Trevithick built a working one in 1803.
Then the scene moved to the Durham/Northumberland coalfield, but progress was slow, and intermittent for 15 years, as various mechanisms were tried, and manufacturing/QC problems were dealt with.
even so, "Locomotion No1" of 1824/5 was still very primitive.
Yet by 1829, the "project" had cohered, bacause the parts/design necessary for a useful feedback system had been realised and made: blast-exhaust giving automatic fire-draw, multi-tube bolier giving hugely improved heating surface, and hence efficiency, separate smokebox to deal with unburnt residues, a reasonable valve-gear for ease of operation and expansive working (hence greater efficiency again) and a water-jacketed firebox incorporated into the build. The "Rocket" class had all of these improvements (though the integral firebox was only on the last one - "Northumbrian", I think).
By 1833 the fully "modern" Locomotive had arrived, in the shape of "Patentee".

Look at that development profile, though.
Steam power, Newcomen 1710, higher-pressure Watt 1776-84, first mobile power, 1803-08, 20 years halting development, and then, between 1828 and 1833 it all came together, and the world has never been the same since.

There is no doubt that Geordie Stephenson, and his rivals and collaborators changed the planet more than any politician, even, or especially Boney.
But who, except for a few mere techies, like us, really understands that?


174: If I'm not mistaken, work on the steam locomotive was the precursor to much of what became thermodynamics.

172: Instead of forming an item on the spot, how about extremely transportation. Select from a catalog, press 'enter' and boom! The item is shot towards your location at several hundred kilometers an hour; twenty minutes later, after having a quick one across the street, you pick it up.

While Lawn Gnomes might be made more cheaply on the spot, it might be more economical for anything much more complex to be delivered. As to the hypothetical transport network, pneumatic tubes are an oldie but goodie(no petroleum was burnt in the shipping of this product! says the green sticker.) Vinge also wrote a few stories which featured a rocket-powered UPS.


174: If I'm not mistaken, work on the steam locomotive was the precursor to much of what became thermodynamics.

172: Instead of forming an item on the spot, how about extremely transportation. Select from a catalog, press 'enter' and boom! The item is shot towards your location at several hundred kilometers an hour; twenty minutes later, after having a quick one across the street, you pick it up.

While Lawn Gnomes might be made more cheaply on the spot, it might be more economical for anything much more complex to be delivered. As to the hypothetical transport network, pneumatic tubes are an oldie but goodie(no petroleum was burnt in the shipping of this product! says the green sticker.) Vinge also wrote a few stories which featured a rocket-powered UPS.


I'm speculating aloud about the state of 10-15 years hence garage tech.

Well, a lot can happen in 15 years.

small scale integrated circuits, for example. According to wikipedia, aerospace programs purchased almost all of the available small scale integrated circuits from 1960 through 1963, and almost alone provided the demand that funded the production improvements to get the production costs from $1000/circuit (in 1960 dollars) to merely $25/circuit (in 1963 dollars)

Of course, it's tagged with [citation needed], so...

The first mobile phone in the US (Motorola DynaTAC 8000X) received FCC acceptance in 1983. It weighed 28 ounces, was 10 inches long (not including the antenna), had 60 minutes of talk time, and took 10 hours to recharge.

It cost $4,000 in 1983 dollars.

fifteen years later, 1998, the Nokia 6160 candybar phone was 5.2 inches tall, weighed 3.4 ounces, . I'm trying to find a 1998 price and I"m not sure, but one review I found makes it sound like it sold for $70. Of course, the cost of the phone might be $200, but is offset by the 2 year contract, but still.

history of cell phones:


Anyway, if you can buy a working prototyper now for 10,000 dollars, I don't think its terribly insane to think that in 15 years you might be able to buy one for $500 or something, and that 15 year from now version might be way better than today's 10,000 version.


Heinlein on pneumatic tubes and rtelated transportation infrastructure (as cited in Technovelgy web site):

Robert Heinlein (1951)

Bounce Tube - pneumatic tube system for people
Robert Heinlein (1956)

Copter Harness
Robert Heinlein (1954)

Jump Harness - rocket pack
Robert Heinlein (1961)

Motorized Combat Boots - quick in a crowd
Charles Stross (2005)

Ramsbotham Gate - wormhole
Robert Heinlein (1955)

Ring Road - mag-lev old-style
Robert Heinlein (1953)

Rolling Road - public transport
Robert Heinlein (1940)

Tumblebug - a gyro-stabilized monocycle
Robert Heinlein (1940)

"We hit the sub-basement and went at once to the express tubes. A two passenger capsule was just emptying; Dak shoved me in so quickly that I did not see him set the door combination..."

"The few minutes we had been crammed in the vacutube had been long enough for me to devise a plan..."

From Double Star, by Robert Heinlein.
Published by Doubleday in 1956

And, ancestral in his writing, though not our reading:

Heinlein's "For We the Living" [written 1938; published 2003, posthumously] "describes a system not unlike a subset of the web, by which one could request information through something like a terminal, and it's printed on demand and delivered by pneumatic tube."


G.Tingey @ 173: That cost factor is precisely why I was talking about the "copy shop" or "service bureau" business model upthread at 140. It's a classic successful business model for technologies in the transition phase. (However, it's a short-term business model unless you can figure out how to parlay it into a number of other niches. Lots of print shops and copy shops went bust after copier prices dropped far enough.)

One factor I haven't seen discussed much yet is time/object. Contra Dave Bell's suggestion of Lawn Gnomes While 'U' Wait, the current tech takes about 5-6 hours to produce a single moderate-sized ("breadbox") solid plaster object. The prospective lawn gnome purchaser will get awfully bored waiting. No doubt the future tech will get faster, but how fast will it get faster?

Note that "printing" time doesn't merely set the parameters of how the tech is used, it also sets some lower bounds on unit pricing. Assume you have a $50K machine which takes 6 hours to produce a single object, and which you need to amortize over 2 years - then you have a basic $17 cost due to equipment depreciation built into every object, even if you're running the machine non-stop day and night.


Clifton @ 179

Well, let's see. I've taken some printer specs from here and tried some SWAGs. Assume that a 3D printer works enough like an inkjet for reasonable approximation. At circa 1500 dpi (about 5X the resolution of low-cost photopolymer printers, my printer, which cost about USD$250 three years ago, takes about 3 minutes to print a single photograph, which usually means printing every spot on the printable area of the page, so it ought to be slower than laying down a single layer of a component, which can be optimized to skip over the voids. Rather than being that pessimistic, let's assume the average layer is 30% populated with material (there's no reason for a garden gnome to be solid, and in fact the plastic ones are injection-molded, IIRC). If we assume an average object of about 25 cm in height, which is certainly of the right order for current machines, and a layer thickness of 0.2 mm, which is reasonable using a grainy material like metal for sintering or polymer beads for photopolymerization without milling down each layer, we're talking about (3 * 50 * 25 * 0.3 * 0.04 for the difference in resolution) minutes, about 45 minutes. That's faster than current tech for 3D printers primarily, I think, because inkjets are optimized for speed using inks that dry rapidly and don't need an additional light exposure or curing time. So better raw materials and cheaper, higher-power light sources would speed things up without requiring tinkering with the motors or moving parts.


Garden statuary in general, time could be a big issue.

But remember, it's an example of a product which doesn't need complex structures, where the ability to manufacture to different sizes might be an advantage, and where there is a wide range of potentially makeable objects. which cannot be practically stocked locally.

(Incidentally, hollow figures would be an advantage in many ways. It's one of the little manufacturing tricks that made Britains so big in the toy soldier business.)


That reminds me.

How do objects get coloured?

A lot today is done with self-coloured plastics, and a fab might be able to handle that--not four-ccolur printing, but more like the Pantone system to handle three or four different specific colours.

One of the big developments in model railways, back in the late Seventies and the early Eighties was the development of automatic painting. The quality of the paint schemes jumped remarkably with the ability to use machines instead of humans. Better paints too. The over-glossy finish of the Sixties (without the dust and water in a scale atmosphere, the same gloss paint a foot away is too shiney to look as though it is 70 feet away) was replaced with a more convincing finish.

I can see an object being fabbed with a standard base that lets a painting machine hold it in a precise orientation for the finishing.


Are there any other small scale manufacturing techniques, or at least large scale manufacturing techniques that could be scaled down? I remember the big scare in the 80's was an isotopic separation scheme with lasers that could be done in a small room. Presto! Everyman able to supply himself with ingots of U-235, suitable for Little-Boy type atomic bombs.

Is there such a thing as a Universal Manipulater on the drawing board? That would be a _big_ step towards autonomous small-scale manufacturing. So would composites that do the more traditional work of steel and aluminum for structural support, at least if they're easier to work with in the formative stage. Stampers and rollers don't seem to be the kind of tools that can be miniaturized.


Dave Bell,

There are rapid prototypers today that use multiple colors (I think they mix 3 or four differently colored plastic resins together for a wide range) on the same scale as the spatial features they can produce. Go here and scroll about a third of the way down to see some example fabrications.


Check out the FabLab at MIT's Media Lab. They're finding out what people would do with fabbers by letting them use existing tools like computer-controlled laser cutters and such. They're not wedded to additive 3D printer technology, so they're looking at any kind of fabrication using reasonably available materials that can be controlled by a computer.


Dave, Bruce:

No, those are not bits of colored plastic mixed; it's far simpler than that. I've seen a presentation/lecture with a demo of the Z-Corp machine's output, and got a full explanation of how it works.

The Z-Corp design is brilliantly simple; I mentioned it briefly upstream but didn't explain. It literally prints plaster objects in 3D.

The material is specially prepared fine plaster of Paris with some binders, which is blown onto a table which can be controlled in height; that gives you your Z axis. The machine then runs a standard (more or less) HP inkjet print-head over it, "printing" the dots that should be solid onto it with plain water. The water reacts with the plaster to bind it to the solid dots below and around it; it hardens fast enough that by the time one layer is finished the machine can lower the table, blow on another fraction of millimeter thickness of plaster, smooth it out, and print the next layer.

For color? Simply use a color print-head with standard inkjet color inks. The color goes all the way through.

My number of 6 hours was a rough estimate of what it really takes right now as I recalled from that presentation; maybe as little as 4 for smaller objects, but definitely on that scale. For your estimate based on photo-printing, you need to multiply current time to print an image by the vertical resolution of the object, and I think you'll get the same ballpark number.

The reason it's still valuable despite this limitation is in replacing things that it would take days of skilled labor to construct.

The company running one here has architectural firms among its clients. Building a large architectural model for presentation, from their CAD drawings for it, formerly could take days of crafting by several junior architects. Now they can sub it to the machine and get back a model (perhaps in sections to assemble) the next day.


Something I haven't seen mentioned among the very interesting speculations here: if these things actually take off (as assumed for the purposes of this thread), we're going to run out of old newspapers awfully fast. It is possible the folks who would like to DRM this would be best served by controlling the feedstock supply. Your friendly neighbourhood printer company could simply refuse to sell more than a certain quantity to anyone.

Of course, there are ways around this too.


@ #186
Of course ... just set up your own coppicing rotation!


"paper is the new cast-iron"

I recall when we had a copy of the "Children's Encyclopedia", and one of the things in the index were sort-of picture indexes of things, like parts of a suit of armour, and medieval polearms. Amongst there were lists of things made of concrete, and celluloid, and cast-iron.

Cast-iron wasn't a Victorian invention, but it was used for an incredible variety of items in that era. Not all depended on the structural quality of the material. Those decorative spikes on the railings were cast iron.

There was a page on papier-mache as well. It didn't tell you about the iron armature that stopped the papier-mache chair from collapsing.

There are problems with casting, from how the mould splits. You can't let a mould-section hook around the object being cast.

I wouldn't be surprised if some fab systems had problems with some elements of an object's shape.



Dave et al A lot of Victorian paper maché furniture was made pressing pre made card into moulds as laminates and gluing them in place. It was a cheap trick to make them look like Japanese lacquer ware. If you are working with well designed paper slurry and depositional process you could probably get something reasonably robust. though I don't know how it would cope with the torsion strains of a wriggler.


Maggie, that suggests other ways of upgrading a basic fab-thingy.

(Crossover with something I just saw elsenet: somebody asking about vacuum-forming armour for costuming.)

Vacuum-forming and GRP depend on making a mould, which can be a significant one-off cost.

A fab could make a one-off 3D object. That could be the mould (plaster-derived fabbing), or it could be a shell which needs reinforcement (papier-mache with a backing layer of GRP to take the loads from straps?).

Next step...

If you can get a good 3D scan of an actor's face, you could make it a lot easier to make facial prosthetics, for film work and such. The current process starts with a female cast taken from the actor's face (not comfortable) which is then used to make a male mould-half, used as the base for the shape of the prothetic. Scan-and-fab would let you make both halves of the mould, either scanning a model of the final face, or working from a wholly digital model.

This doesn't need vast leaps in the tech. It might even be possible now. And you could also do CGI preview to check the look, and tweak such things as wrinkles around mouth and eyes to match the actor's face shape.



You've reminded me that quite sturdy furniture can be made by laminating corrugated cardboard into thick sheets and cutting the sheets to make the parts of the furniture. I've seen a couch made as a demonstration; it was ugly, but strong enough to hold several adults, and fireproof besides (at least, you could put a cigarette out against it without it catching fire).

Designing load-bearing pieces for 3D fab using corrugations might make paper much more useful.

Dave Bell,

I think 3D laser digitizing is up to scanning faces (it's been used to create a CGI triangle mesh from an actor's face as a basic form for animating from motion capture for several years, IIRC). I'm not sure if the holdup in using it for prosthetics is conservatism, or the time and human effort required to get a scan that has a really high resolution at joins and places of highly variable curvature like the chin and the nose, and then tinker with it as required to get final fit.


This is where new technology comes in to its own, understanding the new ways to play with it while not forgetting all the lost by-ways of existing tech. The useful thing about 3D printers is their abillity to make moulds easily. You don't 'need' the actors head. You can then knit a complex shape to reinforce what ever binder you like. A bit like what the Northwest Composites Centre http://www.futurecomposites.org.uk/index.html arn't saying they are doing. But what do I know.


Here's a nice, very recent article on RepRap in PC World.


I wonder how many patents we've piut the kibosh on today.


An obvious interaction of fabs and art:

Masks and puppetry in theatres. We're not talking close-up-quality prosthetics, but think of The Mikado with caricature masks of well-known public figures. Or any other piece of theatre.

And you'll never be able to ban Guy Fawkes.



I love the idea of making puppets with a fabber. You could get some pretty nice joints and movable eyes, etc.

And check out the Fabaroni. It's a 3D printer that extrudes pasta.


I came across a link to a 3D printer made of Legos that built objects made of chocolate. But when I first saw the link I thought it said that it created objects out of Legos which were made of chocolate, which started me thinking.

Suppose a printer manufactured tiny Lego-like bricks, with Lego-like buttons on top and bottom and dado-like interlocking fingers on the side. It could use any material that could be poured into a small mold and cured by heat, light, chemicals, or evaporation of a solvent (I'm talking about less than 1 mm on a side, preferably only 200 - 300 μ). Then they're stacked in a delivery tube whose end is precisely positioned at the next place where a brick needs to be inserted. If placement is close enough to allow the large ends of the dado to engage, the the brick will self-register with its neighbors as it is pushed down. Then the end of the delivery tube steps to the next point where a brick is needed. You do all the bricks in a layer then pull the tube up one Z increment, just like an extruder. There are obviously some interesting problems, like how to register bricks with only vertical neighbors, and choosing materials that allow you to make bricks with such fine details, but it would be fun to play with the idea.


Anyone who watched tonight's Doctor Who Confidential will have seen examples of ares in which a 3D printer could do useful things. The episode centres on a particular established villain/monsyer combination, involving both prosthetic masks and animaytonic mechanisms.

I've already mentioned the idea that 3D scan/print processes could turn out facial prosthetics by making the moulds more easily. You also have the potential for a 3D digital model which can be split into components using different materials, rigid and flexible (there's a glove)

(I am trying to avoid spoilers)

Once you have a basic digital model of a small part, used many times with varying sizea, you could make them easily. Again, I've mentioned the potential for scaling, and the problem that some parts shouldn't change size.

At some point there is going to be somebody with a general-purpose animatronic tail/tentacle system, complete with a little computer translating human command input into something the servomechs can use to get the desired movement. (This is something that you can already get in CGI models, so that a single Side-Side control on a tail adjusts all the individual joints of the tail in a sensible way.)

Though by the time the fab-tech is good enough, the film and TV people will probably be doing the job with CGI, using some markers on the actor's costume for the CGI system to anchor the tail. (Remember, a month or three back, the demo of using the Wii remote as a position sensor--the problem is similar.)