« The future, Indian-style | Main | Historical contingency »

And the beat goes on ...

Two sets of page proofs checked, one to go.

When not grovelling over a hot galley proof, I've been upgrading Macs — I've just installed Snow Leopard on all the members of the herd that would run it, and replaced my other half's DTP workstation — the only current member of the herd that wouldn't. (Memo to self: write another book to pay for the Mac Pro.) A handful of minor annoyances have come to light, but nothing earth-shattering (scanner drivers that crash under repeatable and avoidable circumstances; an external hard drive's power supply that decided to go on strike just when it was really needed: that sort of thing).

I should have the storm of page proofs nailed down by the end of the week, in time to actually take a couple of days off before I get back down to writing. (And the ongoing project to drag the look and feel of my website kicking and screaming into the twenty-first century.)

Meanwhile — to continue the IT theme of this entry — Microsoft have startled me by demoing something that actually looks well-designed and potentially useful. (As I'm definitely not a fan of Microsoft, you should take this statement to be extremely unusual.) The smartest comment I've seen on the subject of the Courier project is that it looks like it started out as a spoiler for the XO-2 but has turned into Microsoft's answer to Apple's Tablet. This could be very interesting; with the exception of Netbooks (basically cheap laptops cannibalizing the industry's margins from below) and smartphones (basically your old-fashioned PDA mashed up with a mobile phone) we haven't seen a new form factor really take off in ages — tablet PCs and swivel-screen laptops have failed to be more than a tiny niche market since the late 1990s. However, we seem to finally be seeing a lot of noise about ebooks; Dan Brown's latest apparently shifted a dizzying, huge 5% of total sales in ebook formats, and everyone and their dog are launching dedicated readers. Something like Courier or the iTablet, however, look far more useful to me: general purpose computing platforms that also kick sand in the face of the ebook readers on overal functionality. If both major OS vendors are now taking them seriously ... who knows?

(Edit: For a contarian viewpoint, here's a salutary reminder of how Microsoft do spoilers whenever it looks like someone else is threatening their cash cows. Is Courier real? Who knows? What we do know is that it means they're staying up late in Redmond, worrying about the iTablet ...)




Courier is definitely exciting in terms of form factor and interface, but unless those are something more than just touch screen LCD displays (color e-Ink?) I doubt the current ebook readers have much to fear.

Also, I kind of hope this is just a reference design to be licensed to other makers. Microsoft has a short and spotty history with selling their own hardware.

Zune - Me-Too ipod at a competitive price with comparable features... ok, sure.

Microsoft Surface - The future is a hugely expensive big ass table!


Don't worry Charlie, they screwed it up already. There seems to be evidence that the thing will be wirelessly powered.

Sure, that makes for a sleek design. But recharging the battery will require you to put it on some sort of pad, which means that you won't be able to move it in more than one axis (perpendicular to the pad) while it is recharging. And of course you won't be able to indulge in the luxury of simply swapping the battery, since that would ruin the sleek design ...

So, you have a gadget that won't run for much more than 10 hours and will inevitably run out of juice when you most need it, at which point it will be crippled.


As @ben noted, tablets won't get a leg up on ebook readers until they have ambient light displays. I can partially close my laptop to shield it from the sun, but non-folding tablets are completely at the mercy of the lighting conditions (unlike laptops, which are only mostly so.)

Ambient light displays (color electronic ink, reflective LCD, or some yet-to-be-invented tech) can be used in full sun or a dark room. They are go-anywhere, which is the niche that I see as most hotly contested right now.


I saw this on Gizmodo this morning (or last night... too much stuff!)

I thought it was a great idea. I am definitely attracted to a tablet form factor (I doodle, I spend too much time in meetings, I take lots of free-form notes) and to eBooks (I read way too much and paper is just so damn heavy!).

Nothing yet provides what I need (a Kindle that also does 'work'?) Something with low power demand for long-stability views (reading a word doc, a slide, or a spreadsheet) but fast response for changes (animations, video, edits, etc)

I saw a link (somewhere) to a dual layer screen that was e-Ink plus oled. You would need the oled for video, the e-ink for low power, long retention. With a properly written window manager, you could take advantage of both (assuming you could switch segments of the oled/e-ink on and off at will) giving truly the best of both worlds.

If this does not yet exist - then consider this to be my contribution to 'free' enterprise. Use it if you will (H/T Max). Switching between modes should be seamless, BTW, and essentially invisible to the user! (I imagine the biggest challenge would be matching the color gamut -- but maybe not doing so would be a selling point! grey for static/low power/ - green; full color for active/high power)


Well...MS Labs is cutting edge in UI design. Will they market it? Who knows!


Tony: that variety of screen is what the One Laptop Per Child folks use in the XO-1. It's being commercialized for laptops and netbooks -- should be hitting the streets in 2010. (Originally: in Q3/2009. I'm padding a little, for the usual productization delay -- but note that it's a development of a tech that's already out there in reasonably large production volume via OLPC.)


The Raven @ 5:
I was amazed to hear that the head of MS Labs told one scientist he was recruiting that he would never have to worry about his work being used in a product. But then the people I know who've worked there say they didn't have to think about products at all. It's all very well so support basic research which may or may not give any return, but insisting that your entire lab do that seems, um, counter-productive.

Charlie @ OP:
MS learned all their marketing strategies from IBM. Remember the 390 series computers? Years of PR, and not an ounce of engineering; they never intended to build them, just wanted to kill CDC and Cray. So I doubt that Courier is real. Even if it is, think of how well the OS will be designed.


charlie@6: thanks for that. glad to see that my predictive abilities* are still functioning to the same standard :~P

* I have highly-functioning integrative retroactively-predictive synthesis disorder (I can clearly see into the past, but never know exactly what I'm looking at)


Part of it is that Bell Labs model. Most people think that Bell Labs was mainly a pure research operation. Not so. Bell Labs was the engineering arm of Western Electric, which made all of the Bell System's equipment, and ATT. It is even in Wikipedia. Basic research was a very small part of Bell Labs work. MS doesn't get it. But then, it's a different landscape. The grand old US research labs depended, first, on monopoly profits and, second, on government funding, largely for military projects. The government funding has dried up. MS has monopoly profits--but who knows for how long? Google is likely to outlive them.

But personally, I'd bet on China.


Forgive my going OT, but I see Amazon (.co.uk) are now listing both Fuller Memorandum (Orbit paperback) and Trade of Queens (Tor hardback) for next year.

Searching on "Stross" also gives "A Touch of the Sun" (1956, VHS, Frankie Howerd).


Charlie: I looked at the XO screen -- similar, but no banana... (I knew I'd seen something like it)

In the XO there is an active LCD (which still takes power to refresh each frame), and a colored backlight (similar to a crt phosphor mask) resulting in a matrix of R G B sub-pixels. Each active LCD pixel maps to a single color sub-pixel (resulting in a lower efffective color resolution).

My scheme is also two layers.
Layer 1 (top) is e-Ink. Grayscale - should be at least 8-bits deep (since we want this to be a *nice* display!)
Layer 2 (mid) is an active-LCD. Color - should be capable of high-gamut response (so 12 bits deep per sub-pixel)?
Layer 3 (base) is an LCD backlight (distributed).

Layer 1 needs to be reasonably transmissive (when 'clear') to allow the underlying LCD pixels to shine through.

Each layer is segmented (how many segments? with a trade off between build cost/power use/control/fuglies!) into individually switchable rectangular areas (layers share the same segmentation).

An individual rectangular segment can be switched as Layer 1 (low power, grayscale ink) or Layer 2/3 (high power, color lcd with backlight). make the segments small enough, and each pixel becomes individually switchable (ink or lcd)

Active page elements are rendered directly on the color plane (so as to provide reasonable response to typing, video frames, etc). Inactive (static as indicated directly by application, or unchanged during a specified time-out interval) are rendered onto the ink plane.

The visible screen is a combination of the two planes. (as you draw ink on the screen, the renderer 'lights' the penned pixels in color [plus surrounding pixels in the rectangular 'segments']. As pixels remain unchanged - they dim to grayscale [pretty!]

Switching modes is relatively 'cheap' since that's essentially what LCDs do each frame in any case (we're simply making the frame blanking interval *much* longer)

Now that I've said this - I want one of these. Low power - except when you actually want to use it.
windowed video that uses only enough power to drive the small window (not the entire screen). readable in all lighting conditions. And immediately responsive. and did I mention low power and readable in all lighting conditions.

Obviously - the controlling system should allow direct control over the screen management (I may want speed, or I may be happy with smears - it should ultimately be the user's choice)

Give me color e-ink and it's even better. But I'd settle for B&W (and being an aesthete I actually like gray!)

Does anyone know any good screen fabricators out there?


Both the Codex & Courier had interesting interfaces, based heavily on transclusions. I liked Courier's ability to drag contacts into a transclusion, and Codex's focus on face to face collaboration. Very cool stuff, collaboration is totally the post-pc break we're waiting for.


The Raven @ 9:

There were some corporate labs that focused on research into the core technologies of the parent corporation: HP Labs (though the corporation rarely used what they produced), Wang Labs (ditto), and Tektronix Labs (the corporation actually got products out of the labs, but shut most of it down in the early '90s anyway). Seems like there's a fundamental disconnect between product management and lab management no matter how well focused the lab may be.

Re: Microsoft Surface
No matter how high the vapor pressure of the Microsoft announcement, there was one significant effect of Surface: it resulted in an active DIY movement. If you can find an old video projector lying around, you can build a multi-touch surface for under US $100, based on plans from a number of sources. A cheap projector might cost you $400 or so, so the cost will still be well below what a commercial system would cost (>US $2,000, at a guess).



Sounds intriguing. I'm not sure it's possible with e-ink as currently designed, though. As I understand it, e-ink screens are opaque, and this is why they can't be back-lit. The opaqueness is what gives it the good bright light (sunlight) capability - it reflects the light that comes in rather than trying to overpower it with a backlight like my PDA. The black/grey colours are also always in the screen, just not on top reflecting. Because of this, I'm not sure a transparent mode is possible.

I'd love to see it, but I think it'll need a lot of development. What we might see earlier is something with two screens, either book shaped like the Courier project or front and back - one e-ink for low power, one LCD for colour and quick response. I suspect that this would cost as much as two devices, though, which makes it pointless.


O hardware gurus: is there any reason an LCD couldn't be backlit by ambient light? By using a translucent back panel or similar? (I know the Citroen C4 does this for the LCD speedo, not sure how similar that would be to a laptop screen)


Chris @15: The problem is that an LCD panel is optically "thick"; it's got a lot of shutters and colour filters to get in the way of the light passing through it and ambient light is not intense enough to provide good levels of colour at the front of the screen. Ambient lighting is also variable depending on the environment whereas a backlight has a constant intensity that can be controlled and set by the user.

B/W screens and e-ink panels can use front-illumination and low-power backlighting as they are optically "thin"; there are no colour filters and usually only one shutter per pixel and the transparent conductive tracks running across the screen are a lot less light-sapping.

Backlights are getting better and consuming less power; newer displays have red, green and blue LEDs for backlighting rather than using white-emitting fluorescent units. This means that the panel isn't pumping a wide spectrum of coloured light through the pixel filters but narrower bands of colour which can be spectrally matched to the filters. It also simplifies the design of the next generation of high-gamut LCD panels on the drawing boards which have five or six colours per pixel (ROYGB(V)) rather than just three-colour RGB.


I must admit I see computing devices today as existing on a sliding scale; rather than as discrete items. So Courier, iTablet and so on are not new or revolutionary, rather they are attempts to find a sweet spot in the market. Maybe it's a sign of a mature market, people no longer buy a computing device because it does something they have never done before, rather they buy one that best fits their needs.

For instance, my personal needs lead me to a smartphone in my pocket (I don't need one, but baby computer + GPS + wifi + Android = toy for Dad), a netbook when I'm out and a moderately serious gaming rig at home. Yes, a laptop will do more but it's heavier to carry around all day and I don't need the extra oomph when I'm using it. Yes, a Mac is prettier but it won't play the games I want to play. And consoles don't do Empire: Total War.

Anyway, what I'm trying to get at is that what can be referred to as non-traditional selling points will become prominent. Rather than technology, performance, and new killer apps, it wll be battery life, size and ease of use. And so I probably won't get a tablet/ebook reader as I can do that on my phone well enough for the limited time I get an opportunity to, while a colleague of mine commutes a couple of hours a day on the train so he spends the time with a Kindle and loves it.


There are outdoor LCD displays already, engineered for the tablet and "toughbook" markets. The trade-off is in resolution; the display has to have tiny "skylight" cut-outs in it to allow sunlight to penetrate down to the reflectors, which makes the display coarser as naturally they can't have pixels overtop the tiny light-pores. Frankly, though, I wouldn't want one for a computer.

I see a lot of reference to Courier as an Apple "spoiler", but looking at the spec (twin 7" screens, touchscreen, lightweight) to me it looks like an outgrowth of their UMPC/Origami concept; I own one of those with a single 7" touchscreen, trade-paperback sized, less than a kilogram mass. (Note that Origami was announced back in '05 and released in spring of '06... predating the current version of the iTablet rumour.) I like the idea in general, with the only thing spoiling my interest is that it'll be running an embedded OS instead of a generalised one.

-- Steve


@Stephen: re /transparency of eInk/

AFAICR most eInk out there is essentially bi-state (flipping according to charge) and opaque.

I've got two thoughts on how to build my ink layer - both would need essentially transparent beads.

First thought would use a tri-phase panel where each pixel is repesented by some funky cubic face bead -- imagine a die. material is transparent; face (1) is white; face (6) is black; faces (2) through (5) are clear. Is this even possible? Don't know - but it should be :). It would be *very* cool, and this kind of control would allow color ink (similar to the 'color' LCD in the XO -- when transparent it would allow the underlying RGB to show through - simpler than trying to align a matrix of differenly colored 'beads' I would imagine)

An alternative is a face material that is essentially transluscent (white face is reflective from above - transluscent/transparent from below) When the ink is flipped to 'white' - the pixel will reflect ambient light - and also transmit the color of the layer 2 LCD. When flipped to 'black' the LCD would also be 'off' so the pixel would display as black. The bead would still need to be mostly transparent.

This is more feasible (from a materials perspective, and definitely from a control perspective) but greatly reduces the transmissivity of the panel for the LCD -- requiring higher power to get the same brightness... that might eliminate any savings from the use of the ink.


I still want one, tho'!


Another prototype toy (from Intel this time) involving multiple displays on portable kit:


I figured they used three small touchscreens because nobody was making a single long thin display that would fit in the same space. If the concept grows legs though it could spawn a market for such super-letterbox displays. I wonder if that rumbling noise is the photocopiers working overtime in Cupertino?


@13: "No matter how high the vapor pressure of the Microsoft announcement, there was one significant effect of Surface: it resulted in an active DIY movement. If you can find an old video projector lying around, you can build a multi-touch surface for under US $100, based on plans from a number of sources."

Those techniques predate MS Surface, but they were research curiosities and now hobbyist curiosities. We need someone to tackle the OS problem for augmented reality, and until that is done, I think all these techniques will remain curiosities. (It would also help if a couple of the patents on 3-D tracking technology would just go away, or at least have the grace to expire early.)


It's also about combining different technologies in one unit. That lets them try out all sorts of UI stuff. I think we could be seeing, as well, some of the effects of pushing beyond the mock-up to effective prototypes without expensive tooling.

I can see a different screen-bezel being turned out for a particular case on a one-off basis. That sort of 3D printer product where you can accept a loss of durability.


Wake me up when there's any evidence at all of Courier being anything other than research and/or vaporware and/or spoilerware. For instance, when there's an actual photograph or video of actual software running on an actual device. This "demo" video is like a sketch of a concept car: a couple of years before a concept car, five years before any of the ideas make it to an actual car in a real showroom. If Courier ever makes it that far, you can more-or-less guarantee that Microsoft will manage to bugger it up in the meantime. For instance, by making it run some descendants of some versions of Windows and Office.


Nick@23: "...you can more-or-less guarantee that Microsoft will manage to bugger it up in the meantime. For instance, by making it run some descendants of some versions of Windows and Office."

If they buggered it up that way, they'd have to build another bank vault to store all the money they'd make from it.

Office is the environment in business software, it's like desks and chairs and lights. It's what made Microsoft tens of billions of dollars of profits, maybe more, over the last 15 years or so. Getting it onto another platform means more sales, both of the platform because it runs Office and more sales of Office because it runs on that platform (see Crackberry for an example). The platform itself doesn't really matter to Microsoft, that's hardware and it isn't their "core competency" but Office in assorted incarnations IS Microsoft to a large extent. The Windows OS is only there in many cases to allow business users to run Office and nothing else on a PC.



B/W screens and e-ink panels can use front-illumination and low-power backlighting as they are optically "thin"; there are no colour filters and usually only one shutter per pixel and the transparent conductive tracks running across the screen are a lot less light-sapping.

Has anyone heard anything lately about the Plastic Logic ereader? It was originally marketed for release 2009, but that's been pushed back to early 2010. There have been all sorts of reasons offered as to why: a color option, a wireless option, superior refresh rates, etc. I do like the form, I have to say, and if they can include some sort of decent annotation option this would be ideal. I have about 5,000 pages of material that I either have to duplicate for the office or actually lug there, and I need a display screen a little bit larger than what is usually offered. Also, you can't really see it here, but the device is quite flexible; see this in the demos shown afterwords.


David@10: that's because the producer is one Raymond Stross, whom I believe is related.


I worked on a tablet app, so I was paying attention to tablet market progress - that is, it's disappointing market entry. Most consumers decided they preferred desktop or, increasingly, notebooks as their tradeoff points between usability and mobility. Most of the market was specialized apps like the one I worked on (if it'd ever shipped). Few, indeed, ran MS Office on them.

The recent love for netbooks, I think, is because they're cheap, especially as more and more of the market is the less-well-off, whom don't afford MS licenses. The second Google ad link for netbooks was walmart There's also a smallish 1st-world high-mobility market, and the continuation of specialized apps that like small form factors. Even if Courier ever launches, it's hardly going to be cheap, with the multiple screens and haptics and high software dev overhead.

Yeah, Sneddon, yeah, no doubt Ballmer WISHES it'd see that kind of success, even if it actually ships in time to be relevant. Poor Ballmer; he has to live in the real world instead of his dreams. Ballmer has exactly ONE success so far, XBox 360, and he got that by taking a HUGE loss, to try to put other console makers out of business. Maybe if he took enough of a loss on Courier to take it down to netbook levels....