Last week, as we intimated was about to happen, Microsoft Tablet PC product marketing manager Neil Laver popped round to display his wares. He very decently also proposes to let us play with one for a couple of days RSN, but an hour or so spent with him poking at the beast helped answer some questions, while raising quite a few more.

First of all, what is it? It is The Register's opinion that Microsoft itself has had a great deal of trouble figuring this one out, and still hasn't entirely decided. This however won't necessarily prove fatal to the project, because there are designs covering the obvious form factors in the works from numerous manufacturers, and the market (or possibly Intel) will therefore be able to decide. But the early Microsoft demo units as first Bill-ed at Comdex were slate-like Transmeta Crusoe units, whereas quite a bit of the action right now seems to be in the convertible area, as second Bill-ed at Comdex last year.

These seem largely Tabletised notebooks running Intel chips, the obvious exception from the big OEMs being the Compaq Evo, which will use TMTA and will look a bit like this. So, over a period of two years we would seem to have evolved from a situation where a Microsoft vision of an ultra-light, long battery life device has been convertibled by the major PC companies into something about the weight of a notebook (lightweight model) with about the same battery life. In mitigation, these villains appear to be praying for Banias, but they'll be punting out juice-monsters for at least a year till that happens.

Enough, however, of this chippery. That is not why I brought you here, and not what I was talking to Neil about. His demo unit is an Acer TM-100, for which you can find a handy run-down here. This is a convertible type unit, with a screen you can twirl around so that when it's closed the screen can face out. Thus, you have stylus control over your apps and the ability to write on an electronic notepad.

That is, if you want to write on a notepad that weighs several pounds and has a battery life of maybe three hours. We at The Register have some difficulty convincing ourselves we'd want to, and see this kind of hardware and form factor as likely to be a fatally munged compromise that doesn't really allow what you ought to be doing with something tagged 'tablet.' One of the points of Tablet PC, according to Microsoft from Bill downwards, is that a Tablet is a full-function Windows XP PC and then some. Neil meanwhile insists that this is what the users say they want; they want to be able to carry all of their stuff around and have it ready to hand.

We tried to dispute this, asking if this was what the IT management people said they'd let the users have/do, but he seemed sure that was the case, so you must meet a richer, more carefree, class of IT manager when you're with Microsoft. Anyway, setting aside the 'all your stuff/just enough stuff and a network' argument for the moment, the actual Tablet PC bit of Tablet PC is, we reckon, quite nice.

The Journal application, which is the one that turns your computer into the expensive pad of paper, is the bit that illustrates this. And the underlying argument is a pretty good one. Typing on a normal notebook computer gets in the way in meetings, and can even get you into trouble. So it's better to take handwritten notes unobtrusively. This being where The Register lives, we tend to agree. Even small keyboard devices intrude, because the other party is disturbed if your fingers are flying ('Cripes! What am I saying?') and disturbed if you're not doing anything at all ('Is this guy asleep or what?') But, erm, this also applies if they can see what you're writing on a large pad of paper (or screen). Unobtrusive is the reporter's notebook under the table where they can't see you're doodling, or the tape machine they've forgotten you switched on.

That said, Journal isn't a bad stab at it, would work on a convertible in the kind of meeting where you're all at least supposed to be on the same side, and might be OK for more confrontational engagements in one of the smaller form factors.

Did we mention handwriting recognition? Well we're not going to, much. Microsoft claims the recognition will be about the best available, but is downplaying this, rightly, we think. Recognition is less important than people tend to think, because in many cases you'll simply be referring to notes, which you can read just as you would if they were on paper, rather than converting them all, or trying to write whole finished documents and/or novels. So the model Microsoft is pushing is much more digital ink, with recognition as an extra you can get into if you like. Handwritten annotations to pre-existing documents are also a potential benefit, and again we feel we speak with some authority here.

Back in the days of good old typesetting it was quicker for a sub-editor to scrawl corrections onto a printout of a story (or even a typescript, as they were often in those days) than it is now to make the changes to an electronic document via a word processor. Even today marking corrections on printed proofs then having one person with a huge Mac do them is often more efficient than having huge Macs for half a dozen people, all of whom are going to have to load huge files in order to make the corrections. Plus, there's a massive version control advantage to having a single paper copy that must be signed off by all the relevant people.

No, don't press send, just wait until you've read the next bit before you fire off that email, because we know what you're going to say. Yes, the fact the Tablet is not paper does indeed change things, and could undermine what we've just said, depending. But digital paper is an interesting notion to kick around, and if it were executed properly within an organisation it could bring some of the control and organisational advantages of paper. Locked documents that revisers can only write on? Different coloured inks for different revisers? And there are plenty of other examples of how ink-enablement could add dimensions to applications.

Writing on the Acer was fairly comfortable, and it made a good stab at picking up my scrawl and figuring out what it said, even though I wasn't trying. Neil claims that working with the machine tends to make you improve your handwriting a tad, but as I said, we're not going to cover recognition much, in this pass at least.

One of the best and clearest examples he gave of where Tablets might find homes was medicine. Doctors have to move around a lot, they need to access records, and they're not necessarily going to be in front of a keyboard when that need arises. The UK National Health Service is allegedly interested (but not, we presume, half as interested as Microsoft is in the UK NHS), and Neil has a little doodled app that gives an idea of how a Tablet might be used in, say, casualty. Which conceptually works. You'd have a proper application developed that would allow you to start or call up a record, mark the problem areas on a chart, check various boxes linking to other people or services, scribble a couple of supplementary notes if you felt inclined, then press fire and it's off into the network, and treatment commences.

We'll own this is a highly imaginitive illustration of how the NHS works or is capable of working, but we can dream. And, uh, didn't we mention a network there? In which case, why are you carrying around all your records on a Windows XP PC that's just crying out to be stolen, what the blazes are you doing about privacy and confidentiality, and why the hell have you got a Windows XP PC in the first place, rather than a cheap thin client?

The thing is that vertical markets are best served by vertical devices, and the Tablet PC is essentially a general purpose PC with a particular vertical device (pen-ink, in this rev - voice seems to have gone rather quiet) stuck on. Sure, it's an application for a PC, but it could equally well be an application for a CE or even a Symbian device. These might also be general purpose devices in their own right, but they could be cheaper, less resource-hungry devices, with any general purpose functionality crippled as needed.

And other companies have been making this class of device for years. Why haven't they taken off? Well, they have taken off, it's just those companies are in vertical markets, so most of us don't notice them much. They have not taken off much in consumer markets, but the question here is whether that's because people don't want them or because they haven't been tried properly yet. Do people want a lightweight device with about ten hours battery life they can use with 802.11 to cruise the web, maybe with a little keyboard for light work, but you can skip the handwriting if you like for now? Don't know, but we do. Yes, and we realise that's a Psion netBook - sorry, we won't mention it again.

None of the Tablet PCs on the stocks fall into that category, not even the Crusoe prototype, which claims dismal endurance of four hours. Microsoft's nightmare product roadmap does have things that might qualify, but maybe it's whatever Mira's called today, or Xbox 2, or something. And in specialist markets it's CE, which is not Mira really, but hey, wouldn't it be cool if you could take your home Mira unit into work and... No, don't go there.

It's the Tablet PC that you're supposed to truck around with you, and when it comes down to it we reckon it's all Bill's fault. In his presentations on the subject he's positioned Tablet as a full desktop replacement you take with you everywhere, and pitched it as the product that will likely replace the conventional notebook, and then the desktop PC as well. It, and its successors, probably will, and the addition of writing, then recognition, then maybe voice will be significant benefits for the general-purpose PC platform as it is currently defined.

Who should be doing the defining is a separate argument we won't have here, as is the question of whether PC-centricity is increasingly absurd in an increasingly networked world. Suffice to say, for the moment, that Microsoft's determination to preserve and nurture the PC as the centre of all things is not, in our view, wholly positive. But it is nevertheless what has defined the script for the Tablet PC. ®

Further swotting:
Pen Computing has been doing an effective job on the Tablet PC for some time now. You can get its rundown on the software side of the Acer here, and an initial take on Tablet, together with a lot of useful pen computing history, here.