This article was published in PC Plus issue 123 (Jan 97), and is reproduced here for information purposes only. This is the original copy which was sent to the magazine, not the subbed version which appeared on the page. |
Paul Stephens column, PC Plus Jan 97 (on sale Dec 96).In this one I wasn't too impressed by the Microsoft/Intel Net PC specification, and made some predictions which - amazingly - seem to have come true!
|
|
|
[BEGINS]
<Callout> Microsoft’s reaction to seeing Notes on the Net horizon must be roughly that of a vampire to a lorry load of garlic. <Body> That billion-dollar double act, Microsoft and Intel, have been at the standards-setting game again, this time releasing a draft hardware specification for a ‘Net PC’ platform. This is a bit of a cheek, as the Net PC (or ‘thin client’) was dreamed up last year by a group of desperate software and semiconductor vendors as a way of reducing Microsoft and Intel’s domination of desktop computing. Still, when you’re trying to stay dominant by telling everyone else what to do, cheek is half the battle. Last year’s Net PCs were low-cost, ran downloaded Java applications and didn’t necessarily use Windows or an Intel processor. Predictably, Microsoft and Intel’s version is rather different. For a start it’s definitely a Windows platform, and definitely an Intel-powered one, too. It has to be, since the Windows they’re talking about - the rickety but big-selling 95 version - won’t run on anything else. There’s nothing necessarily wrong with that - after all, most people buy ‘Wintel’ PCs already, and that’s what all the software is written for. There’s nothing necessarily wrong with the detailed specifications, either - or not at first glance, anyway. The minimum hardware platform for this ‘low TCO’ (Total Cost of Ownership) machine includes the following: 100MHZ Pentium processor, 16MB RAM, hard disk (for cacheing), 256 colour display, sound adaptor and either a local network interface (Ethernet, Token Ring) or remote communications connector (modem/ISDN etc). All this sounds remarkably like today’s typical (if slightly under-specified) desktop PC, minus the CD-ROM drive which you arguably wouldn’t need anyway as you'd be downloading everything from the Net. The problem lies with the final item on the hardware list. As a requirement of the Net PC specification, the machine must have ‘no end-user expansion slots’. In other words, what you get is what you’re stuck with, as there’s no way you can add whatever new type of device might suddenly become next year’s ‘must have’ peripheral. On that basis it’s no wonder that hardware vendors such as Dell, Compaq, DEC, HP and Packard Bell are lining up to support the Net PC initiative. Any vendor in its right mind would love to sell machines like these - meaty enough to make a few bucks on, and non-upgradeable enough to be obsolete within a year when a compelling piece of new hardware technology pops up. These machines aren’t the thin client devices proposed last year by Oracle and others; they’re simply today’s desktops with their major advantage - expandability - taken out. If I was a corporate buyer, I’d go for the rest of the Net PC package, but insist that my supplier put a slot or two in as insurance. If I was a private buyer putting down hard-earned cash I’d be even more insistent, as I’d be far more likely to actually plug an extra bit of kit into my single machine than the average corporate would be to upgrade a couple of thousand. My guess is that there’ll be plenty of vendors willing to go along with this, and that the ‘true’ Net PC will end up as marginal a product as its virtually identical predecessor, the diskless workstation. That won’t necessarily upset Microsoft or Intel though, as the NPC will still have done its job, killing off the alternative contenders and keeping the desktop a Wintel zone. SURRENDER This surrender seems slightly premature, given that Navigator’s still the most widely-used Web browser. However, with Microsoft giving its Internet Explorer away, tying it closely to the forthcoming Office 97 and embedding it right into the next release of the Win 95 desktop, NetScape is really bowing to the inevitable. That’s a smarter strategy than simply watching your market share plummet, but a shame nevertheless, since it will give a clear run to Microsoft’s over-complex ActiveX (alias OLE) controls and the nasty Visual Basic language that drives them. Meanwhile if NetScape can’t handle Microsoft on the client side, quite how it thinks it’s going to find a way round Windows NT, with its integrated Web server and close-couple development tools, is anyone’s guess. It seems that the Microsoft of the Net is going to be Microsoft after all. Or at least it would, except for some rumblings from the direction of Lotus. Its people are going round making the not unreasonable point that Notes, a system which specialises in secure, replicable storage of rich data types, is the ideal thing for providing what commercial Web and Intranet sites need most, namely secure, replicable storage of rich data types. It’s produced some software too, called Domino, which acts as an interface layer between Notes databases and Web servers, allowing you to create transactional applications which run through ordinary Web browsers. The effect is not unlike the clever dynamic HTML pages you can build using Microsoft’s FrontPage designer, except that users can type their credit card numbers into them and feel confident they won’t leak out the other end. All this is probably enough to switch off the party mood in Seattle. Notes has been something of a bête noir to Microsoft folk since the heyday of groupware (a full two years ago), when the threat of it forced them to make some credibility-sapping claims about their Exchange messaging system and Cairo release of NT. The company still doesn’t have a real answer to Notes (partly, it must be said, because Notes poses such an amorphous question), and their reaction to its appearance on the Net horizon must be roughly that of a vampire to a lorry load of garlic. Can Notes thwart Microsoft’s Web-dominance ambitions? It certainly stopped it from defining groupware as a Windows-based concept, and it could just do the same for the commercial Web. It’s early days yet, but watch this space.[ENDS] (C) Paul Stephens 1996. All rights reserved.
|
||
