PC-oriented dealers and, to a lesser extent, manufacturers, will have some new software and hardware considerations to think about and plan for during 2000. Of course, most of the computer buying public will continue using Microsoft operating systems, but, as is evident at any Comdex show, it pays to watch trends in wireless devices, flat panels, communications technologies and the upgrade plans of the percentage of the market that use Macintosh or Linux, et al, as well by examining trends in these categories as well.
Legacy Free Computing
The iMac influence has strongly permeated the PC industry, as boldly coloured plastics, stylish cases and legacy-free architectures continue to demonstrate consumer appeal. However, legacy-free computing may also prove a hit at the corporate level, as new systems such as Compaq's iPaq computers, with their complete lack of serial and parallel ports, ISA and PCI slots, redefine the concept of low-cost, low-maintenance desktop workstations. The iPaq, like the iMac, uses the Universal Serial Bus as its primary method of input/output. Weighing in at 10 pounds and running Windows 2000, it does not include a built-in screen or floppy drive, but comes with 64MB of RAM, a 500 MHz Celeron processor, 3.2GB hard drive, Ethernet, sound and VGA out, for US$499. Additional storage devices, such as a CD-ROM, DVD, LS-120 floppy or a second hard drive can be added via a so-called "MultiBay." A Pentium III version and other options are available at extra cost.
And Compaq and Apple aren't the only ones straddling the line between PCs and internet appliances. HP, IBM, Gateway and Dell have all revealed plans to start selling smaller, stylized computers in the near future. Even AMD has apparently caught iMac fever and, despite the fact that it, along with many other major vendors declined to exhibit at Comdex Fall '99, showed a curvy translucent plastic PC called the "EasyNow" at a nearby hotel. And of course, Intel made a splash earlier in the year with its "EasyPC" initiative.
For those not ready for quite as drastic a step forward, there are new motherboards such as Abit's BE6-2 and BF6, with their single ISA slot and five or six PCI slots, respectively, or newer models from a variety of manufacturers based on Intel's newest i820 chipset. And, of course, the support of USB by all new Macs, Windows 98 and Windows 2000 (and, for those sticks-in-the-mud, Windows 95 OSR2 plus the USB supplement) also suggests that legacy-bound parallel-port scanners and serial port devices are soon to hit the bargain bins, with their IRQ hassles and other conflicts.
Next-gen Chipsets
Intel's Oct. '99 revision of the Pentium III architecture, adding full-speed cache and 133 MHz bus support, and the company's Nov. 15th announcement of the VC820 and CC820 motherboards (or "Desktop Boards," as they are now referred to in Intel's euphemistic parlance), ostensibly for the PII and PIII only, suggests that the "Camino" generation of motherboards will be the preferred option for those seeking a high-end PC in 2000, despite the market's so-far lacklustre response to the multimedia benefits of Streaming SIMD Extensions (SSE) that Intel persists in pushing as a way to "enable the Internet" -- instead of focusing on the games market, where they really matter most. But it will be interesting to see whether the next generation of Celeron chips, with support (whether officially, or via end-user overclocking) for a 100 MHz bus (or, perhaps the VIA Cyrix Celeron clone, with its support for a 133 MHz front side bus), again cannibalize PIII sales and hamper AMD's ability to effectively compete, as the 98/99 Celerons did to a significant extent.
But it's no surprise that the buying public will accept a product that does less, if it costs less. Take a look at the success of WinModems -- almost every PCI modem on the market at this point is a WinModem, despite the fact that we in the computer press have been gamely recommending against them for years. And the development community didn't seem to mind the architectural shenanigans that the MMX chips used to do their magic. However, there's a good chance that WinModems will become Linux friendly in Y2K -- after all, if the hackers can crack the DVD encryption, as they did in Nov. '99, writing a Linux driver for a software-based modem technology doesn't seem completely implausible -- at least until the buying public learns that solutions like IBM's MWAVE, Apple's Geoport Telecom Adapter and the "Windows only" modems of the last couple of years were all relatively troublesome and compatibility-challenged products, as many have already noticed in increasingly shunned parallel-port drives and scanners. And if a product that costs less does more, as AMD's mighty Athlon may again demonstrate, as it did in 1999, Intel may face some additional pressure.
We expect Intel's relentless push toward 700 MHz and faster CPUs to meet increasing consumer skepticism in the year ahead, at least with business and "average" entertainment-oriented PC users. While hardcore gamers will always benefit from a wicked-fast PC, at a certain point, when you're already getting 100 frames per second in Quake III and the billion-odd pixels-per-second fillrates of the Voodoo 5 or its kin, you start to look for other values, as automobile buyers did several decades ago.
Windows 2000 and Beyond
There's little doubt that Windows 2000 will be a big success. NT is a bear to configure, and Win98 is, in the minds of many, simply too flaky for serious business use. All it takes is the mainstream media to enthuse about Windows 2000's dramatically improved stability and the legions of users fed up with crashes in Windows 95 and 98 will give it at least a second thought. And, when hardware or software comes out -- as it inevitably will -- that performs better on Win2K, the desire for improved stability and convenience that we saw with Windows 95/98 will push many forward. And, it's safe to say that Windows 2000's plug-and-play support and the power-smart potential of ACPI (Advanced Configuration and Power Interface) features is a giant step forward from that of Windows NT, especially when coupled with a motherboard that supports the Instantly Available PC (Suspend-to-RAM) spec, such as those based on the i820 chipset.
Still, Giga Information Group, in a recent study, estimated that the average cost of upgrading to Windows 2000 Professional will be about US$1,640 per desktop system. This alone will restrict many companies on NT or Win98, at least until pain-versus-gain studies adequately demonstrate that this cost is more than offset by savings in training, productivity or uptime.
Given the similarly grim manual IRQ configuration procedures necessary back in the days of DOS and Windows 3.x, we were surprised how many people chose to stay with Windows 3.x, given Windows 95's improvements in reliability. With the high percentage of machines that were upgraded to 32 MB or more RAM due to sharply declining prices in RAM and hard disk prices during the first half of 1999, we still harbor a glimmer of hope that the even higher cost of upgrading to Windows 2000 estimated by Gartner Group is far too high.
Of course, it's likely that the ultimate success of Windows 2000 will be determined by the recommendations of IT staffers, who will hopefully find that their PC users actually don't like the endless fiddling that DOS-based versions of Windows seem to require. If they are told by a trusted source that their computers don't have to be as trouble-prone as they appear under Windows 9x, it would be my guess that many of these people will jump to Windows 2000.
Microsoft Millennium betas will probably appear as a consumer-purchasable "preview release" well before the final product's likely mid-2000 release, with the unified 9x/NT code-based "Neptune" following on in 2001. Microsoft has been seeding the early adopter markets aggressively with all of its recent OS and big-ticket releases, and it seems likely that the company will continue this trend in 2000. However, even if Millennium is derided as a cash grab in some quarters (and it will be!), it is clear that Apple has done well with its similarly minded Mac OS upgrades in the recent past, so there is every indication that this trend will occur with PC consumers as well.
Despite its advances, Millennium isn't looking like all that big a change from Windows 98. It will be interesting to see how many people downgrade from Windows 2000 Workstation to Millennium, due to compatibility or performance issues, or if, in fact, Microsoft even supports such as move. Thank goodness for products like Power Quest's Partition Magic 4.0, that can convert NTFS partitions to FAT16 or FAT32 and vice versa. Despite the fact that Microsoft has decided to support FAT32 in Windows 2000, there are certain functions that require NTFS, and we expect more than a few percentage of people to find, as we have, that Windows 2000's compatibility isn't all that Microsoft's early claims of its driver support as a "superset of Windows 98" to come up short. And, although Symantec has recently changed its mission statement to reflect its success as a provider of enterprise-wide security solutions, it is likely that subscription-oriented products -- probably sold primarily via the web -- that keep virus protection and other system-maintenance tasks up to date -- will prove to be a growing area in Y2K. More than application service provisions, which are predicated on more bandwidth (and patience) than most users currently have, security, anti-virus and other system-level driver software updates, such as those pioneered by technologies such as Oil Change, Symantec's Live Advisor, and Microsoft's own Windows update are likely to prove popular with reliability and convenience-seeking users in the year ahead.
The Linux Wildcard
And then there's Linux. While both Linux vendors and Microsoft have shown test results that prove, without a doubt, that their systems outperform similarly clocked systems running the other's OS in both stability and raw performance, we think that CorelLinux (based on the Debian 2.2.12 Linux kernel), with its Windows-friendly networking and utterly simple configuration procedure, is the Linux most likely to see massive gains in the year ahead. However, we think Linux's gains during 2000 will be hampered by the fractured market that already sees the distribution of Linux-based software and upgrades complicated by divergent schemas. In one corner, there are standards in RPM format for Red Hat users, in the other(s), "TARballs" and so on. Despite the efforts of Corel, with its Archive Administrator, and others to address this issue, it brings additional complexity to an OS already seen as too complicated for mass adoption. Still, with the increased emphasis on plug-and-play configuration and graphical sophistication seen in Corel's nascent Linux distribution and others of its ilk, Linux is almost certain to gain in popularity in the coming year (look for Transmeta, with its Jan. 19, 2000 "Crusoe" processor announcement to add another twist), and it is equally assured that its ease-of-use quotient will dramatically improve -- largely, we think, thanks to the emerging dominance of the KDE environment favoured by Red Hat and Corel.
We may also see even more blown-out-of-proportion Microsoft "FUD" stories about Linux's 'fatal flaws' than we saw during 1999, when Microsoft apparently noticed that Linux was a genuine threat and began generating dubious reports of Linux weaknesses. There are, however, some legitimate concerns about the quality of software and support for Linux, which already suffers from widely varying software and driver quality. Still, this doesn't diminish our feeling that retail-packaged Linux distributions and "how-to" books are going to be hot sellers in the year ahead.
Do or Die Time for AMD?
In hardware, we expect 2000 to be do-or-die time for AMD and/or VIA, perhaps the only Intel competitors that really matter anymore. Despite an unexpectedly profitable fourth quarter for AMD and delays in the i820, Intel's follow-up to the i810 "Whitney" chipset (with its weak i740 graphics architecture), the i820 will be the chipset that does for the PIII what the BX chipset was for the Celeron and PII -- enough of an improvement, with its UDMA66 and 133MHz bus architecture, that yet another round of motherboard upgrades, new RAM purchases and hard drive upgrades, will take place. And, despite the lower price and high-end performance of the Athlon and the technical merits of the forthcoming "Sledgehammer" (64-bit, 32-bit x86-compatible AMD processor), we expect the next-generation Celerons and price-competitive PIII chips to continue to result in more red ink in 2000 for AMD, despite its encouraging Q499 results. With that said, we're expecting prices on RDRAM (Direct RAMbus memory) to rapidly decline during the first half of 2000, approaching price parity with PC100 memory, rather than the 40 percent premium it currently commands.
It is unlikely that Apple's significant market gains during 1999 will approach anything resembling parity with the Wintel market in the year(s) ahead, but the company is building good products right now, and they were the stock market's darling during the final months of '99, even exceeding Microsoft's per-share stock value in November. But, when you consider that a relative unknown like Akamai, best known for its bandwidth-allocation and routing technologies (shown off to a reported 16 million online viewers during the Nov. 9 keynote at Apple's QuickTime Live! event), can achieve a market capitalization of US$15 billion dollars in a few weeks of public trading, it makes the sale of an old warhorse like Apple (with a market value of around $5B) seem a persistent likelihood.
Packet Writing
Another technology that finally caught on big-time in 1999 was CD-R technology (which we were predicting would soon dominate the market back in '97). It's safe to say that the next logical step in this progression is rewritable DVD. What's not as clear is why backup software that supports CD-R and CDRW is still so limited in availability and poor in quality. Is everyone happy with packet writing (the method of writing a compact disk that doesn't require the user to write the whole CD - or even an entire session - at once)? We're sure not. Although with packet-writing software such as Adaptec's Direct CD, one can simply 'Save' files to a CD-R, as if it were a floppy disk, the software isn't reliable enough for us, nor, do we suspect, even more forgiving users. And, unlike the competing PacketCD offering from CeQuadrat, Adaptec's software doesn't play well with others, often causing trouble if a CD-R recording package from a cometing vendor is installed on the system. And then there's the issue of market confusion caused by the competing standards of DVD-RAM and DVD+RW (set to be exacerbated by the fact the Hewlett-Packard had the bright idea back in June of pre-announcing a DVD+RW drive that would, upon its release in Dec., write discs completely incompatible with every drive currently on the market; an idea that the company thankfully came to its senses about and cancelled in Nov. '99). HP and other DVD+RW supporters now say their second-generation models will not have this problem and will support up to 4.7GB of data per disc.
Until the market decides on the writable DVD format(s) of choice, CD-R and CDRW remain the technologies most likely to put the long-suffering floppy out of its misery, despite the levy on recordable media in Canada expected, at this writing, to be in place by Jan. 2000. Things don't look so bright for Iomega, which lost bags of money in 1999, with its "Clik" drive efforts and the increasingly soft market for tape-backup solutions. Although Zip disk users are not as problem-plagued as Jaz owners or those who got suckered into buying a SyQuest product in 1999, 100 MB just isn't a lot of storage anymore, and cost-per megabyte ratios of Zip, Jaz and LS-120 discs just can't compete with CD-R media.
We're still waiting for a killer backup program of the quality and reliability of Ahead's Nero or CeQuadrat's WinOnCD, combined with the ease of use of Adaptec's Easy CD Creator, that makes CD-Rs easy enough for mere mortals to use as a complete and incremental backup solution. By the end of 2000, we think such a program will have been released, but at this point, it's not clear whether it will come from Adaptec or any of the other CD-R software developers, or one of the system-level backup experts such as Veritas/Seagate, Symantec, Power Quest, or another player. (Plextor, with its version of Ghost customized to write directly to CD, is on the right track.) In short, we think CD-R will be widely used as a backup method by this time next year, and writable DVD drives that can also record CDs will be the buzzword du jour in the next generation of products from JVC, Sony, HP and others.
With the growing emphasis on e-commerce at the websites of almost every manufacturer and dealer, 2000 could be the year that the traditional wholesale channel really start to hurt. And while rebates are currently all the rage, we think that this, too, will fade during 2000. If everyone is doing it, it's not an effective market differentiator anymore, and consumers will choose the discount at the cash register (or online), when given the choice.
The "PC Plus" Era
We're not expecting that 2000 will prove to be the year that we see the ubiquity of FireWire I/O and desktop digital video publishing as a "killer app," as apparently envisioned by Apple's Steve Jobs, who took the first step of releasing DV-capable, FireWire-equipped iMacs in Oct. 1999. But he was pretty much on the money with USB (and, for that matter, with coloured plastic!), and FireWire sure beats the pants off of SCSI in terms of ease of configuration, as USB proved superior to parallel and serial-based I/O. It's more likely that products conforming to Version 2.0 of the USB spec will hit the market in 2000, further establishing USB as a good enough spec for the PC-using populace. As USB and home networking technologies take hold, products such as digital cameras and camcorders, DVDs and other gadgets of the so-called "'PC plus' era are bound to boom, as well.
In handhelds, it's currently unclear as to whether support for Windows CE is dying (Compaq, Philips both appear to be fence-sitting), dead (Everex), or whether Microsoft will simply do what it always seems to do... just keep improving it until it gets things right. Nevertheless, the "instant on" and "always available" concept of Palm, Handspring, Psion and CE-based mini-portables has an obvious potential both as an organizational partner for both desktop and notebook users and, probably, more importantly, as a logical progression forward for wireless communications devices. However, the fact that Bluetooth devices don't currently safely coexist within an existing 802.11 RF network environment. is a real problem -- and one that clearly needs to be effectively addressed before Bluetooth solutions can be said to be truly seamless. It's safe to say that Palm didn't read the market correctly with its insistence on web pages specially coded to "clip" data out for display on the Palm VII; we think companies that get Web-on-cellphones technology, Wireless Markup Language (WML) solutions and wireless TCP/IP communications right will emerge as the leaders in 2000 and beyond, as 3G wireless communications devices start to emerge. It is clear that wireless devices already represent a major growth area in the portable computing, telephony, handheld and networking markets, as the R&D and investment efforts of leading companies in these areas show. Too bad North Americans will be among the last to benefit from the wireless connectivity speeds of 3G device -- already delivering up to 2.2Mbps in limited trials. For now, we're keeping an eye on the Yahoo-owned Online Anywhere, WML developers AvantGo, Digital Paths LLC, OmniSky, Riverbed, Spyglass and Wysdom Inc., and hardware developers Nokia, Qualcomm, Ericsson, Palm, and, yes, Microsoft.
Speaking of which...
Microsoft Exchange 2000, which should be out sometime in the Spring of 2000, will very likely be a successful product, and not just because Lotus' email efforts have lost some of their early steam. Exchange, as even the most skeptical Exchange haters will admit, improved dramatically between 5.0 and 5.5, and Exchange 2000 is another significant step forward -- and, because it requires Windows 2000 server, is currently perhaps the best example of a "killer app" for Win2K.
One area that is not, at this writing, clear, is how (if at all) the Microsoft antitrust trial will affect future Microsoft product releases. We happen to prefer Internet Explorer to Netscape, we like the way Office 2000 connects to the web, (try going to the Compaq FTP site and downloading a 'DOC' file with Internet Explorer to see some pretty impressive "DocObject" behavior), and we think the majority of users who now use IE do, too - and it's not Microsoft brainwashing making us think that way. What also isn't clear is whether Mozilla.org will ever get its act together and release a decent version of Netscape 5.0, now promised in "final" form in February, 2000. The company has already conceded that this release will provide "some" support for CSS 2, and limited HTML 4.0 support. Until both of the major browsers fully support DHTML and Cascading Style Sheets, web surfers will continue to experience glitches and unexpected display anomalies.
We also think the Office software wars are due to take a low-cost turn. If you hadn't noticed, Corel has been putting together quite a few deals recently that see its WordPerfect software bundled for free with various hardware configurations. The company already offers the Macintosh and Linux versions of the software for free on its web site. And then, on Nov. 8th, it announced a pilot project with FutureLink Corp. (http://www.futurelink.net) for a new online initiative – subscription pricing via the Internet. FutureLink, it seems, will remotely host WordPerfect Office 2000 so that FutureLink clients can purchase monthly rental licenses for the software.
And then, only one day after Corel announced its plans to offer Corel WordPerfect 2000 as a fee-based hosted (online) service, Microsoft announced similar plans for its Office 2000 Suite. Microsoft on Nov. 9, announced Microsoft Office Online, a new offering of its flagship Office product that enables the delivery of Office 2000 over the Internet. Office Online will, says the company, provide a choice for Office customers who prefer the benefits of centrally managed software (apparently over decent performance) and will be offered to small-business customers through select Canadian partners and Microsoft's bCentral Web services portal (the latter, only in the U.S. at present). As part of an Office Online pilot program, the company announced deals with a number of Office Online service partners, including services from FutureLink Corp. in both Canada and the U.S. and MTT Mpowered, a division of Aliant, in Canada.
And of course there's Sun, which bought Star Division for US$73.5 million, apparently enthused at the potential of its StarOffice software to turn into web applications-for-hire.
We expect to see many other products made available in this fashion in the years ahead, despite the potential for laggardly performance on even the fastest Internet connections.
NT
Already, Microsoft's web site is downplaying NT. Windows NT 4.0 will probably see a service pack or two beyond the current SP6 release, but we expect a replay of the scenario that happened with the 3.51 release when 4.0 started showing up. Despite all the claims that enterprise IT are slow to move and they don't need the latest and greatest, Microsoft's efforts for the new OS have focused primarily on the networked server and corporate desktop, and these people will benefit from its many improvements in web-serving, terminal services, virtual IP addressing and streaming media services.
Still, NT isn't going to disappear from the radar screen overnight. It remains the OS of choice for a vast number of enterprises running servers, as well as a number of professionals who use it (often not altogether successfully) for mission-critical tasks on the corporate desktop.
Among slow-moving corporate adopters, the relatively robust NT 4.0+ service packs alternative made a growing dent in the IS departments of companies that previously were or would have been well-served by Novell's NetWare. And, as device drivers become less of a problem and intranet-based workgroup environments become more common, and the inevitable service packs show up, moving to Windows 2000 won't seem like such a bleeding-edge step.
It's probably safe to say that NT will still be more popular than Windows 2000 in overall numbers, despite the fact that the Windows 2000 systems are likely to represent the bulk of new system and server installations. We also suspect that 2000 or 2001 will be a time of decision for fans of the Alpha chip. Microsoft already bagged NT development for the PPC and Sparc, with the coming generation of high-performance and high-availability Intel-based and Intel-compatible solutions, the onus is on the others to show a raison d'etre. In the past, non-Intel RISC processors consistently outperformed high-end Intel boxes (and still do, in some areas, as any Mac fan will tell you). But recent advances in Intel processors and super-high performance network configurations are changing that equation. We think the days are numbered for the Alpha version of Windows 2000. The NT market - indeed the 32-bit PC market in general - has proven to be remarkably unadventurous. So, look for OS strategies to cater to this mentality, particularly, as the next generation of 64-bit Intel processors further change the rules of the game.
Considering the remarkably low price of today's CPUs, we confess to being surprised at how few people seem to exploit NT's support for multiple CPUs -- a capability well-suited to the demands of web serving, number-crunching, graphics and multimedia tasks. Still, we persist in believing that multiple CPUs will grow in popularity over the next few years, as the power of NT comes to the "average enthusiast's" desktop, thanks to Windows 2000, or alternative OSes such as Linux or BeOS. We even expect Apple to get its act together with regard to multi-processor support in 2000.
A bigger paradigm shift may come about as a result of the Open Source movement, and recent announcements such as the one by Sun, that it would release its MicroSparc design for free, or the announcement from IBM that it was freely releasing the hardware design for Common Hardware Reference Platform (a.k.a. PowerPC Platform) -- a move, in turn, caused by the fact that Motorola and IBM backed away from their early support of running Windows NT on this platform. Free, open hardware designs could lead to the ability to emulate entire systems on a massively connected, asymmetric-processing "virtual supercomputer" running a CPU cycle-sharing program not unlike that which, in 1999, popularized the Search for Extra Terrestrial Intelligence thought the SETI@Home project (currently dominated by Team Art Bell).
Although Apple is enjoying something of a heyday, there are more than a few Mac development shops packing their bags and moving on, too. The market moves faster now than it did a few years ago, and simple ports of a Windows application -- or, worse, a feature-incomplete effort, such as that which was shaping up for the Mac version of Half Life before that game got canned in late '99) are less acceptable than ever before.
Like Hollywood and the music industry, the computer business doesn't seem to want to take many risks right now unless they are fueled by Internet money. In the rush to go public and/or exploit the stock market's enthusiasm for all things ending with "dot com," high-tech venture capitalists and privately held companies alike are more interested in showing rapid growth than they are in investing in R&D, or worse, that boring old bricks-and-mortar retail market. Why focus on retail, with a mere 29% year over year growth, as Symantec's Chris Monette recently noted, when web-based, enterprise-oriented subscription offerings can provide over ten times the growth?
Indeed, the Internet economy is something of a paradox. On one hand, software development, product marketing and corporate America in general became a lot more bottom-line-driven in 1999, and 2000, we are likely will see a lot of too-little, too-late projects get their plugs pulled, and a lot of companies falter, as Packard Bell, Netscape, and, to a lesser extent, Compaq and Toshiba did during '99. On the other hand, we see players like Amazon.com, who've seldom shown more than a hint of profitability, but dominate their categories and compete with a vengeance. The market suspends disbelief and floated the stock bubble for enough of 1999 to at least suggest that the old rules don't necessarily apply. But woe be unto those who are not in the top three or four products in their respective categories. In the Internet economy, the winners don't have to make a profit, but the losers do.
Of course, we'd love to be wrong about Netscape. To mix a metaphor, Microsoft needs someone breathing down its neck to spur it on. And despite the U.S. Justice Dept.'s Nov. 99 pronouncement that Microsoft possesses "monopoly power," and whether or not you believe that it beat Netscape at the market-share game by using an unfair advantage, most people agree that IE is a more advanced browser.
And despite whatever you might think about Apple, it seems clear that Microsoft was more concerned about Java, Netscape and, most recently, Linux, during 1999 than whether Windows was as good as the Mac OS.
Utilities
We don't expect the "bricks and mortar" retail software sales dynamic to change dramatically during 2000, although it is clear that e-commerce sales will start to make a dent in the bottom lines of those who aren't competitively priced. As we saw in the top 10 lists during much of 1999, operating system upgrades, anti-virus tools, MS Office upgrades, games and 'fix-it' software that includes Norton Utilities, SystemWorks and so on, will continue to top the bestseller retail lists. But the real area to watch is the enterprise site-licensing market. This is enjoying about ten times the growth of the retail sector, and, more often than not, these sales are skipping the channel entirely.
Portables
We also think that notebooks are due for their biggest year ever (unlike Windows CE devices, which are fizzling) -- but not at retail. We'd put the money on lightweight Celeron- or PIII-based units (preferably with a built-in CD or DVD drive) with USB, 64MB of RAM, a 5GB+ drive, and an XGA active-matrix screen, running Windows 98. Notebook users, who nearly all want to surf the 'Net and access the company network, will also be prime candidates for an Ethernet PC Card and a high-speed Internet sales pitch. The price-point for full-featured name brand notebooks needs to drop well below the $3,000 and up the "good brands" currently cost, however, before people will pass up those prices on lesser-known mail-order specials.
Of course, just because certain industry analysts don't see a growing market for this-or-that technology doesn't mean that you can't make money from it.
In this business, focus is all-important and, as the vendors out there still selling and servicing niche products will attest, it doesn't really matter what you pick as your niche (in fact, picking a non-mainstream market segment is often a decided advantage!), as long as you are completely committed to it.
It's the Field of Dreams Principle: if you build it, they will come.
Post new comment