Crawford's Corner

Citation

Crawford, W. (1999), "Crawford's Corner", Library Hi Tech News, Vol. 16 No. 12. https://doi.org/10.1108/lhtn.1999.23916lab.001

Crawford's Corner

Perspective

New Narrative Forms on the Web

Michael J. Miller is PC Magazine's editor-in-chief, and his one-page lead essays are frequently thoughtful and enlightening. In the July 1999 issue, he discusses "The Web as a new art form" based on his experience at a Brown University forum.

On one hand, Miller saw the promise: "New technologies now offer the possibility of changing methods of telling stories and writing fiction and poetry." On the other, the single pull quote notes the reality: "Hypertext-based fiction and poetry may reinvent creative writing, but they won't replace novels or poems."

Miller sees that the new forms complement current forms rather than replacing them. For writers who can cope with the demands, hypertext-based creative writing could open some remarkable new vistas ­ but as a new kind of creation, not a replacement. I recently read Hamlet on the Holodeck, and was saddened to realize that the author really did seem to see computer-mediated creativity as the only kind of creativity worth pursuing. Not true, and a stance that endangers the future of computer-assisted and computer-mediated creativity: expect too much, denigrate "the competition," and you're setting up for a fall.

What Works ­ and What Doesn't?

What new forms will emerge ­ either directly on the Internet or using hybrid media? I don't believe we'll know until we see them. It's fair to assume that there will be many experiments that fail ­ and that there will be techniques that could work undermined by poor execution. Effective creativity requires ideas (including dreams), a means of expressing those ideas, and the talent to use that means. The Web may offer new means of expression but it doesn't provide ideas or talent.

We're already seeing a variety of trivial Internet-enabled creations. Some of these are akin to the old picture books with split pages: you could make your own strange creatures, just as you can make your own stories or songs or whatever on the Web. I find it hard to think of these as new forms of creativity; they're amusements, but no more than that.

I've heard of Web-based "stories" where the user is encouraged to follow his or her own path through a network of paragraphs or sections, eventually (maybe) arriving at an ending. That doesn't even require the Internet or hyperlinks; there have been books whose authors encouraged readers to choose chapters at random. Here again, I find such attempts more interesting as virtual parlor games than as creative acts; without some order, there's no narrative flow.

It's interesting to consider the ways that some traditional forms incorporate hyperlinks. Of the many attempted Web magazines (mostly failed), one of the most successful is Microsoft's Slate. My possibly-faulty remembrance is that Slate used to incorporate hyperlinks heavily within individual stories. That isn't usually the case anymore. Many stories end with clusters of links for those who wish to find out more about some aspect of the story, but links within stories usually do nothing more than provide in-place footnotes (or brief expansions of specific points). On the other hand, Suck's daily essays are peppered with hyperlinks ­ and, in my experience, those hyperlinks disrupt whatever narrative flow Suck has. Most of the heavily-experimental Webzines that I've seen tend toward the "peppered with hyperlinks" model ­ and most of them fail to convey messages very well. They usually fail, period.

Hyperlinks aren't unique to the Web. Many CD-ROMs incorporate hyperlinks to provide definitions or branch to related material. For purely informational works, this can be very effective; where learning or understanding is involved, it can be disruptive. Some CD-ROMs and quite a few Web sites deal with this by using popup windows (or smaller secondary Web browser windows): you can view the "interruption" but you're still clearly grounded in the overall article or discussion.

True hypertext fiction or narrative is tough. It requires new ways of thinking, new ways of organizing. It's hard to avoid sound-bite narration, turning an overall story into a linked network of mini-stories. There's nothing wrong with sound-bite narrative, but it can rarely have the sweep or effectiveness of fully organized narrative.

These are musings. I believe we'll see interesting and, eventually, important new forms emerge. The best CD-ROM products use that carrier to offer features that books and videos can't. The best Web sites do the same ­ and, so far, in much the same ways as CD-ROM.

Will there be a Great Hypertext Novel? Will there be a hundred of them, each using the Internet in diverse, creative ways? I hope so ­ not to replace traditional narrative, but to tell new stories in interesting new ways. I also hope I'll be open-minded enough to recognize such triumphs as they occur ­ and that libraries will find effective ways to incorporate these new forms.

Press Watch

Howard, B. "Another kind of drive," PC Magazine, Vol. 18 No. 15, p. 97.

Much of this is a fairly cute column about cars and the dozens of microprocessors that help to run them. It's the last two paragraphs that get interesting. Specifically, the last sentence of one paragraph and the first two sentences of the final paragraph:

"We're also on the verge of active distance control on freeways, and we're within a decade of licking the technical problems that hinder hands-off driving. The 'complexity' of electronics ­ multiplexing control signals over a single wire rather than running hundreds of discrete wires ­ can actually make cars simpler, safer (ever drift across traffic lanes trying to tune the radio?), and eventually, cheaper. It's inevitable."

Whoops. It's inevitable: always a good reason to think critically about the argument being stated. It's probably true that multiplexing control signals will result in simpler wiring harnesses and reduced costs ­ although I'd sure like to see redundant wires, thank you, if that "one wire" is controlling everything the car does. As for making the car cheaper ­ well, the electronics in cars have gotten steadily cheaper over the years, but about the best it's done for the cars themselves is to stabilize prices. That's not the amusing part, however.

The amusing part is seeing Popular Science predictions from the 1960s repeated once again: active distance control on freeways and, a few years later, hands-off driving. But the way Howard has worded it saves him: we should be able to lick the technical problems. Unfortunately, the technical problems don't have a whole lot to do with implementing hands-off driving (or even making distance control work). For distance control, you have to convince freeway yahoos that they should allow their cars to not tailgate the cars in front of them. Well, why not? We've all been convinced to drive appropriately small high-mileage cars ­ haven't we?

Hands-off driving involves a whole host of non-technical issues, most of which require massive government funding and, eventually, some form of regulation. I don't see how any car on a stretch of highway can run hands-off unless every car runs hands-off ­ and until there's a network of roads at least comparable to the interstate system with the embedded sensors and other needed agents, hands-off driving won't amount to much.

What really makes this amusing is that Howard's fellow columnists in PC, PC/Computing, and most other technology magazines are almost unanimously semi-libertarians, opposed to most taxation, most public agencies, most government ­ and particularly opposed to "taxing the Internet," even though sales taxes pay for so much road work and other governmental services. It's possible that we'll see some hands-off driving experiments in some European countries with more pro-governmental attitudes, but when it comes to the United States, I don't see this happening any time soon. Which is sort of a shame.

Fallows, J. (1999), "The slow march of history," The Industry Standard, Vol. 2 No. 24, p. 24.

Fallows initiates a new fortnightly column with this piece, and it should be worth watching. He has been a journalist for most of the last 20 years, just finishing an odd six-month project to develop "better software for writers" at an unnamed company that can only be Microsoft. His theme for this issue is unusual, given The Industry Standard's role as "the newsmagazine of the Internet economy." To wit, "the Net has hardly changed anything."

He believes that future historians will identify two technologies that revolutionized twentieth-century life: electricity and the telephone. Third in importance would be the set of medical and agricultural developments that have roughly doubled human life expectancy worldwide. After that? The automobile, airplane, "secrets of the atom," and computer. Fallows grants that the Internet (and other communications technologies) may change the twenty-first century to a similar degree, but it's a little premature to celebrate that transformation.

He notes four areas where he believes the success of the Internet has been exaggerated: the importance of where you live, "virtual communities," the politics of tyranny, and each person's ability to effectively be his or her own editor. The first one is particularly ironic: if workers in the information economy can work equally well from any location, why are Silicon Valley home prices so high (and still rising)?

Dvorak, J. (1999), "The internet bubble," PC Magazine, Vol. 18 No. 16, p. 87.

I'm not exactly picking on John Dvorak this time around (that's too easy), because I'd like to think he's right ­ but I'm not sure I do. His thesis is this: "It's the information, stupid. I'll argue this until I'm blue in the face: The Net is best for information gathering and dissemination, not e-commerce. Content, in fact, is king."

There are at least three problems with his thesis and the exposition. The first one comes along just a few sentences later: "Nobody would argue that TV's primary mechanism is anything other than content delivery." Really? I've been reading way too many books on media lately (preparing for my own project), and most of them do exactly that. The most common argument is that TV's primary role is advertising, with the so-called content being nothing more than a wrapper around the ads. Some posit something even more sinister: that TV's only role is to hypnotize us all into being obedient consumers.

I may not agree with those writers, but I'm not sure I'd agree with Dvorak about TV. When it comes to the Internet, I'm increasingly frustrated with any sentence along the lines of "The Internet is all about..." or "The Internet is primarily for..." But then, I'm not fond of statements like "Libraries are for ..." followed by a single purpose either.

Is the Internet primarily for information gathering and dissemination? I don't believe so. In terms of traffic, e-mail is still king: that's communication (of a sort), but not necessarily information dissemination. Dvorak says "the Web is a form of publishing that anyone can do," but it's an odd form of publishing at best. The Industry Standard and half a dozen other magazines seem to assert that the Internet is about business, as do most industry pundits. I think they're wrong as well. The Internet is a complex transmission system that carries many different kinds of stuff; I don't think it's "about" any one particular kind of stuff. For that matter, the Web isn't all there is to the Internet.

The Net is used for many purposes; no single purpose does (or should) dominate to the exclusion of all others.

Slater, M. (1999), "The right speed for you," PC Magazine, Vol. 18 No. 16, pp. 100-112.

What's the best CPU, and how much speed do you need? If you're in the Windows world, it depends on your uses for a computer. This article includes performance tests on the range of current "x86" CPUs (Intel and compatibles) and suggests appropriate choices for various kinds of users. One key finding is that, for the majority of users, the hottest CPU may not be the way to go. You're better off using your money for a larger display, faster graphics accelerator, more video RAM, more RAM in general, a larger and faster hard disk, and even a hardware DVD decoder.

Intel won't much care for this article. The editors evaluate Intel's advertising claims that the PIII enhances your Internet experience ­ and basically conclude that it's (at best) minimally true. As noted elsewhere, performance tests show that AMD's Athlon outperforms Intel's Pentium III, offering better performance at a lower price. And for most of us, one of Intel's cheapest chips (say a Celeron-400 or Celeron-466) offers more than enough power and costs the computer maker $500 less than the high-end Pentium-III chips.

Ad and Platform Watch

MacSnobbery?

June 1999 Macworld, p. 31: "Just like you, it won't work on a PC." It's an ad for the TelePort Internal 56K for the Power Macintosh G3. ("Internal 56" what? Presumably, we all know it's a modem.) Then: "At Global Village® we refuse to make products for the PC. In fact, all we do is make modems and modem software for the Mac."

Hmm. Should 3Com/U.S. Robotics run ads saying, "We make more modems for more kinds of computers than anyone ­ so we do it better and for less"? They do, and their Mac modems are considerably less expensive than those from Global Village ­ but they don't have that Mac-only cachet. Surely that's worth the extra bucks, even if Mac-only companies have suffered when there's open competition.

The bottom of the ad shows the chains of Mac dealers who stock Global Village: CompUSA, MacConnection, The Mac Zone, MacWarehouse, and Micro Center. You know that the first and last of those are primarily Windows operations. As for the three "Mac" places: MacConnection is a division of PC Connection, The Mac Zone is essentially the same company as PC Zone, and MacWarehouse has the same street address as MicroWarehouse, a PC operation. Gasp: none of these chains is Mac-only.

Encouraging Insularity

Here's the entire text in a two-full-page ad in the 8 June 1999 PC Magazine and elsewhere:

• •

  In the upper left hand corner, in scrawled italics: "Because great minds think a lot."

• •

  In the center of the right-hand page, in large letters as part of an odd collage: "You're 100 per cent technology. Why go to a bookstore that's 99.3 per cent everything else?"

• •

  Below that: "fatbrain.com {*}" (and, in tiny type on the left page, the usual copyright and trademark claims).

The small Computer Literacy bookstore chain was too specialized for my taste, particularly given the extensive computer-related collections at most bookstores around these parts. (I'd guess that Printers Inc. is about 20 to 30 per cent technology, Tower Books perhaps 10 to 15 per cent, and Stacey's in Palo Alto probably 20 per cent or more. I don't know of any general-interest bookstore in this area with only 0.7 per cent of its stock related to technology; certainly not Barnes & Noble.)

Computer Literacy went for the ".com" cachet and changed its name to fatbrain.com. Now we get print ads celebrating the idea that Silicon Valley folks (and those who want to be like them) should avoid books that aren't about technology (by which they mean information technology, incidentally: no car repair manuals here). Travel? Wine? Art? Fiction? "You're 100% technology."

How sad.

"We finally got it right"

Two-page ad, July 1999 Macworld. The first page has one word and one image. The image is an iMac. The word is "jet." Second page: the box for Microsoft Office 98 Macintosh Edition, the big word "fuel," and a few brief paragraphs of text. Here's the fourth paragraph in its entirety:

"Best of all, Office 98 was built from the ground up for the Mac. From the interface, to the icons, to the sophisticated use, even the experts agree we finally got it right. Which means with Office 98 you can make good on all the reasons you bought that powerful iMac in the first place. To really go places. Faster and more easily than ever before."

After the last "paragraph" ("Microsoft Office 98 Macintosh Edition. Prepare for take-off."), there's a box with the following: "This software suite demonstrates that, for the first time in its history, Microsoft truly understands what 'Macintosh elegance' means. ­ David Pogue, Macworld." And, below that, a logo indicating that Office 98 won Macworld's 1998 Eddy for Software Product of the Year.

I find this all a little bemusing. Many Mac owners regard Bill Gates as Satan Incarnate, but Microsoft applications have dominated the Macintosh for more than a decade. Until recently, Word and Excel were even more dominant on the Mac than on Intel-based PCs (that may still be true). And now, Microsoft is advertising that it took this long to "finally get it right."

There's little doubt that Office 98 is "built from the ground up" for the Mac. Windows owners will be aware that the Mac edition is the only edition: Office for Windows has gone directly from Office 97 to Office 2000.

Is Office 98 the most "Mac-like" set of applications Microsoft has ever produced? I don't doubt it, and as a Windows user I'll never know. Is this ad some bizarre combination of boasting and admitting that Microsoft has never understood a market that it's nonetheless managed to dominate? Apparently.

Working together: MacDrive 98 3.0

Macs do a fair job of reading PC-formatted diskettes. What about the other direction? Two competitors offer Windows software to open Mac files, and both now support HFS+, the latest Mac file system. The October 1999 Macworld reviews the $70 MacDrive 98 3.0, which has some slight advantages over MacOpener.

Once either program is installed, Mac diskettes work pretty much the same as PC diskettes, and the same goes for Zip, Jaz, CD, and even external disks. If you work in a mixed Mac-Windows environment, MacDrive seems like an easy way to reduce cross-platform conflicts. It won't work on Windows 3.x ­ but you really should have moved to Windows 95, 98, or NT by now!

Perspective

Windows 98 Utilities Redux

The lead item in Crawford's Corner for July 1999 concerned my early experiences with a new computer, Windows 98, and utilities for Windows 98. The two utilities I discussed were Mijenix PowerDesk Utilities 98 and Norton SystemWorks. Then, in September 1999's "Product Watch," I noted a rave review in PC Magazine for the new Fix-It Utilities 98 from Mijenix. There was an error in that note: the name is Fix-It Utilities 99, not 98.

At the end of that note I said, "I've been using Norton SystemWorks happily enough, particularly since it includes CleanSweep. But given Mendelson's enthusiasm, I may just try this alternative. Since I haven't, I'm not endorsing it ­ yet." A little parsing of those statements may be in order.

"Happily enough" is about as enthusiastic as I can be for SystemWorks. I've never been a devoted Norton fan. Indeed, I only started using Norton tools after Symantec purchased Central Point and either absorbed or discontinued the PC Tools competitors ­ which I always preferred. To my mind, the highlight of SystemWorks is CleanSweep, which Symantec purchased. SystemWorks in general isn't terribly well integrated, I've never been all that fond of the design of Norton Utilities and Norton CrashGuard seems to cause more crashes than it corrects. The system monitors seem to chew up a lot of resources and WinDoctor seems to find problems at the most inopportune times, asking you to interrupt your writing or database work in order to correct bad registry entries that happened hours or days ago. I like much of what Norton SystemWorks does and consider it a bargain, but certainly not the best of all possible utility packages.

Installing Fix-It 99

After preparing Crawford's Corner for September 1999, I found two plausible choices. I could upgrade to the new 2.0 version or I could buy Fix-It. SystemWorks 2.0 apparently eliminates one of SystemWorks' most annoying features: WinDoctor runs at startup, then gets out of the way. There are other improvements, most of them fairly minor. Since I already owned a Mijenix program, Fix-It turned out to be cheaper than upgrading to SystemWorks 2.0; I decided to give it a try.

The most unnerving aspect of installing Fix-It 99 was uninstalling Norton SystemWorks. That may not have been mandatory, but seemed reasonable. Competing real-time utilities would almost certainly conflict and would certainly use even more system resources. (Both program manuals warn against running competing utilities in some cases.) Fortunately, Norton's uninstall works fairly well and lets you choose which elements to remove ­ an important point, as you'll see shortly.

I set up Fix-It using the recommended options for real-time and startup operations, then ran an initial set of tests and operations. As with any set of utilities, it will take a while to see how well Fix-It performs overall ­ and, ideally, I won't see CrashProof (Fix-It's crash protection) functioning very often.

The agonizing part of any new Windows utility package is defragmenting the hard disk for the first time ­ and that's more true with Windows 98, where intelligent file organization is the rule. Every vendor has a different idea of how disks should be organized; the first defrag run means moving from one method to another. The process took about half an hour on my 10GB hard disk, but that's a one-time problem.

Clear Differences

Is Fix-It better or worse than Norton SystemWorks? As is frequently the case, the best answer to that multiple-choice question is "Yes."

Yes it's better, for at least the following reasons:

• •

  It comes with PowerDesk Utilities, a first-rate file manager and Explorer replacement that has performed beautifully ever since I started using it. See my July 1999 commentary for more comments.

• •

  Fix-It itself offers a clearer and more coherent interface than SystemWorks.

• •

  Utilities seem to be organized better and offer clearer explanations of what they're doing and your options. The Registry cleanup routine is much more understandable than WinDoctor, and I found similar improvements elsewhere.

• •

  Fix-It's real-time modules place lighter burdens on the CPU and operating system than those in SystemWorks. One difference is clear from the system monitors (available in both utility packages). With Norton SystemWorks running, my computer had 44 program threads operating before any applications were opened. With Fix-It, even with Norton's CleanSweep still operating, I count 36: a significant reduction in the number of background tasks. There is no equivalent to SystemWorks' continuous WinDoctor and drive checking: Fix-It (rightly, in my opinion) regards these as periodic maintenance tasks. (One magazine's testing showed SystemWorks chewing up 20 per cent of computer power while Fix-It uses about 5 per cent.)

• •

  Fix-It includes a nice set of modules to customize Windows appearance and operation. Some of these functions are available elsewhere (particularly in Microsoft's "PowerToys"), but others require direct Registry editing. The Fix-It module is clear and offers good help. Among other options, Fix-It makes it trivial to add QuickView to all file right-click menus.

• •

  You can run a single-click "FixWizard" to do a full-scale system tuneup ­ and there's a single Undo that will reverse decisions across all of Fix-It's functions. (I mistakenly accepted all the "caution" boxes in one Registry checkup and had to undo the changes: the action was clean and nearly instantaneous.)

• •

  The Fix-It CD includes versions for Windows 95, Windows 98, and Windows NT ­ and it's not easy to find good utilities for Windows NT.

  Yes it's worse, for at least the following reasons:

• •

  Fix-It doesn't include a competitor to CleanSweep. While well-designed Windows98 applications don't need an uninstall monitor (Add-Remove Programs works properly), I'd be loath to do without such a monitor when installing and removing CD-ROMs. I left Norton CleanSweep running: there's no reason to believe it will interfere with Fix-It.

• •

  Fix-It doesn't add its own tab to the Properties display for hard disks (as does SystemWorks), so you can't just click over to use Fix-It's disk checking and defrag tools. That's a minor nuisance at worst: it adds one more keystroke to regular maintenance (and, of course, you can schedule the tools to run automatically).

• •

  The system monitor doesn't offer some of SystemWorks' options, such as free GDI percentage. On the other hand, the system monitor is more cleverly designed, with an AutoHide facility so that it can operate like the Taskbar.

• •

  The disk defragmenter seems to be slower than Norton's. It may be doing a better job of optimizing in the process; I can't really tell.

  This story isn't complete, and may never be. So far, I don't miss Norton's features. Is my system more responsive? I think so. Is CrashProof better than CrashGuard? I may know more after I've run Ventura a few times (almost nothing else ever crashes on my home system). To be continued, possibly.

Product Watch

Bad Idea Watch

You never know where bad karma will crop up. In this case, it's page 10 of the September 1999 Stereo Review's Sound & Vision. The news watch mentions that Ç-Dilla (now owned by Macrovision) is developing technology to "prevent a computer's CD-ROM drive from playing music CDs, making it impossible to copy them with a CD-R drive or post them on the Internet."

The AudioLok system would add garbage data to an audio CD so that it looks like a CD-ROM ­ but a defective CD-ROM. Regular CD players would ignore the garbage ­ but other little tricks would assure that you couldn't copy an AudioLok CD in a consumer CD recorder either, or pass a digital data stream from a player.

From the perspective of the electronics industry, this is wonderful on several counts. It not only makes the recently introduced audio-CD recorders useless, it also rules out all the high-end CD decks with separate digital-to-analog processors. Sure, it would prevent people from "stripping" MP3 copies (or pure digital WAV copies) of CD tracks ­ but it also would mean that you couldn't play audio CDs on your PC while you're working. The wrecking of audio CD recorders is particularly interesting, given that the higher price of audio CD-R blanks goes to pay a royalty fee to record publishers.

With any luck, this incredibly anti-consumer development will do even less well than Divx. I would think that equipment companies that build digital processors or CD recorders would sue any record company that used this process, at least any such company that's part of RIAA and, thus, the bargain for royalties. It seems highly improbable (or at least flatly hypocritical) that Sony (which is one of the world's three largest recording companies) would even consider such a move, given (a) that they're a leading producer of CD decks of all kinds, including recorders and (b) the presumed corporate memory of Betamax and Hollywood's legal attempts to shut it down. Literally tens of millions of CD-ROM and DVD-ROM drives are being sold every year (mostly installed in new PCs); can record publishers be so stupid as to want to cut out that playback market?

I'm sure Macrovision will assure record publishers that AudioLok has no audible effect on audio players. That's probably what they say about Macrovision itself, but for several years we watched Macrovision-encoded VHS videos with a wavering, unwatchable top inch or so of the screen. Our new TV cured that, but I'm satisfied that such assurances can be hollow.

I'll be watching this one, with the hope that nothing more is heard from it. Personally, it's fairly simple. If any new audio CD carried a label indicating AudioLok, I wouldn't buy it ­ and I would send the publisher a nasty letter saying why. If AudioLok wasn't indicated but was there, I would return it as defective. I haven't started recording compilation CD-Rs yet (but probably will next year) ­ but I do most of my CD listening while working at my PC. I sure don't have these surround speakers on my desk to listen to Windows sound effects!

How does this affect libraries? Some of you probably read my May 1999 American Libraries article about pitfalls in preserving digital resources. Right now, audio CDs are an exception: they can be cleanly copied to hard disk, to other digital media, or to CD-Rs or CD-RWs. AudioLok would add audio CDs to the growing list of inherently transitory media. Isn't that wonderful?

Flat PCs Continued

Recent editions mentioned Gateway's Profile and NEC's Z1, the first two "flat PCs." Both PCs consist of desktop LCD displays with the rest of the PC built into the display's casing. The September 1999 PC World reviews both systems, with the overall conclusion in the headline: "Flat-PCs offer more style than substance." On the other hand, T. Capen's dual review concludes, "Would I buy one? Not as my primary PC. But as a second system, either flat-panel machine would be very tempting."

Both systems come with 15"-viewable displays. NEC puts the speakers, DVD-ROM drive, diskette, and power switches in an extended panel beneath the LCD itself; the wireless keyboard includes a pointing device to emulate a mouse. Gateway puts the speakers in the display's pedestal; the diskette and DVD-ROM drives are behind the LCD, mounted vertically.

The strong points of both PCs are that they're sleek, small, and self-contained: like notebook computers but with better display ergonomics. The weak points are expense, expandability, and power. While both systems are reasonably complete, neither one offers much in the way of expansion. Gateway uses an AMD K6-2-400 and charges $2,299; NEC uses a Pentium III-450 and asks $2,499. In either case, you'd probably pay around $800 less for a traditional system with a slightly larger CRT.

Big Flat Displays

Without boring you about the virtues of 18"-viewable displays, I'll just repeat that I love mine. Now you can get that much screen space in a lightweight thin LCD display; you can even mount these displays on the wall. The two units reviewed in the 21 September 1999 PC Magazine are both essentially the same display: a Sharp LCD panel shipped with an ATI Xpert LCD graphics card, but with different electronics for the two displays.

Native resolution for both displays is 1280 × 1024, which I find to be the most "natural" resolution for this screen size, and both displays had fine focus, no jitter, but narrower color ranges than CRTs (typical for LCD). There's a problem with the 1280 × 1024 resolution, which reflects the actual LCD elements: that's a 5:4 ratio, whereas all other standard Windows resolutions are 4:3 (640 × 480, 800 × 600, 1024 × 768). Thus, in addition to the usual problems in scaling displays on LCD panels, you have a problem with proportionality. (Why isn't the SXGA resolution 1280 × 960, which would retain the 4:3 ratio? Ask Microsoft. Not only isn't 1280 × 960 a standard resolution, Windows 98 doesn't offer it as an alternative.)

The problem with these two displays is just what you'd expect. Sony's fine CRTs with this size viewable display go for $640 (or $960 for the truly flat screen). These two sell for $3,000 and $3,880 respectively: from three times as much as Sony's best CRT to five times as much as the very good one. For that matter, as Poor notes, you could buy two 15" LCD screens (including graphics cards) and still have enough left over for a pretty good computer; that would get you the same screen area, albeit in a clumsier configuration. For what it's worth, the clear choice here is the Princeton DPP800: it's $880 cheaper than the ViewSonic VPD180, easier to set up, and has three times the warranty on the backlight.

Portable PC Dictation

A few months ago, Dragon introduced the Naturally-Mobile package, a lightweight digital mobile voice recorder designed to be used with Dragon's Naturally-Speaking voice-recognition software. Now there's a competitor in a field that could get interesting. The 22 September 1999 PC Magazine reviews Lernout & Haspie's $230 Voice Xpress Mobile Professional, which includes L&H's voice recognition software and the Olympus DS-150 digital voice recorder.

The voice recorder uses two AAA batteries and weighs a mere two ounces; it's half as thick as Dragon's recorder and holds up to 75 minutes of speech. You can speak directly into the recorder or use a headset mic (included), and there's voice-activated recording as well as reasonably effective environmental noise compensation.

These devices (and voice recognition software in general) won't replace touch typing, but they make great alternatives for those who can't or don't type well, and the portable recorder seems like a wonderful idea for people who think best away from the keyboard.

Review Watch

Desktop Computers

Brown, B. (1999), "Where to shop in PC land,"PC Magazine, Vol. 18 No. 15, pp. 168-94.

Here's an unusual approach for PC Magazine: rate the combination of shopping experience and PC. The magazine selected 16 sources of low-cost PCs for home use, typically looking for a system costing $800 or less (without display). They sent out secret shoppers and evaluated the experience in each case. It's an odd lot: they went to Gateway's Country Stores (which are showrooms, but which will take orders for home delivery) but not the Gateway Web site; they went to Dell and Compaq Web sites for refurbished units rather than new ones; and they rarely had much bad to say about any store, including Wal-Mart.

The single Editors' Choice goes to Gateway Country, where they purchased a Gateway Essential 400c. They found the staff knowledgeable and courteous, you can try out the system, and this particular system is more powerful than most low-cost PCs and offers fine value for money. I won't argue with that: other than additional memory, an upgrade to a DVD-ROM drive, and a bigger display, my newish home PC is basically identical to the Essential 400c, and I still love it after half a year.

Ryan, B. (1999), "The honor roll," FamilyPC, Vol. 6 No. 8, pp. 98-109.

Reviews of five desktop PCs and four notebooks with particular attention to their suitability for students (presumably college students). They set a maximum price of $1,500 for desktop configurations and $2,000 for notebooks.

Among desktops, the top score goes to Apple's iMac: its compact nature suits dorm rooms, although they do advise checking to make sure that the campus systems all support Macs. Although I've seen Mac zealots proclaim about the stereo sound from the iMac's internal speakers, this review takes a more realistic perspective: "One drawback is the tinny sound from the iMac's built-in speakers; if you want to use the iMac to play audio CDs, you should invest in a good set of headphones or external speakers."

I was a little surprised at the second highest rating: the iDot 333M2V, one of the rock-bottom cheap PCs. It gets points for being dirt-cheap ($608 with modest display and adequate RAM), but they didn't take off for the iDot's total lack of productivity software, not to mention total lack of track record on technical support. I think most students would be better advised to choose the third or fourth place choices: the Gateway Essential 466c ($1,403, but that includes the most well-rounded system in the bunch, including a Canon inkjet) and Dell's $1,429 Dimension L466c.

The four reviewed portables cover such a broad range of the low-end spectrum that it's hard to compare them well. The highest rating goes to Compaq's $1,999 Presario 1685 with its better-than-average speakers and DVD-ROM drive, even though it's hefty at seven pounds, has a smaller screen than the second-choice Dell Inspiron 3500 ($1,958), and offers short battery life. I would probably favor the Dell, since it's a little lighter, has a gorgeous screen, is faster, and has better battery life. The other two systems are lightweights: Acer's $1,999 TravelMate 330T at 4.1 pounds and Sony's $1,999 Vaio 505TR SuperSlim at 3.1 pounds. The Acer squeaks out a Recommended rating. The Vaio doesn't; its screen is by far the smallest of the group, and its keyboard is "nearly full-size," which translates to a keyboard that touch typists will hate. Any serious college student needs to be a touch typist; the keyboard problem should rule out the Vaio for this application.

Venezia, C. (1999), "Athlon vs. PIII at 600 MHz," PC Magazine, Vol. 18 No. 16, pp. 41-4.

Here's the hard reality: unless you're a gamer, doing high-end graphics work, or planning to do video editing, PC hardware caught up with software demands about half a year ago. Celeron-400 systems offer all the speed most people need for most applications, and in most cases you're better off adding RAM or getting a larger display (with more display RAM) rather than going for a faster chip. (For that matter, if you do high-end graphics work or video editing, Windows probably isn't your best platform anyway: look to the new G4 PowerMacs.)

But some people always need the hottest machine on the block, and this review covers the fastest Windows systems available as of late summer. The big surprise (if you haven't heard it): for the first time ever, a competitor's CPU performs faster than Intel's on all standard measures. AMD's 600MHz Athlon runs faster than Intel's 600Mhz Pentium III ­ and the Athlon is available in a 650MHz version. (By the time you read this, Intel will probably have 650MHz CPUs and possibly a new 800MHz line.)

The brief review covers four speed demons, all surprisingly inexpensive ($2,099 to $2,678 with 128MB RAM, 18" display, DVD-ROM drive, V.90 modem, and at least 13GB hard disk). Editors' Choice is the Dell Dimension XPS T600 ($2,678) for its overall value ­ but the fastest (and sleekest) of the four is IBM's $2,598 Aptiva S Series 860, using an Athlon 600.

Digital Cameras

Grotta, D. and Grotta, S.W. (1999), "The closest yet to film," PC Magazine, Vol. 18 No. 15, pp. 56-7.

This mini-roundup includes two-megapixel units from Olympus, Ricoh, Sony, and Toshiba, costing $700 to $1,000. Three cameras offer true 1600 x 1200 resolution; the fourth, the Editors' Choice Ricoh RDC-5000, pushes up to 1792 x 1200, a full 2.3 megapixels. It's also one of the two least expensive units at $700 and has generally good features.

McClelland, D. (1999), "Digital cameras develop," Macworld, September, pp. 78-84.

It's been five years since Apple's $749 QuickTake 100 offered 640 x 480 pixel images for less than $1,000. Today, the lowest-end camera in this 14-camera roundup (the $399 Olympus D-340R) costs about half as much and captures four times as much detail in an image: 1280 x 960 pixels, the baseline for true "megapixel" cameras. One of the units in this group goes all the way up to 1600 x 1200: just under two million pixels. (Some others claim 1600 x 1200 pixels, but achieve it through interpolation; the $999 Nikon Coolpix 950 actually captures all those picture elements.)

Although this group review doesn't describe each camera in any detail, it does offer print samples of their output (as refined with standard Photoshop tools) and some reasonable buying advice. The Coolpix 950 gets one Editors' Choice nod, since it offers pictures as detailed as $15,000 digital cameras and fine versatility. But it's expensive and not as convenient as some others. Their other choice is the "camera to take on a social occasion," the $699 Olympus D-400 Zoom. It's the most likely to get a good shot without much fuss, and comes in a sleek body based on Olympus' sterling Stylus film cameras.

Displays and Graphics Adaptors

Case, L. (1999), "3-d graphics boards," PC Magazine, Vol. 18 No. 15, pp. 41-8.

This first look roundup covers six graphics adaptors that offer strong 3-d performance at mainstream prices. So far, of course, most 3-d requirements come from games. That could change, but at the moment even the 3-d features of spreadsheets and presentation systems don't use hardware acceleration.

Still, these are fairly reasonably priced boards and offer first-rate 2-d acceleration as well. All but one of them carry an astonishing 32MB of display RAM, and that exception (with a "mere" 16MB) incorporates a TV tuner and TV input and output, along with on-the-fly MPEG2 compression. Most of these cards come from the manufacturers of the chips themselves; at this point, nVidia is about the only producer that sells entirely to other boardmakers.

But then, guess which chip offers the best performance? The nVidia TNT2 Ultra, used in Diamond's $200 Viper 770 Ultra, the Editors' Choice for performance and features. If that's too much to pay, look at the $150 3dfx Voodoo3 3000 AGP ­ or, if you need TV features, consider ATI's $190 All-in-Wonder 128.

Gowan, M. (1999), "17-inch CRT monitors," Macworld, October, pp. 32-3.

The opening sentences to this brief roundup state that big graphics displays cost too much for most of us, and suggests that 17" (16"-viewable) displays "fill the bill" for most applications. That omits the middle ground, today's shallow, reasonably priced 18"-viewable displays. Meanwhile, these six units cost less than $500.

Unfortunately, none of them is all that great. Sony isn't in the lineup, but NEC is ­ and its unit ties for third with two other displays. No display measures up to Apple's own $499 Studio Display, but the $469 Iiyama Vision Master Pro 410 comes close and gets the same 4-mouse rating. Both the Iiyama and Apple use aperture-grill screens (Trinitron or equivalent); the Iiyama has a flat screen.

Poor, A. (1999), "5-pound projectors," PC Magazine, Vol. 18 No. 16, p. 53.

Portable graphics projectors are getting cheaper and lighter, although "cheap" continues to be a relative term. This review covers two systems that cost a lot less than the $12,000 champions of just a year ago. Compaq has an odd-looking $4,500 projector, the MP1600 Microportable, that weighs a mere 4.2 pounds and stands on edge. It offers 1024 × 768 native projection, uses DLP technology rather than LCD, and offers very good brightness, uniformity, and contrast. There is some blurring in the corners and colors seemed to lack "punch."

The remarkable value is Lightware's $2,995 Scout, still a lightweight at 5.2 pounds. It's only about half as bright, has much lower contrast, and runs at 800 x 600 native, but it offers vivid color and (unlike the Compaq) includes built-in stereo speakers.

Portable Computers

Gore, A. (1999), "PowerBook G3s," Macworld, September, pp. 34-5.

Two recent PowerBook notebooks bring true portability back to Apple's notebook line: these units weigh less than six pounds (with CD-ROM drive and battery, but without AC adapter), while the earlier PowerBook G3s were a hefty eight pounds. The G3/333 ($2,499) and G3/400 also offer decent battery life, although not quite enough for Apple's silly advertising claim that you can watch Austin Powers twice on one charge.

These are fast systems and reasonably sleek (1.7 inches thick). Both include 64MB RAM, 8MB video RAM, a 14.1" display, Fast Ethernet, and a V.90 modem. The G3/333 adds a 4GB hard disk and 24 x CD-ROM; the G3/400, 6GB hard disk and DVD-ROM. If you're a Mac user who's eyed the sleeker Windows notebooks, these look to be very good Mac alternatives ­ fast, reasonably light, and decently priced (though hardly bargains).

Printers

Mitchell, K. (1999), "Color ink-jet printers," Macworld, September, pp. 36-7.

Mac users have been a bit hard up for inkjet printers; for a while, Epson was the only game in town. This review considers a new Canon and two new HP DeskJets that offer good alternatives for USB-equipped Macs. Interestingly, ratings for the three printers and two new Epsons are identical: four mice throughout.

For home users, Macworld recommends the $300 HP DeskJet 882C. They suggest Canon's $320 BJC-6000 for small businesses thanks to its versatility and large ink supply, noting that the other three general-purpose printers are also good choices. One Epson is to some extent a specialty printer: the $499 Stylus Photo 1200, which offers tabloid-size output and excellent photo quality (with the right paper) and color adjustment, although its plain-paper text output is mediocre.

Scanners

Fraser, B. (1999), "Slide scanners show off," Macworld, September, pp. 86-91.

This is a specialized roundup: seven scanners that handle nothing but film ­ 35mm slides and negatives, in some cases Advanced Photo System cartridges and 4 × 5 film. But if you do need to digitize extensively from film originals, these scanners offer much higher resolution than ordinary desktop scanners ­ at a much higher price. The Editors' Choice, Nikon's Super CoolScan 2000, runs $1,899; others in this roundup cost as much as $16,995! The Nikon includes the snazzy Digital ICE technology, which can remove dust spots and scratches from scanned film by using special techniques to separate out the protective surface layer of film from the colored layers. Incidentally, they're not saying that the Nikon produces the best scans (the $12,000 and $17,000 scanners offer even better quality), only that it offers the best combination of high quality and reasonable price.

Grotta, D. and Grotta, S.W. (1999), "Scanning made simple," PC Magazine, Vol. 18 No. 15, pp. 196-224.

Twenty new scanners costing less than $500, all but three of them with USB or parallel connections rather than SCSI cards. I'm not sure I go along with the end of the subhead ("No home or small office should be without one"), but there's no question that you can now get high-quality scanning at a reasonable price ­ and without opening your PC.

Unfortunately, they don't show scan examples from all 20 units; they do show one of the worst (and it's fairly dull although not terrible) and the best (which is far better). These are all flatbed units; although you can attach autofeeders to some scanners, pagefed scanners have almost gone the way of hand-held units: too limiting for the market.

Two Editors' Choices are offered in two price ranges. If you're really on a budget, they recommend Visioneer's $130 OneTouch 7600 for its sharp scans, exceptionally easy operation, and good document-management software. If you want the best scans and can afford a little more money, there's the Epson Perfection 636 ($300, SCSI) and 636U ($230, USB). The Epson pair offers outstanding quality and very fast scanning with very good software.

Security Software

Morris, J. (1999), "Protect your PC and your privacy," PC Magazine, Vol. 18 No. 15, pp. 107-36.

How paranoid should you be about your computer and computing? That depends on your situation. For example, high-speed Net access usually means that your computer is always connected to the Net. That saves time when you want to go somewhere on the Web ­ but it can also open your computer to attack from the outside at any time.

This article offers a reasonable overview of the issues and includes reviews of a range of software to protect yourself. Most of you won't need all of this, but you might want to consider how many precautions do make sense. Their Editors' Choices are eSafe Protect ($39) as a general-purpose security suite, InvisiMail Deluxe 4.0 ($45) for secure e-mail, Cookie Pal 1.5 ($15) for cookie management, SpamKiller 2.6 ($30) to cut down on spam, and Password Safe 1.7 (free) to manage passwords (and generate random passwords).

Roberts-Witt, S. (1999), "Protect your business," PC Magazine, Vol. 18 No. 15, pp. 151-66.

Businesses need more protection than individuals do ­ and libraries are businesses in that respect. I suspect one article won't give you the information you need as your library becomes more enmeshed in the Web, but this one's a good start.

In addition to some overall coverage, the article reviews offerings for two key aspects of business Internet security. FireWall-1 4.0 from Check Point gets the Editors' Choice as a firewall ($7,995 for 100 nodes). Axent's NetProwler 3.0 and Intruder Alert 3.0 ($7,995 for NetProwler, $1,995 for Intruder Alert and $995 for the server agent) get the nod to patrol network traffic, to identify attacks or irregular behavior and respond to them. They review some vulnerability scanners, but offer no Editors' Choice.

The Details

Crawford's Corner is written by Walt Crawford, an information architect at the Research Libraries Group, Inc. (RLG). Opinions herein do not reflect those of RLG or MCB University Press. Comments should be sent to wcc@notes.rlg.org CD-ROMs and DVD-ROMs for review should be sent to Walt Crawford, 1631 Columbia Drive, Mountain View, CA 94040-3638; Windows only. Visit my Web site: http://home.att.net/~walt.crawford