Progress toward moving the TidBITS mailing list from Rice University is continuing, and, barring unforeseen difficulties, this should be the last issue of TidBITS sent via the Rice LISTSERV. If you’ve recently subscribed, unsubscribed, or altered your subscription, please bear with us for a week or two while we clean up loose ends. We’re essentially maintaining two parallel versions of the TidBITS list right now, and there are bound to be some rough edges. In the meantime, if you are not on the TidBITS list and wish to subscribe, send email to <[email protected]> rather than working through the LISTSERV. Thanks for your patience! [GD]

Amelio Outlines Realistic Mac OS Strategy — In a move seen as simply bringing Apple’s official system update policies in line with current practice, Apple CEO Gilbert Amelio announced at Macworld Boston last week that Apple intends to deliver future enhancements to the Mac OS in a series of incremental updates rather than as large, monolithic packages. What does this mean for real people? First, it means Mac OS 8 won’t emerge from Apple fully-formed; instead, components will trickle out as Apple finalizes them, starting with components like Open Transport, OpenDoc, and QuickTime, along with Finder enhancement like multiple simultaneous copy operations and "docking" folder windows. This announcement also means developers won’t be receiving the long-delayed DR1 release of Mac OS 8, most recently promised for next month.

Although moving away from enormous system releases should allow Apple to deliver technologies to users more quickly, I hope this official commitment means Apple will show some leadership in clarifying the confusing morass of system software and component releases. We’ll know soon: the first incremental system update is scheduled for January 1997, with another release following in July 1997. [GD]

<http://www.devworld.apple.com/devnews/ devnews0808.html>

Mac OS Runtime for Java — Apple has posted a pre-release of Mac OS Runtime for Java, which is essentially designed to put a Java virtual machine into the Mac OS. I can’t honestly recommend this release to anyone but Java and OpenDoc fanatics, but the runtime does allow users to run Java applets in a stand-alone viewer and within an OpenDoc document. It runs on any Mac with a 68030 or better and System 7.5. Apple says the Java runtime will work with the shipping version of Cyberdog 1.1 (due very shortly), but it doesn’t work with previous versions, including Cyberdog 1.1 betas. [GD]

In an article in the 15-Aug-96 Wall Street Journal, Lee Gomes outlines a "virtually unknown" Microsoft development group in San Jose focusing exclusively on the Macintosh, and Microsoft’s plans to promote third-party Internet development on the Mac. (I’d provide an URL to the article, but the Wall Street Journal doesn’t make their material freely available online.) I’ve received a number of telephone calls about this article – where did all these journalists get my phone number? – and all I can say is that if this group is virtually unknown, someone hasn’t been paying attention. It’s safe to say most of the Mac Internet community has heard of Internet Explorer for the Mac, one of that group’s projects. Microsoft started the Mac-only development group well over a year ago, and its employees include a number of solid, long-time Mac developers who have been visible at trade shows and online. This development group is not news.

What is news is that Microsoft is apparently providing assistance to developers working on Internet applications exclusively for the Macintosh, ignoring Windows altogether. It’s unclear what support Microsoft intends to provide, except for money. But why would Microsoft want to help Apple directly? Isn’t Microsoft the enemy?

As Apple centers its computing strategy around the Internet, Microsoft indirectly benefits from a strong Macintosh Internet development community. First, remember that Microsoft typically makes a lot of money off its Macintosh software (particularly Microsoft Office). More Macs sold means more money for Microsoft – much more money than it’s spending on this developer program.

Second, if Apple collapses and Microsoft comes to dominate the desktop computing market utterly, Microsoft is likely to face serious scrutiny from the Department of Justice under antitrust law. (Incidentally, Caldera recently filed suit against Microsoft over DR-DOS on antitrust grounds.) From Microsoft’s point of view, a healthy Apple with a secure – but small – percentage of the overall computer marketplace is infinitely preferable to no Apple at all, and these days compelling Macintosh Internet products are one way to help Apple survive.

<http://www.caldera.com/news/pindex.html>

Third, a strong, Microsoft-friendly, Mac Internet development community opens an independent front against Netscape, in a battle both companies seem to view as a life-and-death struggle. If Mac Internet developers prefer dealing with Microsoft rather than Netscape, Netscape loses part of its battle to dominate the Internet marketplace.

Fourth – and perhaps least apparent: Microsoft has another carrot to dangle in front of Macintosh Internet developers. Microsoft’s Office applications currently dominate both the Macintosh and Windows markets, and it’s no secret future versions for both platforms will be much more Internet-centric. Wouldn’t it be a coup for Microsoft if the next version of Office for the Mac worked seamlessly with all the hottest Macintosh Internet products? What if some of those products were bundled with Office? Microsoft sure isn’t going to ship Netscape’s products with future versions of Office, and most small, fast-moving Macintosh Internet start-ups would seriously consider such an offer – especially if Microsoft had helped fund their product development.

Reaction to Microsoft’s efforts in the developer community so far have been mixed. Some developers welcome Microsoft with skepticism, some with scorn, and some with open arms. Regardless of whether Microsoft truly wants to help Apple or the Macintosh, it’s important to remember Microsoft only puts its money where its business is. Like everything else Microsoft does, if there wasn’t money in these actions, Microsoft wouldn’t be doing it.

Fueled by a melange of Internet-related software, this year’s Macworld Expo had plenty of enthusiastic crowds and product announcements. We’ll cover more of these products in future issues, but in this semi-annual Macworld Expo superlatives article, I chronicle companies whose gimmicks particularly stood out or whose offerings caught my eye.

Slimmed Down Approach — The Aladdin Systems booth wasn’t new, but space-constrained attendees enjoyed the StuffIt t-shirts, which set a new standard for sartorial compression. The shirts, distributed in shrink wrap, were compressed with 50 tons of pressure to the size of a large bar of soap. Aladdin also announced an agreement with Netscape Communications to bundle StuffIt Expander with Netscape Navigator, which will mean that Netscape Navigator will finally be able to handle MacBinary files without the user needing to download another application.

<http://www.aladdinsys.com/>

Tower of Marketing Power — Never one to sit quietly in a crowd, Power Computing made quite a display with an assortment of games and marketing pitches that attracted the largest, most enthusiastic crowds of the show. The height of the action, though, took place just outside, where Power Computing offered bungie jumping from a 225 foot tower. Power Computing’s marketing firm must have had a good time with the promotional boxes being handed out at the MacWEEK Volume Buyers meeting the day before the show: the box looked like the face of a Power Tower, and when you opened the cover a small sound chip screamed as a paper doll flopped from a rubber band. Attached to the bottom of the box with a bungie cord was a t-shirt proclaiming, "We’re fighting back for the Mac." Power was (of course) showing off its latest models (see TidBITS-337 and TidBITS-339), and as with the last few major Macworld Expos, Power Computing dotted the entire exposition scene, with many vendors showing their wares on Power Computing machines.

<http://www.powercc.com/>

Buggiest Product — Pulse Entertainment gave out realistic looking plastic cockroaches and showcased Bad Mojo, a dark, gritty, CD-ROM game where you play the role of a scientist who won a grant to complete research to wipe out cockroaches. Early on in the game, you are dealt a Kafka-esque hand of "bad mojo" and transmographied into a roach. Your mission is to collect information while interacting with objects and other animals, and eventually to gain enough insight to return to a human state, although there are several possible endings. What makes this game stand out is its real-life images and careful attention to interactions and detail. Bad Mojo’s creators crafted a highly realistic environment by researching how roaches move and how surfaces become dirty. Pulse has made a 4 MB demo version available through its Web site.

<http://www.badmojo.com/bmintro1.html>

Most Interactive — MacUser, MacWEEK, and ZDNet teamed up with The Winners’ Club, a series of games which culminated with each player putting on protective eye goggles and entering a small, transparent booth the size of a shower stall with coupons and dollar bills on the floor. The goal was to grab as many of those coupons and dollars as possible while a fan blew them around. When your time was up, you exchanged the coupons for software and hardware prizes, though the people I watched only won the default prize, a t-shirt.

Best Way to Make Money on the Net? Realizing that few companies are making much money from Internet content, but that plenty of companies make money selling physical objects, such as books, via the Internet, Wolff New Media was showcasing its NetBooks series. A NetBook reviews Web sites about specific topics, such as using the Internet to find a job or to follow the next U.S. presidential election. A representative described the company as having a "newsroom atmosphere," with some 50 editors writing books and updating books online with fresh reviews. The company plans to publish one new book every three weeks. You can check out their Web site, which reportedly holds reviews of some 50,000 sites.

<http://www.ypn.com/>

Stock up on Stock — As electronic transactions and record keeping become increasingly common, stock certificates have become increasingly uncommon, despite their often interesting typography and graphic design. Operating under the theory that actual certificates could become valuable collectors items (or at least well-loved wall decorations), One Share of Stock, Inc. sells framed stock certificates. You could pick any stock, but One Share of Stock featured Apple stock at the show. The company offered all comers the chance to spend $89 for a share of $21 Apple stock, delivered as a framed certificate. Considering the commission, certificate, and framing fee, $89 doesn’t seem utterly unreasonable.

<http://www.oneshare.com/>

Staying Alive — Live Objects, components that take advantage of OpenDoc technology, showed up here and there on the Expo floor. Some, like Bare Bones Software’s BBEdit Lite for OpenDoc 1.0 module, are due for imminent release; others, like Nisus Writer 5.0, are still a few months from shipping. I didn’t get to see them, but I heard glowing reports about upcoming Live Objects from Quebec-based Adrenaline Software – Adrenaline Numbers, a spreadsheet part that should import Excel spreadsheets and offers 149 functions, and Adrenaline Charts, for making 2-D, 3-D, and even animated graphs.

<http://www.adrenaline.ca/>

Handiest Product — GBM design showcased mobile wrist supports called the Comfort Point and the Comfortype. You move the supports by resting your wrists on their "contour paddles" while you use a mouse, trackball, or keyboard. You can adjust the paddle to three different angles. The Comfort Point attaches to the back of a mouse (or trackball); the Comfortype looks much like a pair of Comfort Points mounted on a track installed in front of a keyboard. Adam bought a $20 (show special) Comfort Point, and we’ll see how he likes it.

<http://www.comfortpoint.com/>

Loudest Party — It’s hard to give this award to any one party because it seems that all the parties we went to this year at Macworld were way too loud. These are geek parties, and what we geeks want to do when we get together is talk. We don’t want to listen to music, and we certainly don’t want to listen to music played so loud that we have to resort to screaming to carry on conversations. The worst offenders this year were the Apple party at the Roxy on Tuesday night before the show and the Mac the Knife party on Thursday night of the show. It took days after each for my voice to recover from trying to scream over the deafening decibels. If you’re planning a party for a future Mac show, ditch the music and let people talk to one another.

Serious Font Management — I don’t do much with fonts these days, but I was extremely impressed with FontReserve, a new font-management tool from DiamondSoft. What’s important about FontReserve is that it has a powerful database at its core. The database manages all your font files and stores information about the font names, font IDs, foundry information, version information, and location on your hard disk. Once you have all your fonts in FontReserve’s database, you can easily create and manipulate hierarchical sets of fonts and even do things like create a Finder folder containing copies of all the fonts in a specific set for use by a service bureau. FontReserve supports all font formats, matches outlines and bitmap fonts, removes duplicate fonts (after comparing name and version information), checks for font corruption, classifies fonts according to a proposed ISO standard for font categorization, and organizes your fonts by type, foundry, and family.

<http://www.fontreserve.com/>

HTML software figured heavily into the Expo mix, but – to be honest – I was disappointed not to have my socks blown off by a new product that had been kept under wraps and that explored novel ways of using graphically-based tools to create and manage Web pages and sites. Even so, I found many products to add to my list of software to review. Claris had an enormous presence with Home Page, which is shipping for $99 with a good table editor, a confusing frames feature, and a clunky HTML text view. Gonet appeared with golive Pro, which offers an intriguing sounding outliner and an integrated tag database for $149. Golive, reviewed recently in TidBITS-337, is now available free from gonet’s Web site along with a trial version of golive Pro.

<http://www.claris.com/>

<http://www.golive.com/>

No other authoring tools shipped at the show, though Adobe PageMill 2.0 went into public beta and SoftQuad announced plans to release of HoTMetaL Pro 3.0 within the next month. HoTMetaL Pro 3.0 will list for $159; upgrades will cost $69 for current users of HoTMetaL Pro 2.0 or HoTMetaL Light.

Bare Bones Software was showing BBEdit 4.0.1, which shipped in late June. This version offers multiple undos (a feature HTML authors should appreciate, since it makes experimenting with different tags easier), a new version of the HTML markup extensions, and options in the File menu which enable you to access, edit, and save files on your remote Web server. These options are part of BBEdit and do not rely on BBEdit 4.0’s expanded integration with Frontier.

Many products offered features for creating HTML documents or GIF images in one way or another; I expect that such features will become almost ubiquitous in another year or so. In particular, Extensis, makers of QX-Tools, a QuarkXPress enhancement utility, showcased CyberPress 1.0, a $150 tool for turning QuarkXPress documents into Web pages. CyberPress purchasers will receive a free copy of PageMill 2.0. Those interested in creating the HTML needed to interact with Maxum’s NetCloak and NetForms can avoid complexity through Maxum’s new TagBuilder. TagBuilder users avoid typing HTML by simply choosing functions from a window and dragging them into PageMill 2.0 (or, according to John O’Fallon, Maxum’s president, some other HTML authoring tools). Adobe plans to include TagBuilder with PageMill 2.0; demos of TagBuilder should be available soon from Maxum’s Web site. Another product, Kaetron Software’s StencilIt, bills itself as "point and click graphics for the artistically challenged." The program offered a number of base images, which can be manipulated in a number of ways and saved in a number of file formats, including GIF.

<http://www.extensis.com/>

<http://www.maxum.com/>

<http://www.kaetron.com/>

A saying claims that you can never be too rich or too thin, but the modern equivalent is that you can never have too much bandwidth (and all of you with T3 connections to the Internet can just keep quiet). A new product from ClearWay Technologies promises to provide the bandwidth that many of us want for our Web servers but can’t afford. Called FireSite, the Web server plug-in essentially increases the effective bandwidth of relatively low-speed Internet connections, taking a 28.8 Kbps dedicated modem connection up to an effective speed of 75 to 100 Kbps, for instance.

<http://www.clearway.com/>

Needless to say, it’s not possible to push more data through a 28.8 Kbps modem with FireSite than without. FireSite works its magic by selectively moving data to another Web server (such as the one your Internet provider runs) with a high-speed connection. FireSite then serves small HTML files over the low-speed line, and redirects requests for large files (like graphics and animations) off to a Web server on the high-speed line.

Of course, that’s the most simplistic way to explain how FireSite works. In fact, if that were all there was to it, you could duplicate the technique manually by simply placing your graphics on another Web server and linking them in with absolute URLs. Many people do this already, and products like Maxum’s RushHour (a Web server dedicated to serving large graphics and similar files) aid the process somewhat. But, what about your logs? If the graphics exist on a different server, they won’t show up in your log files. Also, managing files on multiple servers is a major pain.

<http://www.maxum.com/>

FireSite solves these problems through clever use of a relational database that tracks information about files on your Web site. Using adaptive algorithms based on file size and popularity, it moves certain files from your Web server up to your slave Web server, renaming them with ISO 9660 filenames and staying within disk space limits that you set. Then, when someone requests a page that uses one of those files, FireSite intercepts the HTTP request and redirects it out to the slave server. In other words, when someone visits a site that has been "replicated" by FireSite, the overall response rate will seem much faster than it would if all the parts of the Web pages were served over the slow connection. Should the slave server go down (FireSite monitors it constantly), FireSite simply stops redirecting requests until it comes back up again.

ClearWay sells two versions of FireSite. The Standard Edition is initially priced at $349 and only replicates GIF and JPEG graphics to a single slave server. Initially priced at $839, the Multimedia Edition adds the capability to replicate some other data formats such as Java applets and Shockwave movies, and it can replicate those files to multiple slave servers, rotating requests among the slave servers to balance the load. The Multimedia Edition adds a "welcome mat" feature that enables you to select which pages on your Web site users can bookmark – direct accesses to unauthorized pages end up at your site’s home page. Another interesting capability of the Multimedia Edition is its "anti-hijack" feature, which is designed prevent people from using the URL to one of your graphics in an <IMG> tag in their own Web pages. When FireSite detects that happening, it instead displays another graphic that you specify.

What I like the most about FireSite is that it’s transparent. There are a few things to configure via a Web interface, but they’re minor (such as telling it how fast your connection is and how much Web space you have available on the slave server). You can check a real-time log and some statistics on what FireSite is doing for you. It also calculates the effective bandwidth of your site when accelerated with FireSite.

Make no mistake: FireSite isn’t a panacea for all bandwidth woes. You must have a dedicated Internet connection with a permanent IP number for your Web server, and those connections aren’t nearly as inexpensive as the more common "unlimited" personal Internet connections that have dynamic IP numbers. Prices range widely, but a dedicated 28.8 Kbps dialup connection can cost anywhere from $75 to $150 per month, in comparison with the average $22.50 per month price of a personal account. However, since a 56K frame relay or ISDN connection can cost $300 to $400 per month, FireSite can pay for itself rather quickly by ameliorating monthly charges. FireSite’s not cheap, but ClearWay is betting that the savings on the Internet connections are sufficient to make the financial equation come out clearly in FireSite’s favor.

FireSite requires about 3 MB of RAM and can use 10 MB or so of disk space for its database, and it only works with Macintosh Web servers that support WebSTAR plug-ins (also known as the WS*API). It’s probably unnecessary for many small Web servers that can serve all their files over a slow Internet connection with no trouble, but if you’re running a Mac Web server over a relatively slow connection and need better performance from your site, FireSite is well worth a look.

ClearWay Technologies — 888/552-5327 — <[email protected]>

After every Macworld Expo, we attempt to pull the overriding theme out of the hype and chaos of the show. At the last Macworld Expo in San Francisco, it seemed as though every company had an Internet product, or they had at least managed to put the word "Internet" into their products’ names. At last week’s show, a vast number of companies exhibited Internet-related products, and this time they were related to the Internet by more than just name. However, many of these companies lack a strong conception of their products, what problems they solve, and for whom these products solve problems.

I don’t want to name names, because this epidemic of confusion about the Internet seems to afflict most software companies. Only a few firms, mostly start-ups, have the proper mindset to conceive of an Internet product that solves a real problem in an elegant and realistic manner.

This problem may have a number of causes. Internet technology changes rapidly, presenting a moving target. You must determine what your program will do and get it out fast, either before the problem it solves has disappeared or before someone else beats you to the punch. Also, it’s difficult to create a product that works well for new users and appeals to more experienced veterans. And, of course, it doesn’t take much time for new users to become gurus in their own right.

More seriously, I have the impression that most people developing Internet programs don’t use the Internet much, and as a result, their programs lack strong vision. The most telling symptom of this lack of vision is when companies release so-called public betas to solicit feedback from users. Successful programs start with a strong vision and then react to feedback – consider programs from small developers who don’t need to do massive amounts of usability testing or public betas. A good example might be Anarchie, which Peter Lewis wrote to solve problems he saw with existing tools. Peter didn’t start Anarchie by asking users what they wanted; instead he created the feature set he wanted and refined it over the years based on feedback.

Working primarily from user feedback tends to result in programs with scattered feature sets and fragmented functionality. I’d like to see developers spending more research time online, seeing what issues come up in mailing lists and newsgroups, and learning more about the Internet world before formulating product ideas. That research is necessary for a focused product – otherwise we end up with yet another ill-conceived HTML editor or bookmark manager or Internet floor wax that breaks no new ground and solves no existing problems.

Apple’s emphasis on multimedia on the Internet makes sense from a technological standpoint, but I fear that it falls victim to this lack of a reality check. No one will argue that QuickTime movies aren’t cool, but I’ve seen few Web sites that benefit from their use. Similarly, on my 660AV, the Shockwave plug-in from Macromedia seems to conflict with RAM Doubler, and I need RAM Doubler far more than I need Shockwave (which, as far as I can tell, I don’t need at all). In each case, these companies are attempting to leverage existing technology to solve problems that barely exist. I’d prefer to see them make a difference by solving the real-life problems of regular Internet users.