System 7.5 Update 2.0 Custom Install — If you’ve upgraded to System 7.5.5 (see TidBITS-346) and later found that you needed to re-install something from System 7.5 Update 2.0 (which upgrades System 7.5 to 7.5.3), you discovered that System 7.5 Update 2.0 won’t run under System 7.5.5! Apple routinely does this with system software installers to ensure users don’t unwittingly mix-and-match system components that were never designed to work together. However, with the plethora of Mac clones, system revisions, and updates lately, not being able to install components of the System 7.5 Update 2.0 under 7.5.5 can be troublesome – users particularly need to get Open Transport 1.1, MacinTalk, or other components from System Update 2.0.

Apple’s supported solution is to do a clean install of 7.5.3 with whatever additional components you need, then upgrade to 7.5.5 again. However, Apple is also providing an unsupported updater installer script that you can use with System 7.5 Update 2.0 under System 7.5.5 by dragging the script to the 7.5 Update 2.0 installer icon. Back up your system, turn off all non-7.5.5 extensions, and read the instructions first! Although this script may be overshadowed by future component releases (such as Open Transport 1.1.1), reports that many computers may not ship with System 7.5.5 for some time may give this installer script a long lifetime. [GD]

<ftp://ftp.support.apple.com/pub/apple_sw_ updates/US/mac/Unsupported/Sys75_Upd2.0_ Custom_Install.hqx>

Netscape Navigator 3.0 and Older Macs — Netscape has discovered a problem with Navigator 3.0 on Macs with the so-called "dirty" ROMs and more than 8 MB of RAM installed. Machines possibly affected include the Mac II, SE/30, IIcx, and IIx. The problem isn’t subtle – a system error on launch, but the solution is fairly simple. If you turn on Apple’s Virtual Memory or install Connectix’s RAM Doubler before launching, the problem won’t appear. An alternate fix is to upgrade to the beta of the week from Netscape, 3.01.b1, which is either 3.7 MB or 5.1 MB in size. [ACE]

Last week brought the interesting news that hard drive vendor (and TidBITS sponsor) APS plans to start selling a line of Macintosh clones, probably sometime in November. APS initially plans three machines in the M*Power line, two desktop machines based on 160 MHz and 200 MHz PowerPC 603e chips, and a minitower based on a 200 MHz PowerPC 604e chip. The motherboards come from Motorola, and APS plans to differentiate its clones by allowing customers to specify precise hardware configurations on top of the base configurations.

The basic specifications for the two desktop machines, the MPower 603e160 and MPower 603e200, include three PCI slots, 16 MB of RAM, 8x CD-ROM drives, 1 MB of VRAM (upgradable to 4 MB) and 1.2 GB hard disks, with prices at $1,399 and $1,599. The 603e160 ships without a Level 2 cache, whereas the 603e200 comes with 256K of Level 2 cache, and both machine can take up 512K of Level 2 cache.

The $2,599 M*Power 604e200, which comes in a minitower case, features five PCI slots, 24 MB of RAM, a 8x CD-ROM drive, and a 2.5 GB hard disk standard. It has 2 MB of VRAM and a 512K Level 2 cache.

Although these prices include ADB keyboards and mice, they do not include monitors, and I expect most people will want more RAM as well, given the rapacious RAM requirements of many of today’s popular programs.

<http://www.apstech.com/aps-mpower.html>

When I asked APS vice-president Paul McGraw about the clone market, he explained the move, saying, "We’re already supporting most of the system, and this is the natural leverage of the knowledge, inventory, and internal systems that we already have." In other words, APS has proven that they know how to assemble, sell, ship, and support a wide variety of storage devices – moving up to Macintosh clones is a logical extension of these skills.

In addition, APS is probably looking to evolve its product line for the future. Hard disk capacities continue to increase as prices plummet (the smallest disk APS carries is now over 1 GB). The obvious result of those facts is that within a relatively short time, most normal users will have all the storage they need for some time. High-end users will always continue to purchase 10 GB drive arrays and the like, but the market of people upgrading from 80 MB disks will continue to shrink.

From an industry standpoint, I think it’s great to see APS making Macintosh clones. Although I’m of course biased since I’ve long liked APS, I think it’s important for Apple and the Macintosh market in general to have more clone vendors, especially those that are already well-known within the Macintosh hardware market. Also, it appears that APS’s machines will come in on the lower end of the price scale, which can only help get them in the hands of more people.

One way to make a Macintosh Web server perform feats of what look like magic is to employ the help of products like Maxum’s NetCloak and NetForms. NetCloak enables you to serve different pages to different browsers, and to include counters as well as date and time notations. It can also perform a number of special tricks, like only showing portions of a page based on the time of day, or randomly changing a page. NetForms helps with managing forms: it can automatically create Web pages that contain messages sent in from people surfing the Web, and it can flexibly send email containing the contents of filled-in forms. For instance, when you subscribe to the TidBITS mailing list via our Web site, that form is made possible via NetForms.

Maxum was an early player in this field, and what I’ve reported above is hardly news, though it does explain how some Macintosh-based Web sites operate behind the scenes. What’s new is a free add-on from Maxum, called TagBuilder 1.0. Realizing that a sizable number of Web authors are migrating to graphical HTML authoring tools like golive and PageMill, Maxum created TagBuilder to address the problem that people attracted to graphical HTML authoring tools won’t be excited about typing lots of Maxum-specific HTML extensions (which are necessary to make NetCloak and NetForms work), especially since these HTML authoring tools typically offer pitiful automation options.

The TagBuilder window provides a simple, outline-based interface that displays tags for use with NetCloak or NetForms. Once you see a tag that you wish to use, you simply drag it into the HTML editor of your choice (assuming that you can drop text into your HTML editor). Regrettably, Maxum has yet to implement copy and paste for those editors that don’t support drag & drop. TagBuilder also provides a helpful description of each tag, but NetCloak and NetForms are deep products, so users need an understanding of when they’d want to use which tags.

I’m pleased to see Maxum creating a simple product that helps people use Maxum software in conjunction with software from other companies. TagBuilder is a good example of how a company can provide useful functionality without locking users into a specific ancillary product.

TagBuilder ships with NetCloak and NetForms and will ship with Adobe PageMill 2.0. Anyone interested in playing around with demos of Maxum products or with TagBuilder can find them at the Maxum Web site.

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

Maxum Development — 630/830-1113 — 630/830-1262 (fax)

My article in TidBITS-346 about why products may not be mentioned in articles prompted some additional suggestions and a few queries worth addressing.

Tom Negrino <[email protected]> writes with a reason that no one has accused me of yet:

Surprisingly, you neglected the reason for ignoring a product that stuns me the most when I’m accused of it:

"You’re being bought off by Apple / Microsoft / Adobe / Joe’s Software Company not to mention a product."

As I said to my girlfriend last month when Apple was delivering the usual load of gold bullion onto the front porch, "I can’t believe that some people question the journalistic integrity in the Mac business!"

Oy. My checking account wishes it were so.

Peter Rosenthal <[email protected]> writes:

As a small Mac developer (Interactive Media Corporation) I really appreciated your article in TidBITS-346, "But You Didn’t Mention." It was refreshing to get a glimpse of the other side of the review process. One issue that is frequently on the minds of smaller developers that you didn’t cover is the impact of large advertisers on the review process. Do they muscle their way into the editorial side, and impact the objectivity of reviews?

Adam responds: It can seem like there is a bit of collusion at times. However, in all the writing I’ve done for Macintosh publications, I’ve never had it suggested that I change what I’ve said because of an advertiser, and any editor will adamantly deny that there’s any link between editorial and advertising. That’s true, but only from the editorial standpoint. The advertising sales people know what products are being reviewed or written about in any given issue of a magazine, and they try to sell specifically placed ads to the companies mentioned. That’s why you can read a review of Microsoft Word and see a full-page ad for Word on the next page. It’s unlikely that the editor or writer knew about the ad beforehand, but the ad sales person definitely knew about the article beforehand. If there was any cross-over, it would likely be in smaller publications where there’s less separation between editorial and advertising.

The problem is that when you’re accused of this as an editor or writer, all you can do is deny it – there’s no way of proving anything. I can say (and it’s true) that we’ve never avoided mentioning a product in TidBITS because a sponsor asked us not to, but if someone believes that I’m toe-sucking pond scum who’s out to destroy some small company anyway, my denial won’t carry much weight. That’s why reputation is worth a lot in this world, especially on the Internet.

Chris Harvey <[email protected]> writes:

When reading reviews of hardware categories in industry magazines, occasionally a major product will be missed in the review. I find these omissions extremely frustrating because it becomes difficult decide on the best product. I can understand it if the review comes out at a time when models from a certain manufacturer were changing, but often reviews will miss products that have been available for months and aren’t due to be replaced for many more. So what’s the story? Is there a reason for this other than bad organization? Does this sort of thing frustrate people in the publishing industry?

Adam responds: Hardware is a royal pain, because unlike software, most hardware is merely loaned to the publication for review. So, along with the general difficulty of dealing with some companies, you have the problem of short supply (since hardware is expensive to produce in comparison to software, companies may only have a few review units available at any one time) and the problem of units that are damaged in shipping or are otherwise dysfunctional on arrival. It’s also possible that a new model is planned, but ships later than expected, making the previous model stay in the market longer than anticipated. And yes, writers trying to do a complete round-up of some category of hardware find it extremely frustrating when they can’t include one major contender.

We almost never review hardware in TidBITS because it’s not worth the effort of receiving the device, unpacking it carefully so you don’t lose anything, testing it within a relatively short time period (30 days is standard), and then packing it up and sending it back. Thus, we mainly publish reviews of hardware that someone has bought and wishes to review for us.

In addition, proper hardware testing and comparison is an extremely expensive proposition, since you need a test lab, and an organization the size of TidBITS simply doesn’t have the time, staff, or money to get into that business.

I’ve been working on and thinking about the Internet for many years now, and I’ve seen a lot of technologies come and go. Most don’t stick around for long because, frankly, they have problems. Perhaps they weren’t well thought-out to start with, or perhaps the implementation never comes together, or perhaps the company in question isn’t sufficiently with it.

I say this by way of introduction of a company and a technology that I believe bears serious watching in the future of the Internet. The company is Intermind, and their hot new technology is called Intermind Communicator. Intermind isn’t a startup, having been the marketing force behind The Internet Adapter (TIA – a software SLIP server), but they don’t have the baggage of old technology or old ideas about how the Internet works. That’s refreshing. And their new product, Intermind Communicator, reflects the company’s fresh ideas and realistic knowledge of the Internet.

What Does It Do? Intermind Communicator is, at its base, a new way of communicating on the Internet. It fits snugly between email and the Web, both of which have their pros and cons. Email is active, in that you directly send a message to someone else. But, it isn’t an efficient way to communicate with lots of people (running mailing lists is work!), and it doesn’t have the appeal of the Web’s media flexibility. The Web, on the other hand, may be able to exploit fonts, graphics, and layout, not to mention hypertext linking, but it’s passive – readers must actively seek you out, and you can’t automatically draw someone back to your Web site every week.

Intermind Communicator relies on the Web for transport and display, but builds in the active part of email, changing the dynamic of the communication process. Let me give an example of how this will work with TidBITS, since TidBITS is also being published via Intermind Communicator, using what Intermind calls a "hyperconnector," a small file that contains information about the item being published, including publication name, description, polling frequency, and so on.

First off, you install Intermind Communicator, which runs on your computer and uses a Web browser as its interface. If you want to receive TidBITS via Intermind Communicator, you subscribe to our hyperconnector, which is a matter of following a Web link to download the hyperconnector file and automatically add it to your Intermind Communicator database. Once that’s done, Intermind Communicator reads the contents of the TidBITS hyperconnector and checks our Web server for updates once each day. TidBITS only comes out once each week, but checking once a day helps spread the load, since not everyone’s computer will check in for updates at exactly the same time. For instance, if you don’t turn your computer on until Friday, Intermind Communicator checks our server then and picks up the current issue of TidBITS at that point.

That’s great, since there’s no possibility of a bounce, as with our mailing list, but all our Intermind Communicator subscribers still get the issue promptly. Since Intermind Communicator is mediating the communication between us, more things are possible. We know how many people have used our hyperconnector (although we know nothing about individual subscribers – not even email address, so Intermind Communicator protects privacy even more than a mailing list), and subscribers can personalize the parts of an issue of TidBITS they get.

For example, perhaps you only like reading about Internet issues and articles that Tonya writes. If we have set up the appropriate topics (a relatively small number) and categorized articles within each issue – which we plan to do eventually – then Intermind Communicator would only retrieve those articles when it snagged a new issue of TidBITS. This isn’t the same things as a keyword search: since we as the publishers do the categorization, you are far more likely to get what you want.

We can publish TidBITS in several ways via Intermind Communicator, and either way works fine with the topics. First, we could publish just headlines and URLs to the articles on our Web server. That works well for people whose machines are directly connected to the Internet. Second, we could send out the entire text of an article instead of the URL to it, which is better for people who want to read TidBITS offline. Although choosing between those two methods is up to the publisher, I expect that we will support both methods eventually; for the moment we actually use a third approach – more on why in a bit.

How Does It Work? I’ve been talking with Intermind for a number of months, and here’s how I understand the inner workings of Intermind Communicator. At its base, it’s a database combined with a Web server that only works for a Web browser on the same machine. The database has a query language that allows the interface (which Intermind does entirely in HTML so as to piggyback on Web browser development) to access the database through the kernel. And of course, the database can also access the Internet, making for a five-layer system that looks something like this:

User interface

Query language

Kernel

Object-oriented database

File and communications infrastructure

Intermind designed the program to be highly modular, which improves portability between platforms (Macintosh, Windows 95, Windows NT, and Unix eventually) and makes it much easier, for instance, to rewrite the entire interface in Java, if Intermind chose to do that.

In essence, I see the utility of Intermind Communicator being that it gives more control over communication to both parties. Publishers know that information is being delivered to readers in a timely fashion and with the advantages of HTML, and readers get to say what aspects of a publication interest them and control influx of data more completely.

The Business Plan — So how does Intermind make money to justify the year of development time and support their 60 or so employees? Intermind Communicator comes in four flavors, all of which are available in one downloadable program.

• First, there’s the free reader, which is available to everyone, and enables you to subscribe to hyperconnectors.

• Second, there’s the free non-commercial publisher, which enables you to create and publish hyperconnectors.

• Third, Intermind plans to license Intermind Communicator to intranets on a per-seat basis, just like any other site license.

• Fourth and finally, there’s the global publisher version, which is for people who use the program in their commercial ventures on the Internet. There’s no set fee for the global publisher version, since Intermind recognizes (I talked with them a lot about this) that there are many different types of commercial ventures on the Internet, and no one license scheme would work for everyone. If you’re interested in the global publisher, you can talk to Intermind about licensing details.

  
  

In my mind, what’s important about this system is that anyone can subscribe to hyperconnectors for free, and anyone who wants to publish non-commercial information via Intermind Communicator can do so for free.

The Other Shoe — Now that I’ve talked up Intermind Communicator, it’s time to drop the bombshell. The Macintosh version isn’t yet shipping. I hope it says something that I consider Intermind Communicator sufficiently important to discuss at length in TidBITS when the Mac version won’t be available for a few months, probably until Macworld San Francisco in early January.

Here’s why I’m not worried. Drummond Reed, one of the founders of Intermind, is a long-time TidBITS reader and has made a point of keeping me informed about Intermind Communicator over the last nine months or so. He’s also worked hard to involve Guy Kawasaki, since Intermind feels extremely strongly about the importance of the Macintosh market, especially on the Internet. The goal from the beginning was to ship simultaneously on Macintosh, Windows 95, and Windows NT, but shortly before last week’s long-planned official ship date, it became clear that the Mac version simply wasn’t ready and the other two versions were. That sounds bad, but the reason why this happened is that the lead developer on Intermind Communicator is a Mac programmer, and when the core code took longer than predicted, he wasn’t available to finish the Mac version in time. With one of the company’s founders and the lead developer behind the Mac, I’m not worried about getting a Mac version soon.

Everything I’ve seen of Intermind Communicator, I’ve seen under Windows 95 on my Compaq Contura 400, which is a 486 laptop that I previously turned on about ten times a year. I can’t say that I like using the machine, but a Web browser is a Web browser, so using Intermind Communicator has been pretty much trivial, even when I published last week’s issue of TidBITS as a test of its publishing capabilities. I can’t wait for the Mac version to appear.

Looking to the Future — What I’ve described above is version 1.0 of Intermind Communicator. As with any 1.0 release, it lacks lots of features that Intermind dearly wanted to include, but had to pull to get it to ship on time and be understandable to an audience unfamiliar with the communications shift Intermind Communicator offers. I’ve run into a few of those limitations while trying to publish TidBITS. For instance, I first tried to set up topics like "Reviews" and "by Tonya," but I soon discovered that version 1.0 doesn’t allow multiple messages per topic. So, if Tonya wrote two articles in an issue, I couldn’t put both into the "by Tonya" category.

Then I decided to publish only two topics, "Announcement Only" and "Full Issue." That way, readers could get the top-of-issue blurb and live links to all the articles, or they could have the entire issue delivered as a single file. That ran afoul of two problems. Netscape Navigator Gold 3.0 under Windows 95 won’t allow more than 30K of text in a text field (remember that Intermind Communicator’s entire interface is inside a Web browser). I worked around that one with minor editing, only to run into a 16K limit on a message for Intermind Communicator. Once these problems are fixed in a future release, we’ll look at publishing individual articles in topics. Until then you can only get the announcement of a new issue of TidBITS, and you must read it on a Windows machine.

<http://www.tidbits.com/hyperconnectors/ tidbits.con>

You can also find TidBITS in the Global Directory of hyperconnectors that Intermind maintains as a way to simplify finding new hyperconnectors.

Intermind has other big plans for the future. Right now, the only type of data that Intermind Communicator can transfer is HTML, which they chose to reduce testing time and to keep the product more understandable. There’s no technical reason for that limitation, though, so imagine a future version that can transfer any type of data, perhaps including updates to programs you use (using a hyperconnector configured to check in once a month, for instance). Or, what about address information sent out via Intermind Communicator? Rather than try to get everyone to change their contact databases (we moved a year ago and still receive snail mail at our old address), you could just publish your information in a hyperconnector and have it checked every few months, ensuring that no one using Intermind Communicator would ever be out of date by more than a few months.

I’ve also enjoyed talking with the folks at Intermind over the last nine months because they have the best grounding in hypertext theory that I’ve seen in an Internet company. They’re fully aware of things like Ted Nelson’s proposed Xanadu system and issues surrounding micropayments and the like. Although there’s no telling if those ideas will ever appear in a version of Intermind Communicator, it’s a good bet that they’re being considered. Although I tend to be fairly cynical about new technologies, Drummond and the others at Intermind have never failed to come up with good answers to my questions. Add to that the fact that I don’t philosophically disagree with anything I’ve seen or been told about how the product and business plan works, and you see why I’m being so positive.

Will Intermind Communicator succeed? Will it be a paradigm shift in how we use the Internet to communicate? I can’t answer those questions – no one can. I can say that I believe that Intermind is a smart company, and Intermind Communicator is a good product with great potential. Other smart companies with great products have fallen by the wayside for reasons that could not have been predicted, and the same could be true of Intermind. I sincerely hope that it isn’t and that Intermind Communicator becomes the next "must-have" Internet product.

You can read more about Intermind Communicator on Intermind’s Web site at the URL below, and you can also download the Windows 95 and Windows NT versions of the program. One neat touch – since everything Intermind Communicator does comes via HTML, Intermind has visual demos of how the program works online as well.

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