Previous Issue | Search TidBITS | TidBITS Home Page | Next Issue
This week's late breaking news comes in the form of yet another ugly virus (INIT-29-B). We also share comments about hard drive reliability, muse further on the state of mergers in the Macintosh world, and take a look at the perceived lack of women programmers in programming-based discussions online.
Copyright 1994 TidBITS Electronic Publishing. All rights reserved.
Information: <email@example.com> Comments: <firstname.lastname@example.org>
This issue of TidBITS sponsored in part by:
I'm extremely pleased to announce that Tonya has resigned from her position at Microsoft to devote more time to TidBITS and other projects, including books and Internet adventures. Perhaps the most obvious indication of her change in hats will be that from now on we will share some of the email that has swamped me recently (I receive upwards of 3,000 messages a month). And, of course, we're looking forward to the freedom to spend a little more time away from the Macs and the nets. So, if you send a note to me and Tonya replies, that's why. If you wish to send her mail directly, her address is <email@example.com>. Oh, and for the cynics in the crowd, this is not an April Fools Day joke, although Tonya's first day away from Microsoft was April 1st. We especially liked the chronological importance of the combination of April Fools Day, Good Friday, and the weekend including both Easter and the Daylight Savings Time "spring forward" day (which Tonya has always hated with a passion, since she'd rather lose an hour in the middle of Monday afternoon than an hour in the middle of the weekend).
IBM and Motorola have announced a 100 MHz version of the PowerPC 601 chip that is the current mainstay of the Power Macintosh line. The added speed over the existing 80 MHz chip will no doubt be welcome if Apple decides to use it, perhaps late this year, but we can't help but think that the announcement is merely the latest salvo in the "Mine is bigger than yours" marketing war between Motorola and Intel.
Rick Holzgrafe <Rick_Holzgrafe@taligent.com> notes that we implied last issue that the Drag Manager's functionality (dragging text to the desktop to create a "scrap" file, for instance) will appear "automagically" in existing applications. It does not; applications must be re-written to take advantage of the Drag Manager. We also are flagellating ourselves for talking about the Drag Manager in the future tense, since it exists already and is supported by utilities such as Dayna's ProFiles. I can't wait for its functionality to appear in Internet applications including Fetch, Anarchie, TurboGopher, and Mosaic.
Paul Westbrook <firstname.lastname@example.org> and others tell us that Applied Engineering has gone out of business. It appears that a slow market for accelerators was the death knell for the fifteen-year-old company.
PC Pursuit long distance services from Sprint were discontinued as of 01-Apr-94 (no April fooling). The services provided fixed rate charges on long-distance calls, offering a useful way to regularly dial long distance to online sources. According to the message from Sprint, fewer customers were using the service, and Sprint wants to focus on providing local dial-in access to online sources. I wonder if other carriers will step in and offer a similar service. Although large online sources often have local access numbers and an increasing number of Internet providers are popping up all over, people who don't live near a major town with access numbers or Internet providers will be around for a long time. Thanks to Jeff Fischer <email@example.com> for forwarding the information.
A number of people disagreed with my statement about all hard drives being approximately equal in reliability. I based that statement on my experience and on the fact that when I looked at the MTBF (Mean Time Between Failure) ratings for modern drives, all stood between 150,000 hours and 500,000 hours - 17 to 57 years of non-stop service between failures. People who had problems with drives in the past have a point, since MTBFs used to be around 10,000 hours, or just over a year of non-stop service. If the drive is a file server, 10,000 hours is not enough, but if you power on the machine for just a few hours each week, 10,000 hours should last quite some time. I'd like to know exactly how a manufacturer determines MTBF ratings - I know of no empirical test that condenses 57 years of non-stop service into a few months. That's not even taking into account additional wear from being turned on and off repeatedly and other environmental stresses.
A low MTBF could be balanced by a generous warranty period. Standard warranties range from ninety days to five years, and frankly, I'd recommend that you go for a longer one, at least two years. Five years is great, but if the drive dies in the fourth or fifth year, you may end up discarding it anyhow to buy a new one that's far larger or faster. Five years ago I used a home-assembled 30 MB Seagate 238R (5.25") hard drive in a case that sounded like an airplane and had room and power for four half-height drives. Now I use a 1.2 GB Quantum (3.5") that's far faster and more spacious. If I still had that 30 MB drive, I doubt I'd use it seriously even if its warranty was still good.
As an additional warning, David Stodolsky <firstname.lastname@example.org> writes:
The exact same drive model obtained from different sources may be of substantially different reliability. When Quantum sells drives to Apple, Quantum knows that the drives must pass rigid requirements and will be tested by Apple and sent back if they fail (perhaps with hefty penalties and cancellation of the delivery contract). The same drive from La Cie, wholly owned by Quantum, may be one that did not meet Apple's standard! (I got four bad drives in row from La Cie.) Even worse, Quantum has released new models under the same model number. In some cases, manufacturers have been known to swap parts among drives not up to spec to get a drive that then passes (barely). Ehman was found to be shipping used drives as new. This is outright fraud, but it can happen, as I found out with one of my drives.
Apple tests different drive models before approving models for its machines. Some aren't approved, and if you buy one of these drives you may have trouble, especially if you run A/UX or some high-end multi-media applications. Know your drive supplier and the specs on the mechanism if you want to avoid trouble. For the average user, a reliable supplier with good service in case a drive must be returned is of paramount importance.
A variant of the INIT-29 virus was recently discovered, according to an announcement from Professor Gene Spafford at Purdue. The new virus strain, dubbed INIT-29-B, behaves similarly to the original INIT-29, which appeared in late 1988. It alters applications, system files, and documents, and may cause unexpected program failures or system crashes on any Macintosh, under all versions of the System. Both strains spread quickly and widely if not checked. John Norstad's announcement for Disinfectant 3.5 adds that there are only minor differences between this new B strain and the original strain, now referred to as INIT-29-A.
INIT-29 viruses alter and infect almost every type of file, including documents, though infected document files don't spread the virus. Infected applications and system files can and will spread the virus; a file need not be opened in order to be infected.
One likely sign of an INIT-29 infection (either strain) is that if you insert a locked floppy disk into the drive, your Mac reports that "The disk 'xxxxx' needs minor repairs" and offers to repair it.
The latest existing versions of Gatekeeper (1.3), Rival, and VirusDetective (5.0.11) are already effective against this new strain. (In Rival's case, you need to have the INIT-29 vaccine installed.)
John Norstad has released Disinfectant 3.5, an update to his free utility that provides scanning, repair, and protection capabilities. It's available on most online services and via anonymous FTP at:
Updates for Central Point Anti-Virus and SAM (Symantec Anti-Virus for Macintosh) may be obtained from their respective publishers on popular online services and via anonymous FTP at:
(This archive typically offers updates for most of the commercial utilities, as well as the latest versions of the free utilities.)
Datawatch has released Virex 5.03, which will be sent automatically to all Virex Protection Service subscribers, and will be available to other registered owners. Datawatch has also provided the following code that users can add to their software to allow it to detect INIT-29-B:
UDV Code for INIT29-B Guide Number = 15753664 1: 0302 3000 1276 0000 / 57 2: A9F0 303C A997 A146 / 9D 3: 2028 FFFC 8180 9090 / 4C
Spaf's announcement also mentioned that version 1.3.1 of Gatekeeper, designed to handle the INIT-9403 virus announced last month, is still unfinished. Using the latest version of the Disinfectant protection INIT along with Gatekeeper 1.3 should provide satisfactory protection.
Gene Spafford -- email@example.com
John Norstad -- firstname.lastname@example.org
Our articles on the recent corporate acquisitions and mergers generated numerous comments and a great deal of discussion.
I'm sure you will receive various messages commenting on your comments on monopoly in the computer arena and wishing for a model more like that of the entertainment (movie) business. Sorry to bust your balloon, but the movie industry is as consolidated and monopolistic as is the computer industry. The Reagan administration ignored the anti-trust laws and today studios own production, distribution, and increasing amounts of exhibition space. The independent cannot be truly independent as they must rely on the largest companies to get films out to the marketplace. Everything is in a mode of merger and consolidation with all the attendant difficulties, such as small, valuable projects falling between the cracks.
Alastair Sweeny <email@example.com> writes:
I like your analogies for software productions, but I think what is really happening is that software will become more and more commoditized.
We now see shareware that is the equal of the $500 programs a few years ago. For example, for photo work, I use the excellent Graphic Converter conversion program by Thorsten Lemke from Germany plus NCSA Image, both of which have all the features of an early Photoshop and more. [And for your FTPing pleasure... -Adam]
My point is merely that there will soon be free and shareware produced by generous hackers that provide the average user all the features they ever need.
Where will this commoditization of software leave Microsoft and Novell? Content. That is why they are moving into providing information [note Microsoft's multimedia titles like Encarta, Bookshelf, and Cinemania -Adam] and building global satellite networks that would seem to be out of their focus.
Dan Neuman <D.Neuman@mailstop.telesat.ca> notes:
I wanted to make a slight correction to your note about Teledesic in TidBITS #219. According to a Reuters and New York Times report, Microsoft has specifically stated that they are not officially involved in this project. Bill Gates owns 30 percent of Teledesic, Craig McCaw owns 30 percent, and McCaw Cellular owns 28 percent. Although the project is estimated at $9 billion, most of the funding will come from outside sources, not the owners.
Nicholas Sturm <firstname.lastname@example.org> writes:
What worries me the most about the mergers of the biggies (and the swallowing of all the independents) is what I saw happen to the textbook publishing industry over the last decade or so. Companies that knew little of book publishing bought the decreasing number of publishing companies and soon we had a mass of almost-correct books. What was an error here and there? After all it was less that one percent incorrect. By modern business standards that was sensationally accurate - much better than the U.S.P.S.'s, "We get over 95 percent of the mail there in two days." But what happens to that remaining five percent?
Where I once selected a book by looking for quality; I now select by reviewing the parts I feel I know best. Then I assume the errors in the parts where I need the author's help will be no worse than that where I can count the errors. Unfortunately - for my students - I can give them the "correction list" only for the parts where I'm confident of the best correction. They are left memorizing (and perhaps I'm left propagating) the mistakes in the other parts.
Pythaeus writes about Borland:
The rumor that I'm hearing is that Novell wants to buy Borland, but Borland's bookkeeping is so bad that Novell can't figure out what the company is worth. Supposedly Borland has been working with an accounting firm for close to a year to straighten out their books. My guess is that Novell, in its attempts to become Microsoft, will buy Borland as soon as feasible. These rumors come from a recent Novell conference.
David Lawrence <email@example.com> passed on the following comments (originally from his Online Tonight radio show) regarding the Aldus and Adobe merger. In relation to my comment about how many graphic designers I know use both FreeHand and Illustrator, David wrote:
You couldn't be more on target. At a press conference following a CEO round-table at the Federal Office Systems Exposition here in Washington last week, I asked Dr. Warnock that very question: who wins here, Illustrator or Freehand? His response echoed your supposition; he said that the new company recognized that most serious illustrators use both products for different things, and that he foresaw no change in the lineup of either half of the new 1,200 pound gorilla.
I also asked him if there would be any attempt to bring in and further develop third-party translation software (such as Altsys's X-Change) and he responded with definite plans of either purchasing those product lines and relabeling them under the new company's banner or developing a new product in house.
As an aside, Michael Dell was also on the panel and spent most of his time pooh-poohing the importance of the PowerPC chip as used in the Power Macs. "With 93 percent of the market Intel based, why would we worry about what's happening with the other 7 percent? It's good for Apple as it extends the life of their line, which they sorely needed, but it's not a market for us."
Much of the press conference was spent on the issues you brought up regarding mergers. Someone asked if the panel saw mergers as a continuing trend; most responded with the shake-out potential as being high. In fact, much was made of the fact that Ad Reitveld, CEO of WordPerfect, was not on the panel as billed. His press people noted "issues revolving around the Novell deal" as keeping him away.
Those of you who read about the Apple/IBM alliance in detail no doubt remember that then-CEO John Sculley proposed that IBM buy Apple outright at one point. I'm sure such a move would have gone over badly in Cupertino, but the rumors I heard toward the end of last week put the IBM purchase to shame. Aldus merging with Adobe has nothing on this deal, and Novell's purchase of WordPerfect and Quattro Pro pales in comparison.
It's quite simple. Can you say "Microsoft?" What better way for Mr. Bill to prove that Microsoft is 100 percent behind the Macintosh?
Yes, that's right. Microsoft is seriously considering purchasing Apple, and although Apple is doing well enough financially, of all the companies in the industry, Microsoft alone has the resources to carry off the takeover. With cash reserves of several billion dollars and a strong stock position, Microsoft may have come up with an attractive enough offer to sway Apple's Board of Directors.
Let me explain why such a deal makes a good deal of sense for Microsoft. As we all know, Microsoft is interested only in the bottom line, and what better way to ensure profits than to control not only the most popular applications that run on a platform, but also the hardware itself? Windows has made a great deal of money for Microsoft because Microsoft controls the game, so to speak. But Microsoft has seldom ventured into the cutthroat PC hardware market, and never at the CPU level. By acquiring Apple, Microsoft suddenly controls not only an operating environment, but a hardware platform that commands 15 percent of the market and that has the recently released Power Macs as a seductive new technology for businesses.
Other benefits abound. Microsoft would pick up Claris in the process, which means that Microsoft Works can fade into the fossil record in favor of the snazzier ClarisWorks. Claris also has a number of programs like Claris Impact and the cross-platform FileMaker Pro that have no competition in the existing Microsoft lineup. As Lotus and Novell enter the software suite game, Microsoft stands poised to change the rules by creating multi-level, cross-platform software suites, all connected via OLE 2.0 for true mix-and match compatibility. Should the deal go through, I'm sure the FTC will have a long look to ensure that Microsoft doesn't take advantage of its position as the maker of both Windows and the Macintosh environments.
And remember, with Microsoft's emphasis on cross-platform tools, how much farther down the road can a core-code version of Chicago (the successor to Windows 3.1) be? With all of that research put into core-code development for Word and Excel, why not use that cross-platform technology at the operating system level? Such an operating system could leave behind the worst of the Windows limitations held over from a senile DOS, and improve in many of the areas where the Macintosh has always led, without fear of a lawsuit. And everyone wondered why Microsoft was happy to license the Windows source code to Insignia for SoftWindows - there won't be any competition from SoftWindows on the Macintosh once Chicago for Macintosh appears with the capability to run all Windows programs at native speeds on the Power Macs.
What's in this for Apple? Well, money, of course, but there's something more - power, and the insanity that set Steve Jobs apart from the rest of the computer industry, an insanity that lives on at Apple. As hard as they try, and as cool as the Power Macs are, Apple's executives have realized that they will never gain more than perhaps a 20 percent share of the personal computer market. That hurts, and although some Macintosh fanatics will liken the sale of the company to selling one's soul to Lucifer, Apple sees it as its best chance to change the world, to mold the world as only Apple could do. An Apple/Microsoft juggernaut, combining Apple's creativity and desire to make a difference with Microsoft's marketing might and attention to the smallest of percentages would result in a company that could easily dominate the entire computer industry. I suspect that appeals to Michael Spindler.
Finally, perhaps the best part about this entire deal, should it ever go through and slide by the anti-trust laws, is that it's complete and utter balderdash. Allow me to wish you all a belated April Fools Day.
by Susan G. Lesch -- firstname.lastname@example.org
This article takes a look at the apparent paucity of female programmers participating in technical computing areas online. We know that women do program on the Macintosh, and some suggest that despite traditional role models and some arguable odds against the possibility for success, women always have numbered among the most brilliant computer analysts, designers, engineers, and authors. Even so, women rarely appear on the nets in technical computing areas. Such a statement is not only difficult to suggest without sounding sexist, but also nigh impossible to document. Still, let's see what we can find out.
Commercial Online Services -- For discussion's sake, take the current figures of 1.5 million CompuServe members and 700,000 America Online members. Programming support for developers centers around the Macintosh Developers Forums, the area of our interest. Personally speaking, I spent three years primarily answering questions in CompuServe's Macintosh Systems and Developers Forums. As my skills grew, I still looked with awe at my colleagues who could correct code and doctor arcane resources. For over four years, between 1989 and 1993, I logged the message boards, looking for female peers, and found they did not exist. Message summaries came back with evidence of 99 to 100 percent male first names. How wonderful it would have been to have found role models back then!
David Ramsey, a programming sysop in CompuServe's Macintosh forums says, "Yeah, it's true: programming is still male-dominated... I don't think it's an issue of whether or not women are accepted in the field - all the ones I've talked to said they never felt intimidated or condescended to by their male colleagues - it's just a reflection of current gender roles in society, where men are expected to be more outgoing than women."
Some may argue that an electronic address allowing only eight or ten characters to identify a person is not gender-specific. Perhaps it would help to point out that the Mac forums of CompuServe require real names and have the luxury of long name fields, which influenced my conclusion. Although "email@example.com" could be either a male or a female, "Chris," "Kelley," and "Bob and Pat" all turned out to be males in CompuServe's Mac Developers Forum when I checked. I must admit, though, that the means to make a truly empirical argument elude me.
Brian Novack, Forum Assistant on America Online's Macintosh Developers Forum, was kind enough to explain the difficulties surrounding determination of gender. "There is no way to accurately gauge such a value. America Online does not ask members to provide proof of gender when issuing a new account and/or screen name. Therefore, unless you conducted a survey of every America Online member who enters the Macintosh Developers Forum by personally contacting them, there isn't any way to find the value you seek in terms of gender. Sorry, it just isn't valid." Maybe, maybe not.
Usenet Informal Survey -- To get the big picture, I recently did an informal survey on Usenet, asking people the question, "In your own experience, how many females [sic] regularly answer programming and technical questions in this newsgroup?" I posted this question to comp.programmer and comp.sys.mac.programmer. I proposed a 1,000:1 male to female ratio as a starting point. This was quickly corrected by an observer who suggested that figures be limited to people of both sexes who regularly participate - by his count, perhaps fifty.
Responses came from people at Apple's Developer Technical Support, Apple's Newton Team, Taligent, Claris, an ex-Microsoft employee, and companies and universities in the United States, United Kingdom, Italy, Germany, France, the Netherlands, Belgium and Australia. One European discredited the survey outright, and another thinks 1,000:1 is close to correct. One pessimist suggested a ratio of 10,000:1. American male respondents said 40 to 50:1 might be accurate. One American, of the four women who replied, seemed to think there were more women than the men did. Avi Rappoport and Jutta Degener were both mentioned, but most respondents could not or did not name a female programmer online. I think now, as was suggested by one respondent after poring over a day's posts to comp.sys.mac.programmer, that 150:1 might be close to accurate.
To arrive at a ratio of any number to one, we need at least one! And that one is rare indeed. Male after male respondent wrote saying Amanda Walker of InterCon is the only female participant in comp.sys.mac.programmer. Although this may be a minor exaggeration, for 1,000, or 100, or 50 answers by men, Amanda Walker is holding up the other end of the fraction. She was generous with her comments about the male to female ratio in comp.sys.mac.programmer, saying her postings are "almost overwhelmingly answers, not questions." Fortune Magazine (07-Mar-94) interviewed Ms. Walker in a feature about the Internet. I would like to thank her for this clear and simple answer to my question above, "Well, there's me."
"My guess would be that it's somewhere in the several hundred to one range, but I'm not sure where. Because of the social dynamics on the net, women are not often accepted in technical newsgroups unless they truly are at the top of their field. Even I myself, with my mind full of random trivia, exercise care to only post an answer when I'm dead sure it's right. Even so I'm occasionally wrong, but it serves to keep my baseline reputation pretty high."
Walker continues, "An interesting comparison would be to take a look at, for example, the proportion of women at the Apple Worldwide Developer Conference. My experience is that there are actually quite a lot of women involved in Macintosh programming, but they tend to maintain a low profile, especially on the net."
Conclusions -- Peter Lewis, author of Anarchie and other shareware for the Macintosh, said in response to my Usenet survey, "It's a sad state of affairs, maybe it'll change one day." With the advent of Women's WIRE, the Women's Information Resource and Exchange (TidBITS #212), I think it time to point out a little joke we have picked up in all the talk about "building" an information superhighway. In large measure, these highways already exist. Both casual and habitual users know the ropes, and they are quickly passing along instructions to new members of online services during the fastest period of their growth in history.
Why do we find such a lopsided situation? Answers tip-toe around sexism, with some suggesting that women avoid areas that allow flaming. It was also suggested that women do not like the intangible qualities of electronic text. But we know that neither theory holds, as there are thousands of women online in areas other than Macintosh programming. I was also surprised to find the notion that the under-representation of women in technical computing online arenas is thought somehow to be men's fault. I should hope there is no need for either sex to coddle women programmers into participating in online exchange on equal terms with men.
Non-profit, non-commercial publications and Web sites may reprint or link to articles if full credit is given. Others please contact us. We do not guarantee accuracy of articles. Caveat lector. Publication, product, and company names may be registered trademarks of their companies. TidBITS ISSN 1090-7017.
Previous Issue | Search TidBITS | TidBITS Home Page | Next Issue