By Joe Barr Originally published in April, 1995 Canopian Embrace 1. an online condition during which the afflicted are unable to stop dialing CompuServe to check for new messages in the Canopus forum. Similar to the "deadly embrace" known to DBAs. Will Zachmann's Canopus Research forum has more message traffic than 98% of the forums on CompuServe. That means it's a very busy place. I was one of the unlucky 5,000 Austinites who lost phone service for a day and a half last week when SW Bell inadvertently cut a cable. When the repair was finally made, my first call was to grab the messages from Canopus that I had missed during my forced detox. Over 750 new messages had been posted in the interim. Thank Baud for off-line readers! I'm not the only Austinite to have discovered this mother-lode of industry insights. IBM'ers like John Soyring, Vicci Conway, and Dave Whittle check in regularly. Ditto civilians like Jep Hill, Scott Swaim, and Steve Shulz. I'm sure there are many others who either lurk (read the messages but don't post any of their own) or who do post but I just don't realize are from Austin. All of us seem drawn to Canopus time after time, looking for the next big story in the industry or the real-scoop on things previously reported. Whether it's the infamous Steve Barkto Incident, the Pentium bug, the London Drugs controversy (more about that later) or the truth behind MS claims for Windows 95, I always seem to read about it first on Canopus. I've had a lot of favorite hangouts in cyberspace. My first was The Source, which later was consumed by CompuServe. Portal was my favorite for a while, and I'll never forget being electronically connected to a couple of dozen friends when we learned about and shared our outrage at the events in Tienamen Square. But when Sprint did away with its flat-rate, unlimited usage and went to an hourly fee for access most Portalites (myself included) became cyber-homeless. Many of us eventually migrated to Bix, primarily because they sought us out and offered a flat-rate access fee. I thought I had found dweeb heaven on Bix. But like Sprint, they only used the flat-rate scheme as bait. When they thought they had you hooked they switched to an hourly charge. The Bix interface was hostile enough to weed out neophytes pretty quickly. People were scorned if they even used the menu system, so you either learned and used the command interface or you kept quiet about it. The interface alone ensured that Bixen could find their way around a computer. It was the on-line electronic acid test. But even so, I met some truly unforgettable people there who have made huge impressions on me. There was Roedy Green, a personal hero, fighting for the rights of the unwashed herd of computer-buying consumers. He raged against rude, insolent, and arrogant install programs, unscrupulous hardware and software vendors, and always tirelessly gave of his most precious commodity - time - to make complex technical issues a little more easily understood. He was the inspiration for me to write the Dweebspeak Primer. Internationally known and recognized professor Hugh Kenner, the most literate dweeb of all, hangs there too. The good professor writes software for fun, to do things like count the number of words in Samuel Beckett's plays. Then he explains the significance of the word count to the play. His tome 'The Pound Era' still explains the mysteries of T. S. Eliot and Ezra Pound to this generation of English majors. There were just too many writers and dweebs of significance to recount them all. Ward Christensen, for example, who set up the world's first BBS and is truly a hero of the PC Revolution. Jerry Pournelle, who writes Sci-Fi novels and columns for Byte magazine, rants and raves there about all manner of things. Oh well, no place is perfect. But that was then. Today Canopus Research is the on-line forum which really has it going on. Will started the forum in the fall of 1992, after giving up his lucrative columns in Ziff-Davis publications in protest of editorial interference by Microsoft. He founded Canopus Research (the name comes from the second brightest star in the sky, in the Carina constellation, which is used as a navigational aid) over a decade ago. At the time, he was Senior Vice President at IDC, Corporate Research. When CompuServe asked him to start his own on-line forum (his section had been the most popular of all the Ziff forums), he decided to use the same name for it. The name fits. Will is an industry analyst who runs several years ahead of the pack. He predicted the downturn in IBM fortunes while it was still the darling of Wall Street. People scoffed. For the past couple of years he has been predicting a resurgence in the mainframe industry and the end of Microsoft's dominance in personal computer software. People scoffed. He says he is not really an IBM fan, or a Microsoft fan, or really a fan at all. He is merely an observer and a reporter. He sees the winds of change shaping the industry and tells us what things will be like next year, or the year after. Each passing month shows him to be a more accurate weather-vane. He also says that when everyone else has jumped on this bandwagon (the ascendancy of IBM and the sinking of Microsoft), he will happily be predicting something different for the future. But Canopus is more than just Will Zachmann. More too than his wonderful crew of sysops (Bruce Bierman, John Oellrich, and Ben Sano) who slave to keep the threads in order and the flames to a simmer. Bruce works for Microsoft, John for AT&T. This is not just a place for OS/2 fanatics, Canopus has both Work Place Shills and Microsoft Munchkins, but there are more fans of OS/2 there than you normally find elsewhere. It's the in place to be, and to be seen. To meet neat, keen people like Steven Durrett, a retired lawyer who maintains a history of almost every message posted. Or Phil Payne, the London-based industry expert for Sievers Consulting, who presents documented facts of the most interesting kind when all about him are posting rumors or misinformation. Like to mingle with writers? How about Jim Fallows and Randall C. Kennedy? How about Peter Coffee and Charles Cooper from PC Week, Dan Gillmor from the San Jose Mercury News, and Nicholas Petreley from Infoworld. Want the latest on the DOJ Vs MS Vs Sporkin? Watch for Wendy Rohm (from Wired magazine and elsewhere) to give us her insights. Have a technical interest in Windows 95? Peruse the posts by Andrew "Undocumented Windows" Schulman. Speaking of Schulman, his recent revelations on Canopus include a ripping to shreds of claims by Microsoft that Windows 95 offers more robust protection from errant applications than it's predecessor, Windows 3.1. He had only recently shown that Windows 95 is sitting atop the same old DOS that the earlier version of Windows did, too, in spite of MS claims to the contrary. Andrew did more than just challenge the claim. He posted the source code for a program so that others could try it at home and prove it to themselves. What he found is that although each Windows 32 application is given its own space in memory in which to execute, and that its private memory cannot be corrupted by other programs, just as MS claims, all Windows 32 apps do share a common area, in which they can all read and write. Contained in the common area are critical system resources which are protected by more robust operating systems like Windows NT, and to a lesser extent, OS/2. In Windows 95, any application running can corrupt the data in the common area and cause other applications or the system itself to crash. Maybe this is why MS vice-president Brad Silverberg said recently that Windows 95 should not be used for 'mission-critical' applications. Tell it like it is, Brad! Naturally, well-written programs are not going to do this, so if all of your 32-bit Windows apps are bug free, you'll never suffer as a result. But one of the most common types of bugs is a wild pointer, resulting in reads or writes to the wrong address in memory. This is the great vulnerability of Windows 95, and very possibly the reason an MS vice-president has to make the sort of statement Silverberg did. Two weeks ago a startling post by a Canadian Canopian claimed that London Drugs, a chain of stores, had an agreement with Microsoft that prohibited them from displaying OS/2 on their shelves. I knew it couldn't be true. Microsoft is hard at work trying to have Sporkin removed as the presiding judge at the hearings on their consent decree. This sort of restraint of trade would blow up in their face at just exactly the wrong time. Still, I decided to track it down. I got the phone number for the store in Edmonton, Canada, and called and asked for the computer department. Color me surprised when the clerk told me that it was true, that they could not display Warp, although they kept it in the stockroom and sold it on request. I asked to speak to a manager. Said manager repeated the same thing, that for 'certain considerations' from Microsoft, they kept OS/2 off their shelves. This was too hot to hold for the next issue of Tech-Connected, so I passed it on to a well-known and highly respected newspaper man. A couple of days later I learned that the store manager was now claiming the story was false, that it was only an 'on-line rumor.' He said they had a copy of Warp sitting in the store window as he spoke. I received mail from Microsoft claiming it was only a shelf-space deal, like the kind that grocery stores make every day. I received mail from IBM asking for the details, and from the PAC called 'Fight MS', so they could check it out as well. I also got a lot of mail from other Canadians who had seen the posts on Canopus and who verified that when they visited their local London Drugs there were no copies of OS/2 in sight. If, like the manger in Edmonton now claims, they are displaying Warp, chalk it up to the power of Canopus. Uh-oh, I just peeked at my last batch of messages. Looks like a storm brewing over 'Out of Resources' messages in the Preview Version of Win95, and another over charges that IBM is lying about the installed base of OS/2. Sorry, I gotta go!