|
Week of 4/12/99 Sunday, April 18, 1999 12:00 A (mostly) daily journal of the trials, tribulations, and random observations of Robert Bruce Thompson, a writer of computer books. |
|
Monday |
[Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] Monday, April 12, 1999 If you didn't read the updates last weekend, check back to last week. I posted quite a lot of interesting new stuff Saturday and Sunday. * * * * * I ended up crapping out yesterday and lying around reading instead of doing my taxes. I've been working seven days a week longer than I can remember, and I just decided I needed a day all to myself. So, today is devoted to taxes. * * * * * I mentioned yesterday my belief that Microsoft is attempting to modify
their revenue model by shifting from pay-once licensing to annual rentals.
In other words, a Windows Tax. It struck me that such a tax actually did
exist at one time. In the 17th and 18th centuries, Britain raised a tax
based on the number and size of windows in a building. Predictably, the
imposition of this tax had two immediate results: people stopped using
existing windows (by bricking them up), and people stopped installing
windows with new * * * * * This from Bo Leuf [bo@leuf.com]:
I think their goal is ultimately more than that. I think they want to be hooked into your credit card and bank to make the process entirely transparent to the user. If annual rental charges are good, monthly ones are better. After all, that helps smooth cash flow, and people are used to paying monthly charges for other services like utilities, cable television, etc. If Microsoft has a coherent long-term strategy, I believe that it is to get themselves so intimately hooked in with users' computers that they become inextricable. In my original reply to Bo, I made reference to the new generation of "WinBoard" motherboards described in Nicholas Petreley's 4/5 column in Infoworld. It made some frightening reading. Or it did, until I realized it was an April Fool's column. * * * * * This from Joshua Boyd [catpro@catpro.dragonfire.net] in response to Bo Leuf's comments 4/11/99:
I remember when dongles first arrived on the scene. They were a response to the inadequacies of key disks and installation counters. In theory, the parallel port dongles were "transparent" but it seldom really worked that way. I remember one dongle, for example, that didn't pass some HP LaserJet commands properly, causing printed documents to format improperly. Then there was the "connect me first" war, where every dongle-based product insisted that it be connected to the parallel port in front of every other dongle. That obviously didn't work very well for people who used multiple-dongle-based programs. I also remember one company that purchased a very expensive ($20,000+) dongle-protected program. They had four people who needed to use the program, but each needed to use it relatively infrequently and for short periods. So, they installed the program on all four PCs, bought an ABCD printer switch box, connected all four PCs to it, and connected their dongle to the common printer port. It worked fine. They didn't think there was anything morally wrong with doing it that way, and I had to agree with them. The program was used probably ten hours a week all told, and doing it that way saved them from buying a separate PC and dedicating it to running that program. In addition to being stolen, the main problem with dongles is that they died without warning. I had several customers whose dongles died for no apparent reason. In at least one case, it took several weeks for them to get the dongle replaced, leaving them without access to that program or to their data. I and my co-workers ultimately decided that any form of copy-protected program hindered legitimate users more than it did people who wanted to steal the program, so we stopped recommending them.
|
|
Tuesday |
Tuesday, April 13, 1999 I had 26 new "real" mail messages from overnight when I logged on this morning. By "real" I mean messages from people I actually know or from readers of my books, as opposed to messages from mailing lists and spammers, which I also had the usual number of. Nearly all of the real messages were personal or book-related ones, but a few belong here. So I'm a bit later than usual getting this posted. * * * * * Taxes are finally done and ready to mail except for writing checks and addressing envelopes. I got through yet another year of aggravation, mainly with the IRS and North Carolina DoR, but also with Intuit and their damned TurboTax package. I hate Intuit with a passion. The most rabid Microsoft-hater hates Microsoft no more than I hate Intuit. What could they possibly have been thinking when they added obnoxious "music" to TurboTax, with no apparent way to turn it off? Every time I started the program, I had to listen to that garbage. It added nothing but aggravation. Intuit's upgrade mechanism needs some industrial-strength fixing as well. For everything that bugs people about Windows Update, it at least gives you options and works most of the time. The same can't be said for Intuit's automatic update program. "One-Click Update" indeed. That was a sick joke. * * * * * Barbara brought me Patricia Cornwell's latest novel, Southern Cross, from the library. I won't put a link to it here, because no one in his right mind would buy it. This is quite possibly the worst book I've ever read, and that's saying something. When I checked the Amazon.com reviews, there were 219 of them with an average rating of two stars, which is by far the lowest average rating I've ever seen for a book by a big-name novelist. One of the reviews even mentioned that there should be a zero star rating available, which I certainly agree with. This book is so bad that it's unintentionally funny. Although she has in the past described herself as a former "computer analyst," Cornwell is completely clueless about computers, email, the web, and similar technology issues. She can't even figure out which direction "downloading" sends a file. Yet she attempts to base major portions of her plot on technical aspects of the web. Her explanations of the technology are ludicrous. For example, she apparently believes that slash-separated components of a URL designate different web servers, and that someone who accesses the HTML page that terminates the URL has bounced through the daisy chain of web servers listed in the URL to get to that document. Cornwell is so clueless that it wouldn't surprise me to find that my 81 year old mother knew more about computers and the Internet, and she's never used either. Cornwell has always been a sloppy writer. Her books are filled with technical and logical errors that could have been avoided with even minimal research. And yet I liked the earlier Scarpetta novels. Reading even those required a willing suspension of disbelief, particularly when Cornwell was blathering on about computers or guns--topics she clearly knows little about--but at least they were readable. It's difficult to see why anyone buys her books nowadays. What's truly incredible is that the woman negotiated a $24 million advance for two books. Her publisher got taken. What's worse is that I can't see any way that these books can "earn out," which is a publishing term that describes what occurs when accumulated royalties on copies actually sold finally reach the amount paid to the author as an advance. I can't see any way, even with paperback sales, that Cornwell's latest two books will ever earn out. That means that the publisher must eat the unearned advance. That in turn means that the publisher will be less generous with advances for other authors, and may in fact cancel books that are already in progress. Ultimately, such huge advances hurt all authors. The mid-list author who counts on a $25K to $50K advance won't be getting nearly that much. A new author, who might formerly have expected a $3K to $5K advance may get nothing at all. We've already seen this happen in the last couple of years with some big-name "celebrity books," which are signed for huge advances and then die in the market. This trend isn't good for anyone in the long run--publishers, authors, or readers. * * * * * This from Bo Leuf [bo@leuf.com]:
I'm sure you're right that if a Windows Tax ever comes to pass people will find ways to get around it. But only some of them, and only to some extent. I really don't have a problem with a Windows Tax per se. Microsoft should be free to license their software on any terms they want. People who don't like the terms are free not to use the software. The way things are going, I suspect I'm not going to like the terms, so I'm going to explore alternatives like Linux. I think Microsoft is between a rock and a hard place. In order to maintain revenue growth, Microsoft must institute new ways of charging for their software. Some ways to do that are with forced upgrades, time-limited software, and by charging for things that formerly would have been free (e.g. Win98 SE). But people hate to shell out money. Just as the US government found that withholding allowed them to charge higher taxes by making collection more-or-less invisible to the taxpayers, Microsoft needs some mechanism that will allow them to take frequent, small, nearly invisible bites out of our wallets. That's why I think their goal is to be able to debit our Microsoft Wallets frequently but in relatively small amounts. People are willing to pay $30 to $50 or more per month for cable television service, so I'm sure that Microsoft thinks we should be willing to do the same for a full suite of Microsoft OS and application software. * * * * * And Bo adds:
Certainly. But the flip side of that is that people will pay for what they want. Monthly cable TV service fees have gotten ridiculous most places in this country, and yet people continue to pay them because they want cable TV. I'd just as soon stop paying cable fees and put the TV into storage. I'd miss watching Buffy The Vampire Slayer and one or two other things, but I could happily live without television. Most people don't feel that way. Same thing with cigarettes. When I moved to Winston-Salem back in 1980, premium cigarettes were selling for $3.67 per carton. I have no idea what they are now, but it wouldn't surprise me if they were $15. But people keep buying them because they want to smoke. $15 may be outrageous compared to $3.67, but it's still a relatively small amount of money to most people, so they pay the price and bitch about it. I suspect the same thing will happen if Microsoft goes to annual rentals. They'd have to start with corporate users, because that market is less price-sensitive and it's a question of needing to have the software to get the job done. You see baby steps in this direction now with volume purchase agreements, annual maintenance programs, etc. Home users will be a tougher nut to crack, but I suspect Microsoft can do it if they focus hard enough on it. But I'll probably be opting out. * * * * * I frequently learn things from my readers. Here's such a message from Ross Fleming [rossflem@serv.net]:
That's certainly news to me. I wasn't aware that there was such a thing as a USB dongle. I don't think we're likely to see them being used for consumer software, however. Dongles of any sort are such a pain in the butt that no software that uses them has ever achieved widespread acceptance. About the closest any dongle-based product ever came to becoming mainstream was AutoCAD, and that's a pretty specialized product. My rule about dongles is "just say no." * * * * * This from Mike Boyle [mboyle@toltbbs.com]:
That's certainly true, but I think it misses the point. Software for heavy iron has always been marketed on a different economic model than mainstream PC software. If Microsoft attempts to extend this mainframe model to PC software, I think they'll find that it doesn't fly. Or at least I hope they will.
|
|
Wednesday |
Wednesday, April 14, 1999 Taxes are done and mailed. That's that for another year. I spent roughly three days doing them, which I understand is about average. The government says that the average is 24.7 hours of work. That's three work days per person. What an incredible waste of productivity. Now I need to get back to work on the books. There's quite a bit of interesting mail today, beginning with several messages from my friend Paul Robichaux, who also writes for O'Reilly. * * * * * This from Paul Robichaux [paul@robichaux.net]:
I've considered getting a dish. But my problem with that is two-fold. First, they block the network affiliates. Barbara watches The Practice, NYPD Blue, and ER and wouldn't want to give those up. We both watch Buffy, which is also network. I know I could do as my friend John Mikol did, and install an old-fashioned antenna, but that involves a lot of cabling and switches if we want to see both dish programming and local affiliates. Or I could continue to pay the cable company for basic cable, which would also require some complex cabling in the house. But that problem is solvable. The real problem is that satellite dish systems restrict you to watching one channel at a time, or two if you buy a dual receiver. We have three televisions in this house. One in the den, one in the bedroom, and one downstairs in my mother's apartment. Although it happens rarely, there are times when all three are in use. I want to be able to tape on one, two, or all three of our VCRs while we watch something else real-time on any or all of the televisions, which would require at least a six-channel amp. I also want to select channels with televisions' remotes rather than by changing them on the satellite controller. The only way to get what I want is for satellite systems to provide a standard broadband feed just as cable systems do now. I don't see that happening any time soon. * * * * * Paul sends this follow-up message:
Thanks. That's useful information. Perhaps I'll re-consider, although it all still seems very complex. * * * * * And still another message from Paul Robichaux:
I'm not sure I believe much that I hear on NPR. You (and they) may be correct about the relative susceptibility of older versus younger viewers to changing their buying habits based on viewing advertising. In the end, that doesn't really matter much. What matters is the advertisers' perceptions of that willingness, and a quick glance at commercials makes it obvious at whom they're targeted. McDonalds doesn't aim much advertising at anyone but kids. And, barring denture cream, dollar-a-week insurance, and adult diapers--things that have absolutely no market with younger people--you don't see many ads aimed at older folks, either. Even the Ensure nutritional supplement (which is surely consumed by almost no one under retirement age) aims its ads at young to middle-aged people. As far as viewer fragmentation, I agree that major advertisers see it as an opportunity. For the same amount of money overall, they can target specific ads on multiple shows to specific markets, rather than running one generic ad that will be seen by a larger number of less well differentiated prospects. We see that now on a show-by-show basis. For example, those shows, like NYPD Blue, ER, and The Practice, that have a primarily upper-middle class viewership get all the ads for luxury automobiles, wines, etc., while shows aimed at teenagers get ads for Nike tennis shoes, Pepsi, and so on. I agree that the idea of running the same program in two versions is likely to catch on. But it spells the death of the ad-laden version in the long term. Those who can afford to do so and are so-inclined--the richer, older viewers--will pay for just the shows they want to watch. In turn, shows they do not watch will generate relatively little direct revenue, and will be forced to depend on an increasing number of ads to make up for the fact that their demographics suck. Soon, such programs are likely to be half or more ads. Or, like the Weather Channel, 100% ads. Have you noticed that the WC runs banner ads even during what is supposedly program content? I think we'll see a lot more of this. * * * * * This from Dave Farquhar [farquhar@lcms.org]:
If Microsoft ever goes to annual rentals, I plan to opt out, too. But it's not going to be as easy as some people seem to think. Microsoft is constantly weaving a tangled web that makes it difficult or impossible to revert to earlier versions of operating systems and applications. I'm not sure how much of it is intentional and how much is an honest effort to integrate applications, but once you start using Microsoft applications, you've essentially walked in the front door of the Hotel California. Getting out of there with your data intact is not going to be easy. And I agree with what you've said about annual rentals aiding the $299 PC concept. Let's face it, with the OS and Office bundles, Microsoft has come to expect a two or three hundred dollar cut from every PC sold. They don't want to give that up, but they can't get anywhere near that amount of money from each cheap PC. So it appears that some sort of periodic payment, whether you call it a rental or not, is going to be necessary for Microsoft to maintain revenues and profits. That's one reason I mentioned the monthly versus annual rental. That way, vendors could sell a $299 PC with "three months free" use of Windows 98 Seventeenth Edition and Office 1901 included. After the first three months, the PC would start transferring $10 monthly payments to Microsoft directly. * * * * * The following from Jan Swijsen [qjsw@oce.nl]:
Well, yes, assuming that they do frequent upgrades. But if they succeed in implementing rentals, they'll be much less inclined to update the software. In its purest sense, rental need not include any support, updates, etc. It would simply be the price you pay to use the software. Much of the feature bloat in software occurred in response to competitive pressures. If WordPerfect had a feature (no matter how useless) that wasn't present in Word, the next version of Word had to have it. And so on. But with Microsoft's dominance in the office productivity market, that pressure has largely disappeared. How many truly new features did Word 97 have relative to Word 95? To the extent that new features are added now, it's because Microsoft wants to provide a carrot that will encourage people to upgrade from earlier versions of the same product. Feature bloat in Microsoft products is now concentrated in markets where they do have competition. They wouldn't be struggling with implementing all these new features for Windows 2000 Server if NetWare 5 wasn't pushing them. If Microsoft implements software rental, they'll be under very little pressure to increase the feature set. My guess is that what maintenance and upgrade effort was expended on rental products would be aimed at reducing bugs rather than adding features. And, again, the whole process would be transparent. Your system would automatically pay Microsoft every month and would automatically download and install bug fixes, etc. somewhat as Symantec LiveUpdate works now. And I think corporate environments would be the first group targeted rather than the last. Although Microsoft can't force people to abandon earlier versions of Windows and Office, they can sure do their best to make it attractive to users to do so. If you want the new features, you upgrade to the new products, which just happen to be rental-based.
|
|
Thursday |
Thursday, April 15, 1999 Tax day here in the U.S. Barbara and I stopped by the library yesterday before going out to dinner. People were clustered around the tax forms. That seems to me to be leaving it a bit late, but what's truly amazing is that people will be getting tax forms until the library closes tonight at 9:00. That leaves them three hours to do their taxes and get them in the mail by midnight. The post office stays open until midnight every April 15th, and has curb service so people can simply drive by and hand their tax forms to a USPS employee to ensure they are postmarked by midnight. Like every year, the library staff will take quite a bit of abuse from people who've left it until almost too late. People cruise in literally two minutes before closing time, expecting to pick up a full complement of tax forms. Usually, they also expect the library staff to offer advice about which forms they need. Obviously, the staff can't do that. Formerly, the IRS provided the libraries with photocopy-able examples of every form. That way, patrons could still get copies of unusual forms that the library didn't stock or ones that they'd run out of. But people who show up two minutes before closing are generally rude as well as stupid. They often get upset because they have to pay ten cents to photocopy a form. I'm sure the library staff spends a lot of time biting their tongues. They're not allowed to explain to these morons that they wouldn't be having this problem if they'd planned a little better. Instead, they're told they have to put up with the abuse. They're even told they're not allowed to close at closing time if there are people still puttering around looking for forms. I'd throw them out at closing time. Teach them a lesson for next year. This year, I understand the IRS provided a CD with forms on it. I'm not sure how much that'll help. I never could figure out why the libraries were passing out tax forms anyway. They're a county government agency. It seems to me that the post office should be passing out the forms, but they don't carry them. Instead, they have a sign referring people to the library. * * * * * Embarrassing moment last night. I need to do some phone wiring in Barbara's office. I no longer have my own butt set and toner, so I called my friend Steve Tucker to ask if I could borrow his. He responded, "you already have them." I was almost 100% sure that I'd given them back to Steve, and told him so. We both did some looking around, and neither of us had them, at least in any of the expected locations. Barbara came to the rescue. She found them in my Lands' End canvas attache case. Urk. "So," I said to Steve, "since I already have them, is it okay if I borrow them?" Fortunately, Steve has a sense of humor. Now I have to call my friend John Mikol and apologize for harrassing him. John borrowed my six foot long bell-hanger drill bit three or four years ago to do some wiring. A year or so ago, I asked him if I could have it back because I needed to do some wiring of my own. He said he'd returned it long ago, which I was almost sure he hadn't. We both turned our houses upside down looking for it, but neither of us found it. He was convinced that he'd returned it to me, and I was just as convinced he hadn't. Six months or so ago, John called to tell me he'd found it. He has his home telephone system controller mounted in the crawl space underneath the main staircase in his home. Doing something else entirely, he was crawling around under there when what should he spot but my drill bit. When he called to tell me he'd found it, I asked him if he'd also found the drill bit guide. He said, "you mean I have that, too?" Come to think of it, John still has them. * * * * * Usually, I wouldn't think of questioning Pournelle's judgement on matters military. But a week or so back, he observed in his View that we could have successfully invaded Yugoslavia with a brigade, but now it would take a corps (two or three divisions). I remember thinking at the time that Pournelle was an optimist in this case. I thought it'd take more like an army group, at least if the Nazi's experience was anything to judge by. The Nazis invaded Yugoslavia with about 200,000 Werhmacht and Waffen SS soldiers in 1941 during Operation Punishment. At the peak, Germany had nearly 700,000 soldiers in Yugoslavia, but never succeeded fully in bringing it under control. In fact, this ill-judged invasion--which was due more to Hitler's pique than anything else--may well have cost Germany the war. All Operation Punishment really accomplished was to delay the invasion of Russia by about five critical weeks. As it was, the German army literally reached the outskirts of Moscow before the Russian winter forced them to retrench. If the invasion had taken place five weeks earlier, as originally scheduled, Germany would probably have beaten and occupied the USSR, leaving them free to concentrate all their resources on Britain. Britain would eventually have been invaded and occupied, leaving the US without a base for operations against Germany. The world would be a very different place than it is now. The morning paper says that NATO has begun to look at what happened during Operation Punishment. It's incredible to me that they hadn't done that already. Perhaps they'll be smart enough to learn a lesson. If the Werhmacht and the SS couldn't do it--and no one could deny that they were some pretty competent soldiers--the likelihood that NATO and the US will be able to do it with a small modern force is small. All we're likely to do is open Pandora's Box. When I said to Jerry that I thought we'd have body bags coming back on the same scale as we did in Viet Nam, he remarked that he was thinking more like Sarajevo 1914. I guess all we can do is hope that sanity prevails. How likely that is with Mr. Clinton making the decisions is uncertain. * * * * * This from Allan Edwards, who asks that his email address not be published:
Well, that question is roughly equivalent to pointing to a group of women and asking me which one you should propose to. There are too many trade-offs, and most of them depend on personal preferences. How much money are you willing to spend? Is getting a small increase in performance worth spending a lot more money? How long do you expect this system to perform at an adequate level, as you define adequate? Do you play heavy-duty games like Quake? How dependent are the applications you use heavily on floating point performance? Do you run an SMP-capable operating system like Windows NT, or are you planning to run Windows 98? And so on. All of that said, here's my take on your question: The Intel Celeron processor offers the most bang for the buck available. You'll be hard pressed to tell the difference in normal use between a system based on a fast Celeron (e.g. the PPGA/Socket 370 Celeron 433) and a system running an AMD K6-III, a Pentium II, or a Pentium III. They're all fast, and the difference between "very, very fast" and "very, very, very fast" is not likely to be noticeable in normal use. The system I use for most of my work has one Pentium II/300 processor. I have any number of faster processors literally lying on the work bench, but I haven't bothered to install one of them. For normal use, fast enough is fast enough. The difference between something taking half a second on the Pentium II/300 and a quarter second on a Pentium III is not worth thinking about, let alone spending much time or money to fix. If I did heavy compiles or something similarly CPU-intensive, it would make a difference. The difference between running a process that takes half an hour on the Pentium II/300 versus 15 minutes on the Pentium III might well make the Pentium III worth installing at whatever the cost. For that matter, those who do iterative tasks (like compiling and linking or recalcing a complex spreadsheet) that take several seconds to a minute or so, but do those tasks repetitively may well find the difference between a 30 second job done a hundred times a day and that same job completing in 15 seconds well worth the price difference for a Pentium III. AMD appears to be in deep trouble right now, with plummeting revenues and profits and an inability to ship anywhere near enough CPUs to meet demand. That doesn't mean that choosing an AMD CPU is necessarily a bad idea. Even if AMD disappeared tomorrow, you'd still have the CPU. But AMD CPUs all run on the obsolescent Socket 7 platform, and there are better alternatives available in Slot 1 and Socket 370. On that basis alone, I'd choose an Intel CPU. The Pentium III does, in one sense, look to the future. Right now, for all intents and purposes, the Pentium III is simply a marginally faster Pentium II. But that will change as software vendors incorporate support for the new SSE instructions. If you are using (or plan to use) software that will be able to take advantage of these instructions (e.g. heavy graphics programs, including games, speech recognition, etc.) then going with a Pentium III now may make sense. Just how much difference the new SSE instructions will make is not clear at this point. I have some Intel benchmarks that indicate that SSE may increase performance by 50% or more for software that is SSE-enabled. But there are few if any commercial applications yet available that support SSE, so what the real-word results will be is anyone's guess. My own guess is that SSE will garner widespread support from software vendors, and that SSE will be very useful for graphics and speech processing applications. But all of that is somewhat speculative, and not on the immediate horizon. In another sense, buying a current Pentium III processor is not looking to the future. Pentium III CPUs you can buy right now run on 440BX-based motherboards with a 100 MHz FSB. Intel plans to introduce a new chipset that will run a 133 MHz FSB (with Rambus memory). But that chipset is not yet available, and it appears that it may be some time before 133 MHz FSB Pentium III processors arrive. If you buy a Pentium III now, you're buying Intel's current top-of-the-line processor, for which you will pay a heavy premium. Unless you are one of those people who has a legitimate need for the latest and fastest processor regardless of cost, the Pentium III is probably not your best choice. So, I guess in short form, my advice to most buyers would be to buy a 400 or 433 MHz Socket 370 Celeron-based system. Or wait another ten days for the Celeron/466 to ship. Spend any money you have left over buying more memory (at least 64 MB for Windows 98 or NT, and 128 MB is better for NT), a bigger monitor, a larger/faster hard drive, or a better printer. * * * * * This from James D. Griffith [ggroup@bellsouth.net]:
I've gotten the Spotmatic battery question several times now, and I did look it up at one point in a 30 year old Pentax book. If you search for Spotmatic on my search page, it should turn up several hits. The latest one lists the battery needed. As I recall, it's a standard drugstore stock item. As far as modern cameras, I'm not the right person to ask. I haven't bought any new camera equipment in at least a decade, and my next camera purchase will probably be a digital camera. Good luck.
|
|
Friday |
Friday, April 16, 1999 Here's a revolting development. The Register ran a story today saying that the search engine AltaVista has announced that it will begin selling placements. By paying AltaVista, a webmaster can ensure that his site is displayed at or near the top for searches on the keywords he chooses. I started using AltaVista the day it opened. For a long time, it was the best search engine out there. Then Northern Light came along, and I switched to using it for most of my searches. AltaVista returned a lot of hits, but its relevance wasn't as good as Northern Light. But now that AltaVista and Northern Light no longer update their indices very frequently I've pretty much given up on both and gone to HotBot. I sure won't be using AltaVista any more now that they're accepting money for placements. I do have an idea for a new slogan for them, though. "All the results that someone's paid us to display for you." Geez. * * * * * Amazon.com is being sued again. This time the suit was filed by a small brick-and-mortar feminist/lesbian bookstore in Minnesota, which happens to be named Amazon Bookstore. They're claiming copyright and trademark infringement, which seems odd. According to Amazon Bookstore, they've only just filed suit because awareness of Amazon.com has only recently become widespread. I guess that depends on your definitions of "recently" and "widespread." This looks like a simple money grab to me. I hope the courts dismiss the suit with prejudice, and point out that if Amazon Bookstore wanted to sue, they should have done so years ago. The likelihood of anyone confusing Amazon Bookstore with Amazon.com has to be pretty small. * * * * * This from Dave Farquhar [farquhar@lcms.org]:
I'm afraid that people who think of an incursion into the Balkans in the same terms as Viet Nam or Desert Storm are in for a shock. If we are foolish enough to invade, I think we'll find that the Balkans will eat divisions and will chew up and spit out armor. I don't think Clinton appreciates just what kind of fight we'll be in for if he decides to put us in on the ground. Even if we succeed in supressing the formal air defenses, every 18 year old kid will have a Strela in one hand and an RPG in the other. We're likely to lose a lot of Abrams and Bradleys, which we cannot replace, and I think we'll find it too expensive in aircraft and crews to sustain any kind of air superiority, let alone the air supremacy we enjoyed in Iraq. If Clinton invades, he's an idiot.
|
|
Saturday |
Saturday, April 17, 1999 This has been a week of plumbing problems. Earlier this week, after we had a minor flood in the kitchen, the plumber came out to replace the Dispose-All in the kitchen sink. The old one had literally rotted through. It was about ten years old, but one expects the motor to die before the casing rots. Then a day or so later, Barbara noticed a pool of water in the basement under our master bath. We thought the shower was leaking, and planned to pry out all the old caulking and re-caulk. But then yesterday she noticed that water was still dripping when neither of us had showered back there in at least 24 hours. We had visions of having to rip the walls out to get to the pipes. The plumber showed up yesterday afternoon. It was a crack in the supply pipe to the toilet, which he quickly repaired for only $53. * * * * * Microsoft has just announced a new program to promote Windows 2000 Server. For $125US, they'll send you (US and Canada only) NFR (Not For Resale) copies of Windows NT Server 4.0 and Windows 2000 Server Beta 3 now, and a full live NFR copy of Windows 2000 Server when it (finally) ships. They also send you a bunch of supplementary programs (e.g. Services for NetWare) and training materials. This is a really cheap way to get live copies of Windows NT Server 4.0 and Windows 2000 Server. You have to answer a short questionnaire, which seems to focus on establishing that you're a consultant or reseller. Once you've done that, you provide your credit card number and they ship the stuff to you. There's currently a two week backlog, which fits nicely with the expected release date of Windows 2000 Server Beta 3. Click here if you're interested. * * * * * And now I'd better get to work on some projects. I promised Barbara to do some telephone work in her office. Right now, the number she has on her business cards is the one we're using for dial-up Internet access. Not good. We actually have only three telephone lines now, down from a maximum of seven. Line 1 is our main home phone line. Line 2 is my mother's main phone line. Both of these terminate directly to the telephone controller, so anyone can use either line for outbound calls. On Line 1, I pay the phone company a buck or so a month for "Call Forward On Busy" service. If Line 1 is busy when a call is placed to it, that call rolls over to Line 2. Neither Line 1 nor Line 2 actually rings on any telephones. Instead, they both ring on my automated attendant, which delivers what sounds like an answering machine message (great for driving off telemarketers). When listening to the message, the caller is given the option to press 1 to speak with my mother (which rings extension 12), 2 to speak with Barbara (which rings extension 18), or 3 to speak with me (which rings extension 11). All of that is fine. The problem is with the third phone line. It has two telephone numbers assigned to it by the telephone company's Distinctive Ring service. Calls placed to the first number ring with a normal cadence, and calls placed to the second number ring with a different cadence, which sounds like the double-ring familiar to anyone who watches British television shows. So, although this is one physical phone line, and can be in use for only one thing at a time, Distinctive Ring allows me to route inbound calls to different locations. From the demarc, I run this line to a Command Communications RD-4000 ring detector. It has one inbound port for the phone line itself, and four outbound ports to which it routes the call depending on the ring cadence. I have outbound port 1 (for the main phone number) routed to CO3 on my telephone system controller. Outbound port 2 (the fax number) is currently routed directly to my office, where it connects to a modem. That modem is connected to sherlock, which runs both Windows NT fax software and WinGate, the proxy server we use for Internet access. That means that line 3 is busy more-or-less all day long. To fix that problem, here's what I plan to do.
Once I'm done, we'll be using our main telephone line for dial-up Internet access. That shouldn't matter much because when it's busy, calls will roll over automatically to line 2. The caller still hears the same greeting message from the automated attendant, and can still ring any of us. This is all a lot of work, but it needs to be done. * * * * * This from Bo Leuf [bo@leuf.com]:
You make a good point about the Nazi's special groups. I've known many former members of the Wehrmacht and the Waffen-SS. All of them were honorable men, and all of them fought to defend their country, as any soldier does. All of them were disgusted when they learned of the horrible things the Einsatzgruppen, Totenkopfverbande, and similar Nazi organizations had done. They claim, rightly in my opinion, that their own names were unjustifiably blackened because of the actions taken by these other groups, actions in which they had no part nor even knew about until after the fact. Contrary to movie lore, none of the Wehrmacht soldiers I've known hated or feared the Waffen-SS. There was some envy, certainly, because the SS always had first claim on the best recruits and on the latest and best equipment. But every Wehrmacht soldier I've ever spoken with about it said that, given the choice, you wanted the SS defending your flank. And, Wehrmacht or Waffen-SS, these were some of the best soldiers who ever took the field. Prevailing US public opinion, at the time and later, was that one American soldier was a match for any two German soldiers. In reality, of course, the situation was almost exactly the opposite. So in the Spring of 1941, up-to-strength with battle-hardened veterans, fully-equipped and on their way to invade the Soviet Union, a significant portion of the German forces took an unexpected right turn into the Balkans and wasted five critical weeks attempting to subdue an area that has never been subdued since Roman times. They failed. And the United States, which would be lucky to be able to field a Corps let alone an Army Group, expects to succeed. We don't even have the advantages the Germans had: short supply lines and a complete lack of concern with world opinion. We go in with both hands tied behind us. The logistics will be horrible, our forces are largely without combat experience, we will fight the war with the world watching on television, and the guy making the strategic decisions bases them on polls. No way. * * * * * This from Dave Farquhar in response to Bo Leuf:
* * * * * And from Bo Leuf in response to Dave Farquhar:
* * * * * Lunchtime: Well, moving the phone lines and bringing up the proxy server on bastet is going much faster than I thought it would. What I thought would be the easy part--connecting the modem on bastet to the existing port 16 jack--turned out to the the hard part. I couldn't find port 16 on any jack in my office. I put tone on one of the dead jacks, went downstairs to my telephone backboard, and used the inductive amplifier to locate the pair the tone was on. Once I did that, it was the work of a few minutes to cross-connect that pair to port 16 on the telephone system controller. After using a regular phone to verify that that jack was indeed now port 16, I plugged the modem in. I was expecting to have to install the modem, but it was already installed in NT on bastet. I was also expecting to have to install RAS/DUN, reinstall SP4, etc., but I'd already done all that as well. I guess I was looking ahead when I installed NT on bastet some time ago. I created a DUN connectoid for my ISP and tried dialing it manually. Everything worked fine, and I was able to use IE to hit several web sites. That done, I started to install WinGate Pro 3.0 on bastet, and found that I'd already installed that as well. Things were going along better than I thought they would. The next step was to configure WinGate on bastet. I fired up GateKeeper, the WinGate administration utility, and it immediately prompted me for a password. I typed the top-secret master password I use for almost everything (horrible practice, but there it is) and GateKeeper refused it. I vaguely remembered that GateKeeper had a default password, but I couldn't remember what it was. After several minutes spent searching WinGate help files, I finally learned that GateKeeper has no password by default, but prompts you to enter one the first time you use it. Okay, I tried to start GateKeeper again, this time using a blank in the password field. It fired right up, and prompted me to provide a new password. It would have been nice if it had simply recognized in the first place that the password I was entering was the one I wanted to use instead of repeatedly telling me that I was entering an incorrect password. Technically, I was, because the password was blank. But as a usability feature, it'd be nice if GateKeeper would recognize what was going on and take me directly to the dialog to allow me to enter a new password. With GateKeeper working, getting WinGate configured as the proxy server for my network took only a couple of minutes, literally. As a test, I fired up IE5 on kerby and changed the proxy settings from 192.168.111.164 (sherlock) to 192.168.111.203 (bastet). I was able to use IE5 on kerby to hit web sites normally. Next, I reconfigured Outlook 98 on kerby to use the new proxy server on bastet. That, too, worked fine. I expected to have to configure the SMTP and POP proxy services on bastet, but when I used GateKeeper to view them, they were already configured properly. Either I did this myself when I installed WinGate Pro 3.0 on bastet a couple of months ago, or WinGate is simply amazing at configuring itself. I tend to think I must have done it, but I'm not completely sure... With IE5 and Outlook 98 configured, the next step was to configure FrontPage 98 to use the new proxy server. That took about 30 seconds, and I'll test its functioning when I publish this. Then I'll head back to Barbara's office to fix her IE5, OL98, and FP98. Once that's done, we'll both be using the new proxy server and the busy signal problem on line three will be gone. I still need to get line three run back to Barbara's office, but for now everything is working okay. Anyone who calls on line three will get our standard automated attendant, but punching two will ring Barbara. Calls on the fax line are still configured to ring to the modem on sherlock, which still has fax software installed, so we won't miss any fax calls. I'll get the rest of what needs done done this afternoon or tomorrow. * * * * * I've been thinking about digital cameras lately. Pournelle and I will be covering them in the book we're writing, and I don't know all that much about them. So I've been doing some research on the manufacturers' web sites and on various sites like C/NET that do comparative reviews. What I've about concluded is that digital cameras are fine if what you're looking for is a $1,000 point-and-shoot. Other than very low resolution--the best of the "prosumer" digital cameras offer perhaps two megapixels--the real problem is that these things are designed more like low-end autofocus 35s than like SLRs. I don't want a $1,000 point-and-shoot. What I want is a $500 to $1,000 body that handles interchangeable lenses, focusing screens, etc. That's not currently an option on any but the professional grade digital cameras. If I wanted to go out and spend $15,000 to $25,000, I could get most of what I want. Kodak makes such cameras, based on Nikon bodies. The problem is that current consumer-grade digital cameras all use a sensor that's anything from 1/4" (6.25mm) square to 1/2" (12.5mm) square. Compared with the standard 1 X 1.5" (24 X 36mm) 35mm frame, that's tiny. The sensor in a digital camera has only 4% to 15% of the area of a standard 35mm frame. The result, of course, is that a 45mm to 55mm "normal" lens on a 35mm camera would effectively become an extreme telephoto on any current consumer-grade digital camera. So the idea of an interchangeable lens digital camera that can use your current 35mm lenses is out of reach until digital cameras begin to use sensors that are about the size of a standard 35mm frame. Although those are available now, they are extremely expensive. State of the art right now in sensors is just under 100 megapixels. A 24 X 36mm sensor with something like 8,000 by 12,000 pixel resolution mounted in a standard SLR body and selling for under $1,000 would be about perfect. That's unrealistic for now, but Moore's Law suggests that we'll see such cameras in the next two or three years. I can wait that long. But I can't wait that long to shoot some photographs that I need for the books and the web site. I'll start digging around in my 35mm gear and see what I can come up with... * * * * * This from Bo Leuf [bo@leuf.com]. I meant to post this this morning, but it got lost among the mess of stuff in my inbox.
I looked at Google some months ago when I first heard about it. I wasn't impressed, but your recommendation caused me to look again. I have to say that I'm still not impressed. I guess choice of search engine is a very personal thing. I'll admit that the results Google returned were relevant, but their database seems to be both small and outdated. I noticed, for example, that the home page of this site was last visited more than two months ago. Just out of curiosity, I did a search for "wakeolda", which is the name of the site that my friend Steve Tucker brought up back in February. Google returned no results, while HotBot returned the site itself and a page from this web site where Barbara mentioned Wakeolda. Worse, I have a set of keywords that I use to do ad hoc tests on search engines, and Google didn't do very well with them. I use these keywords to attempt to locate obscure pages that I know are present on the web. Stuff that HotBot finds easily, Google returns no results for. If I had to pick the key elements that made a good search engine, I'd say that database size, database update frequency, and relevancy of returned results were the three most important factors. HotBot does well on all three, while Google seems to fall down on the first two.
|
|
Sunday |
Sunday, April 18, 1999 I got tired of working on phone stuff yesterday, so I decided to leave things the way they are for now. Barbara's business line is answered by our home automated attendant, but we're now using a different phone line for Internet access, and that's the main thing. We went out for dinner, and then I decided to curl up with a good book. That OnStream DI30 30GB tape drive was nagging at me though. Finally, I decided to go ahead and install it in the test bed platform just to see how it worked. I spent most of yesterday evening and this morning working with it. Expect a report once I have time to write it up, probably next week. * * * * * This from Joshua Boyd [catpro@catpro.dragonfire.net]:
Well, I don't consider what we're paying now to be reasonable. We pay $34.41 per month (not including premium channels) for something like 45 channels, most of which are garbage. The basic cable service around here is something like $7/month, but that pretty much includes only the local stations and some garbage channels. The so-called "second tier" buys us 30 channels or so, only half a dozen of which we ever watch, and boosts the cost by more than $25/month. A lot of that goes to content providers, which I don't like. For example, although we never watch the commercial-laden Turner channels, we're paying probably a couple of bucks a month to Turner for TBS and TNT. That sucks. Both cable and satellite are all about bundling--making people pay for channels they don't watch--and that flies in the face of the current trend towards unbundling everything. What we need is an a la carte menu, where we could choose individual channels we want to watch (and pay for). People could vote with their dollars, and we'd all be better off. I'd happily pay a couple bucks a month for a commercial-free History Channel, AMC, or Discovery Channel, and I suspect millions of others would too. And that couple bucks a month would be more than enough to replace the payments these cable channels receive both from advertisers and from the local cable operators. I'd really like to see "free" television disappear entirely. Let everyone pay as they go. We'll all be a lot better off. As far as search engines, I've gotten numerous messages from people who like Google. I simply don't find it very useful. Perhaps once they're out of beta and can get their database larger and more current that will change. But I don't think so. * * * * * This from Bo Leuf:
Perhaps. I'll keep trying Google, although I usually use multiple search engines for the reasons you mention. As far as how frequently the indices are updated, you're obviously correct that a lot depends on what the spiders have been parsing lately, but my experience has been that AltaVista data averages a lot older than Northern Light, which averages a lot older than HotBot. * * * * * This from Tom Genereaux [entropy@lawrence.ks.us]:
I certainly have nothing against Google, and I do intend to keep trying it. It's simply that I find HotBot more useful for most of my searches. And your point about it being in beta is a good one. I'll keep an eye on it. As far as Linux and NT, I suspect I may one day do as you are doing. For now, most of the software I use is available on NT, but not Linux. If that changes, I'll certainly give Linux a try as a workstation operating system. I shudder at the thought of using Win9x for anything important. * * * * * This from Bo Leuf:
That would be nice, but I wonder how well it would work. In my experience, things intended to modify a product to do something other than what it was originally designed to do are never completely satisfactory. Although in the case of a high-end Nikon or other professional grade camera that supports interchangeable backs, I could see where the entire functionality of the digital camera could be built into the back, which is essentially what Kodak does with their professional grade cameras. Still, I'm sure compromises would be necessary, and the limited market would mean that back would probably cost two or three times what an equivalent purpose-built body would cost. I still think what we really need is digital camera bodies built to accept standard lenses. In fact, given that linkages nowadays are largely electronic rather than mechanical, it might be possible to design a generic body that could accept adapters specific to each lens maker. We could buy the same body, but you'd buy the adapter for Nikon lenses and I'd buy the one for Pentax lenses. Or whatever.
|