Monday, 14 July 2003
7:16 - My den system died yesterday. We'd had bad thunderstorms Saturday night. The power went out a couple of times and finally failed for good about 9:00 p.m. It came back up between midnight and 1:00 a.m. Sunday. When I sat down at the den system yesterday morning, it seemed to be slow booting, but I didn't think much about it. It did boot, but when I started FrontPage 2000 I got a long pause followed by a blue screen. That surprised me, because I haven't had Windows 2000 blue screen on me more than a handful of times since it shipped, and that's on several systems. When I restarted, things appeared to be running very slowly, but they were running. After a few minutes of that, I decided just to work on the system in my office.
Then, yesterday afternoon, I decided to look at the den system again. I figured maybe I needed to run chkdsk or something, because it was still running very slowly. When I rebooted it for about the fifth time, it started to boot but got stuck at the point where the status bar was scrolling across the bottom of the screen before the boot process continued. I decided that the most likely cause of the problem was that the power supply had taken a hit.
I was never entirely comfortable about that power supply, anyway. It's an older Antec 300W unit, pre-ATX12V, that I'd used an adapter with to route voltage from one of the drive connectors to the ATX12V connector on the motherboard. I popped the cover, pulled the old power supply, and installed a PC Power & Cooling Silencer 400W unit, expecting it to cure the problem.
When I restarted the system, I was sitting on the floor, looking up at the monitor on my end table. It started the boot normally, but froze at the BIOS boot screen. Down at the bottom of the screen, there was a SMART error message, saying that the hard drive had failed. It's a parallel ATA Seagate 120 GB unit, which I installed new in that system when I built it a couple of months ago. It's possible that the drive itself was simply defective, but given the electrical storms the previous night, I suspect that the system itself took a hit.
I went off in search of a replacement hard drive, and found that I had zero P-ATA hard drives on the shelf, if you didn't count the old, tiny, obsolete models. I had plenty of S-ATA drives, but not a single S-ATA power cable adapter. Power supplies will start coming with S-ATA connectors later this summer, but for now I'm out of luck. Oh, well.
10:50 - Articles like this completely miss the point. The issue isn't and never has been open-source software versus closed-source software. The issue is open data formats versus closed data formats. Microsoft is arguing OSS vs. CSS as a red herring. As Microsoft is fully aware, the software has little to do with anything. Application compatibility is the issue, and it all comes down to closed formats.
Open formats would mean the death of Microsoft Office, because open formats would mean that competing applications such as OpenOffice.org could have 100% compatibility with documents created in Microsoft Office. The only reason, and I mean the only reason, that Microsoft still has paying customers for Microsoft Office is that competing products are not 100% compatible with documents generated by Microsoft Office. If OpenOffice.org were 100% compatible with Microsoft Office, how many companies or individuals would ever pay Microsoft again for Microsoft Office? Almost none. How many OEMs would continue bundling Microsoft Office with new PCs when they could bundle OOo for free instead? Almost none.
Microsoft Office certainly has some modules that OOo doesn't provide, notably Outlook. Microsoft Office is also a better product than OOo, in the sense that Microsoft Office is better integrated, has more features, is faster, and is a lot less kludgey. Enough so, in fact, that a fair number of people might pay a reasonable price for the Microsoft product, say $25 to $50. But not all that many in the larger scheme of things. Microsoft might sell 10% the number of copies they do now. Considering the bundled price of Office and the mix of retail versus OEM products and prices, Microsoft might do well to average 30% the current price. That means their Office revenues might be 3% of their current level, which is actually about right.
More importantly, though, just as open formats would ultimately make Office a niche product, the domino effect would also cut Windows sales dramatically. A huge number of people right now could happily run desktop Linux, except that Microsoft Office requires Windows. If the need for Microsoft-branded Office disappeared, so too would the need for Windows. Once again, a significant number of people would continue running Windows. It's a better desktop operating system than Linux for most people. But the "Good Enough is Good Enough" factor would ensure that Windows sales also plummeted. Once again, Microsoft might sell 10% the current sales volume of Windows, this time at perhaps 50% the current price. Windows revenues might be 5% what they are now.
Note that I'm not suggesting open data formats as a way to harm Microsoft, although they would undoubtedly have that effect. I'm suggesting open data formats because they're the right thing to do. Right for everyone except commercial software companies. There's entirely too much data being stored in proprietary, closed formats. What should be our data is instead hostage to the commercial software vendors. Even if those vendors were well-intentioned, the result is that we are losing our data. Just try, for example, to open a document created in a very early version of Word. You'll find that the current version of Word doesn't understand the format. In a few years, that data may well be gone forever, not because it's physically unreadable but because no one has software that understands the format.
Open formats don't solve that problem entirely, because open formats can evolve just as closed formats do. But at least with open formats we have the format documented out in the wild. Years from now, if someone cares enough, he'll be able to read antique data files because he'll know the full details of how they were formatted. That's not true for closed formats, and is a very strong argument in favor of abolishing them. Accordingly, I think RFPs, whether from private companies or government agencies, should always require that any software bid should use only open data formats.
Tuesday, 15 July 2003
9:58 - As of mid-month, my overnight mail for the month has been running 37.7% spam, of which SpamAssassin is catching 97.6%. That means about one out of every fifty spams ends up in my inbox, which on some days means a dozen or more. Still, deleting a dozen spams takes only seconds. The spam problem is still terrible in the sense that it overloads the Internet in general and mail servers in particular. But at least SpamAssassin makes it bearable.
What's really going to annoy me about spam is if I have to start paying for it. I saw in the paper this morning that Roadrunner has instituted a 15 GB/month cap, after which they charge extra. I haven't kept track of average message sizes, but with HTML-rich spam and virus-laden mail, I suspect my spams average perhaps 50 KB each. There are days that I get 100 or more 100+ KB virus messages, so 50 KB is probably not unrealistic. Assuming that I get 300 spams a day, which again is not unreasonable, that means that spam/virus messages alone are taking about 15 MB/day of my bandwidth, which I do not appreciate.
I truly wish that someone would start hunting down and killing spammers. I mean that literally. It's the only thing that's going to work. If we woke up tomorrow and found that the top 100 spammers had all been killed overnight, the spam problem would disappear. That's true in two respects. First, because those top 100 spammers are collectively responsible for the vast majority of the spam we get. Second, because the event would be widely publicized, which would make any would-be spammer think about what he planned to do.
If they caught one of the spammer assassins, the trial would be interesting. If I were on his jury, I'd be strongly inclined to look favorably upon a plea of temporary insanity.
Barbara called Factiva tech support, and was told they'd call her back in a few minutes. It was nearly dinnertime, so we went ahead and ate. Barbara left to go to the gym, and I waited for tech support to call back. The guy who called was very nice, but not very helpful. His only suggestion was to install Internet Explorer 6, which I really, really didn't want to do. I did it anyway, and it indeed solved the problem. But at what a cost. I now have IE6 running on Barbara's system. I hate being forced to install a late build of IE, and I let the tech support guy know that. He just didn't get it. He asked me, "What's wrong with installing IE6?" to which I replied, "You mean other than the fact that it's a security disaster waiting to happen?"
I lectured him at some length about how much I resented his company forcing people to use IE. He said that they chose IE because everyone used it. I told him that a lot of people used it, but by no means everyone. There are millions of Linux users who can't run IE, and, with Microsoft discontinuing IE support for the Mac, that adds millions more users who won't be able to run the latest version of IE. And then there are the millions of people who run Opera.
I also explained to him that there was absolutely no need to code IE-specific stuff into their site. They could build it to support a standards-compliant browser like Mozilla, which would mean the site would work with any browser, including IE. He was very polite, but it was pretty clear that he just didn't get it. That's one of the obstacles that Linux needs to overcome. Moronic companies that code specifically for IE.
I pretty much finished building Barbara's new system yesterday. She decided to name it newton, after Sir Isaac Newton, the physicist and inventor of the type of reflecting telescope we use, which is called a Newtonian in his honor.
Newton uses an Antec Sonata case, which is nearly silent. The loudest sound is the Dynatron CPU cooling fan. Not that it's loud for its type, but cooling a 2.8 GHz Pentium 4 takes some air movement, and the 60mm or 70mm fan that fits atop a heatsink has to spin very fast to move enough air. I'm always tempted to build my own CPU cooler. I'd use a standard copper heatsink, but instead of clamping a relatively small fast-spinning fan to the top of it, I'd build a bracket inside the chassis to hold a large fan, 120mm or 150mm, pointing straight at the heatsink. A fan that large could spin at a much lower rate and still move enough air, which means it could be almost silent.
I used an Intel D865GBF motherboard, a Pentium 4/2.8 processor with a 533 MHz FSB (no Hyper-Threading), two 512 MB Crucial PC2700 DIMMs, a Seagate 120 GB S-ATA Barracuda, and a Plextor DVD-ROM/CD-RW drive. As you might expect, it's fast, very fast. Windows is content with 256 MB of memory, and happy with 512 MB, but it's absolutely delirious with joy in 1 GB.
After I got the system running, I configured the boot order to CD-ROM first followed by the hard drive, popped in a Windows 2000 distribution CD and restarted. Windows 2000 Setup ran uneventfully. Windows 2000 didn't recognize a few system components, including the video adapter, LAN adapter, and audio adapter. That wasn't any problem. Once I had a basic Windows 2000 installation running, I inserted the CD-ROM that came with the motherboard and installed the INF file along with video, audio, and LAN drivers. After uncounted reboots, everything was installed an working. The next thing I did was connect to a network volume where I keep installation files and run the Windows 2000 SP3 (ugh) update. I then connected to Microsoft and installed Internet Explorer 6 SP1 (double ugh).
Now that I have the basic system configured, the next step is to install all the third-party software Barbara needs, including WebWasher, AVG AntiVirus, Nero, IrfanView, WinZip, an SNTP client, Mozilla, etc., etc. Once I have all of that installed and tested, the next step is to migrate Barbara's data and configurations to the new system. She understands that there will inevitably be glitches, but I'll try very hard to minimize those. Barbara's system is critical to her work, so I'll do everything possible to make sure she can start using the new system without interruption.
Wednesday, 16 July 2003
7:57 - Barbara now has the fastest production system in the house. Newton is an Intel D865GBF motherboard with a Pentium 4/2.8 and 1 GB of Crucial PC2700 DDR memory, with a 120 GB Seagate Barracuda S-ATA hard drive and a Plextor 320 DVD-ROM/CD-RW drive, all in an Antec Sonata case. I spent a good part of yesterday getting all the software installed, configured, and tested. I did all that with the system set up in the den, so that I wouldn't interfere with Barbara's current system until her new one was ready to go.
The only interesting part was getting her Mozilla Mail data off her old system and onto her new one. Mozilla puts its mail and other data in the folder \Documents and Settings\<username>\Application Data\Mozilla\Profiles\default\<random-string.slt>, where <random-string.slt> is a generated random alphameric string. I searched the registry to locate that string, and couldn't find it. I then checked every Mozilla-related key, and couldn't find anything that specified it. Obviously, Mozilla looks to the Profiles\default directory to locate available profiles. There was already a randomly-named profile on Barbara's new system, which was of course different from the profile on her old system.
So I shut down Mozilla QuickStart on both systems, made sure MOZILLA.EXE wasn't showing in Task Manager, and then deleted the randomly-named profile from the new system and copied the entire randomly-named profile directory from the old system. I then used Notepad to open PREFS.JS and did a global search and replace of <old-random-string.slt> with <new-random-string.slt>. Once I saved that, I fired up Mozilla on the new system. Sure enough, when I opened Mozilla Mail, it went off and retrieved Barbara's new mail.
Once I'd checked everything one last time, I disconnected Barbara's old system and moved the new one into her office. It's complete except that I haven't yet installed her PDA sync software on it. The old system is going to be re-incarnated as a Linux server at some point, although I'll just leave it sitting on the floor of her office for the next month or so until I'm absolutely certain that I didn't overlook anything.
11:43 - I updated Barbara's system just now to Office 2000 SP3 and Windows 2000 SP3. Applying Office service packs is always a pain in the butt. In the first place, why should I have to apply Office 2000 SR-1 before applying Office 2000 SP3? It seems to me that the latest SP should be cumulative. That's the way most other updates work, including other Microsoft updates. In the second place, why should I have to insert the original Office 2000 distribution CDs? Not just the first CD, which might be reasonable to prove I owned the product, but CD1 and CD2. And not just once each, but twice. First when I installed SR1 and then again when I installed SP3. Geez.
The Office 2000 SP3 installation was not confidence-inspiring, either. The whole time it was installing, the dialog told me it was installing "Office 2000 SR1 Premium", which of course I'd installed previously. When the installation finished, it did at least pop up a dialog to tell me that SR3 had been installed. I fired up Word 2000 and did Help->About, just to make sure.
Once I'd finished all that, I noticed that there was a little Auto-Update icon in my system tray, clamoring for attention. I turned off that little rattlesnake immediately. Can anyone really be foolish enough to tell Microsoft to download all new patches and install them automatically, without even asking? At least that wasn't the default, which was to download the patches and ask before installing them. I may be exaggerating, but it seems to me that the chance Microsoft will get a major update like a service pack right the first time are about 50%. Quite often, they end up withdrawing a service pack because it trashes at least some systems. I have numerous "SP-X" and "SP-Xa" files in my installation directory to prove it. The latest SP for W2K, SP4, has recently been shown to trash AutoCAD. Microsoft and AutoDesk are both working on the problem, but as I understand the problem the upshot is that if you run AutoCAD and have installed SP4, you're screwed.
The little Auto-Update icon got me to thinking about trying Auto-Update. I've never used it before on any system. I prefer to do things the manly way, by downloading the files myself and installing them manually. Having Microsoft take control of my computer and install updates as they see fit scares me. But this is Barbara's system, and she doesn't have the qualms about Microsoft that I do. So I decided to run Auto-Update manually.
When I clicked on the Windows Update item in the Programs menu, IE fired up and displayed a bunch of script code rather than running it. Ah, that's because WebWasher was active. So I turned off WebWasher and ran Windows Update again. Never say that Microsoft has no sense of irony...
was the first thing it displayed. Disregarding the important philosophical issues this dialog raises, I told it to go ahead. After downloading something, presumably an ActiveX component, Windows Update brought up a welcome screen in IE. It told me that it was checking my system and that depending on my connection speed it might take a minute or so. Five minutes later, nothing had happened, and it didn't look like anything was going to happen. Eventually, I decided things were locked up, so I exited and then restarted the process. This time, of course, I wasn't prompted to download the ActiveX control, but otherwise things went exactly the same way. I got to the welcome screen, and then sat and waited, and waited, and waited. Nothing happened. After perhaps 10 minutes, I decided to bag it.
Such was my experience with Windows Update. I don't think I'll ever try it again.
11:00 - Another warning from Roland Dobbins, this one critical.
Mozilla Mail 1.4 acts strangely. About every tenth time I open it, I get a blank display, with all of the expected screen elements present, but none of my accounts, folders, or messages visible. The first time that happened, it scared the hell out of me. I closed Mozilla and checked my data folder, only to find that all of my mail data files were still there and of appropriate sizes. I fired up Mozilla Mail again, and everything appeared normally. This has happened repeatedly on two different computers using completely different mail stores, so it's obviously a flaw in Mozilla itself rather than a local problem.
The other odd thing is that Mozilla Mail now retrieves mail as long as Mozilla itself is open, even when the Mail component is not. I used to run Mozilla Mail both on my den system (store mail locally, do not delete from server) and on my office system (store mail on local server, delete from mailserver once retrieved). That way, I always had all my mail available on my office system, and I could use the den system just to check mail after I'd quit for the day.
That's no longer possible, because as long as I have an instance of the Mozilla browser up and minimized on my office system, which I almost always do, the office system keeps retrieving mail (and deleting it from the server). I could, of course, simply tell the office system not to retrieve mail automatically, but I shouldn't have to do that.
Speaking of mail...
Yes, I frequently forget that many of my readers from outside the US and Canada have never had all-you-can-eat plans.
I don't object to developers choosing tools that bind them to Microsoft. That's their choice. But I do object to them choosing tools that bind me to Microsoft. Usually I simply don't use their sites. But there are times when there is no alternative.
Yes, Mozilla allows you to do the same thing, but Mozilla is in general so poorly documented that I hesitated to try that on a system with live data. What I'd really like to do is the same thing I did with Outlook. For example, I kept my data file as f:/usr/thompson/outlook/mailbox.pst. I could then point Outlook running on my office system and Outlook running on my den system to the same data file, so all of my mail was always available no matter which system I happened to be using. I considered trying that with Mozilla, but in the absence of any documentation I decided that was too risky.
I also retarget my "My Documents" to point to another location, in my case a folder on the file server.
And Bob Sprowl announces reality TV, North Carolina style:
No comment needed...
And an Australian reader chimes in on spam:
Nah. In the US, a jury simply returns its verdict, which the judge is free to accept or not. A judge can set aside the jury verdict and impose his own decision, although that seldom happens.
For temporary insanity, the accused basically admits the action but denies legal responsibility for it based on his mental state at the time of the action. It used to be that temporary insanity pleas almost always failed, but in the last several years they seem to have become both more common and more likely to succeed. Or perhaps that's just my impression based on several high-profile cases.
I always thought the US should have adopted the Scots method, which allows a jury to return one of three decisions rather than only two. Guilty, Not Guilty, or Not Proven. The latter means "we think he did it, but the prosecution didn't meet its burden of proof."
Of course, things are changing in the US. It used to be that there was an absolute presumption of innocence, but that is no longer the case. When charged with some crimes, an accused is required to prove his innocence, which is unconstitutional, but is becoming increasingly common. It began with the modification of rape prosecutions. Formerly, someone accused of rape was presumed innocent. In the absence of hard evidence or a corroborating witness, if a woman claimed a man had raped her and the man claimed he had not, the man was acquitted, which was as it should be. Nowadays, a man charged with rape is in a hideously bad position legally. He basically has to prove his innocence, which is often impossible to do, and he has to do so with both hands tied behind him, so to speak.
Recently, the same perversions of justice have become common with such crimes as child or spousal abuse, and in particular for child pornography. It is possible, for example, for a man to be convicted of possession of child pornography and sentence to life in prison without any evidence whatsoever that he had pornographic images of children in his possession. Similarly, Jerry Pournelle recently commented on the miscarriage of justice in which a man was sentenced to life in prison for sucking the toes of a 10-year-old boy. Disgusting behavior, to be sure, but there is no law on the books that makes that action a crime. So, basically this guy has been sentence to life in prison for committing no crime. We no longer presume innocence in this country, and I think we will come to regret that.
10:46 - You can't get there from here.
With Barbara's new system up and running, I decided it was about time to get my den system running again. It's fine except that the hard drive failed during a thunderstorm several days ago. I built the system only a couple of months ago, so it didn't really need anything other than a new hard drive. I checked my stock and found that all I had available was several S-ATA drives. That system has an Intel D845GEBV2 motherboard, which doesn't have embedded S-ATA support. Fortunately, I had a SIIG S-ATA adapter sitting right there with the drives.
So I pulled the dead drive and installed the SIIG S-ATA adapter and a 120 GB Seagate S-ATA Barracuda. That took all of ten minutes, including giving the system a quick cleaning. I then hauled it back into the den and fired it up. It booted normally, and I ran BIOS Setup just to make sure the S-ATA drive was recognized. It was.
I then put the Windows distribution CD in the Plextor DVD-ROM/CR-RW drive and rebooted the system. Windows XP Setup started normally, but when it had finished loading all of its drivers and asked me if I wanted to install, it told me it couldn't find a hard drive. And, yes, I know I said I'd never run Windows XP on a production system, but I need to do this for a project. At any rate, no joy.
Hmmm. I had the SIIG driver diskette right there, but the problem is that this system doesn't have a floppy disk drive installed. Every time I build a new system without an FDD, I know I'm going to regret it. And every time I do.
Hoping against hope that Microsoft had joined the 80's and decided to support loading third-party drivers from CD-ROM during Setup, I restarted the system and pressed F6 when instructed. Setup again loaded all of its drivers and then prompted me to insert the third-party disk driver diskette. It also helpfully informed me that I had no FDD, and so could not continue. Arrrghhh.
So, I have two possibilities. First, Maxtor is sending me some P-ATA DiamondMax Plus 9 drives next week. Second, I could install an FDD in this system and have done with it. I suppose I'll do the latter. I don't really want to be without my den system for the next week or more, and besides it wouldn't hurt me to install XP with a third-party driver just for practice.
I do these stupid things because I'm stupid.
That was written last night. I ended up installing an FDD in the den system and installing Windows XP Professional on it. I played with XP for a while, and I hate it. It kept popping up gratuitous dialogs to tell me things I didn't want to know, and the only way to get rid of the dialogs was to close each of them manually. Geez. The default desktop is ugly, although I fixed that soon enough. I don't see any compelling reason to run XP. I have to have it installed on at least one system to support my book writing. I think what I'll do is install XP on my current main office desktop system once I replace it with the new system I plan to build shortly.
In the interim, I'm going to blow away the XP installation on my den system and install Windows 2000.
I'm also thinking about installing the standalone Mozilla browser and mail client, whatever they're calling them today. My den system would be a good place to test them. I'm not entirely comfortable using a beta product for something as important as my mail, but Mozilla's betas are better than most commercial release versions, and my den mail is secondary anyway.
12:20 - I've been reading all the stuff about the 86-year-old driver in California who mowed down dozens of people. Every time something like this happens, there are calls for age limits on driving licenses, stricter testing of elderly people, and so on.
Although I recognize that many elderly people need to be able to drive to maintain their independence, the simple fact is that many elderly drivers have no business being on the roads. Society has determined that a certain level of impairment is acceptable for drivers. For example, North Carolina, like many other states, now uses a blood alcohol level of 0.08% as presumptive evidence of impairment. That's not a good yardstick, of course. An alcoholic who downs a fifth a day of hard liquor is probably not impaired at all at 0.08%. I, on the other hand, probably average one beer a year. At 0.08% BAC, I'd be entirely incapable of driving. Despite my 230+ pounds of weight, I wouldn't even consider driving a car if I'd had two beers.
I've always thought that instead of breathalyzer instruments, police cars should carry devices designed to measure reaction time. Just as we've set the acceptable BAC at lower than 0.08%, we could set a minimum reaction time as the threshold. If the driver's tested reaction time was above that statutory limit, the driver would be presumed to be impaired, whether from alcohol, medications, or simply age. The advantage to that method is that it is not arbitrary. If your reaction time is sufficiently quick, you are permitted to drive even if your BAC is twice the current limit. If your reaction time is slower than the legal limit, you are driving while impaired, regardless of what happens to be causing your slow reaction time.
I haven't attempted to research the question, but my guess based on observation and experience is that any 86-year-old person will have a reaction time slower than that of a young person who has 0.08% BAC. If that's true, no 86-year-old person should be driving, for the same reason that the young person is presumed to be impaired at 0.08% BAC.
I'm sure that reaction time can vary depending on time of day, blood sugar level, and other variables, but the statutory limit could be set to take that into account. Every police car could be equipped at small expense with a device that measured reaction time. The test itself could be done in a few seconds, so it would be possible to specify that any person being tested could have, say, five attempts, with one or two failures permitted. Jumping the gun on any attempt would automatically result in failing the test, so attempts to beat the test by prediction would fail.
Also, every driving license examining station should have such a device. Just as everyone has their vision tested when renewing their license, they could have their reaction time tested as well. Failing either test means that person doesn't qualify for a license. After, say, age 65, licenses could be issued for only one year terms, with a test required for each renewal.
Taking such a step would get those drivers who are most likely to be involved in accident (or cause them) off the roads. I'm surprised someone hasn't already implemented such a scheme.
12:00 - I've just about finished the build-out on davinci, my den system. Here's what I've installed so far:
The BIOS update was particularly exciting. Immediately after I installed Windows 2000 and the INF update, I noticed that in the install directory I also had a more recent BIOS version. I decided to fire up the Intel Express BIOS Updater. It warned me that once the update started I'd have no video and that under no circumstance should I turn off the system until the BIOS update had completed. It also said the update might take three minutes. After waiting what had to be at least three minutes with no activity, I looked at the clock and noted the time. Ten minutes later, nothing was happening.
I decided the BIOS update process had failed, so I powered down the machine, popped the lid, removed the jumper on the Normal/Configure/Recover, setting it to Recover mode, downloaded the old-style BIOS update, created a floppy disk from it, stuck that floppy in the den system and powered it up. It lit up the Plextor optical drive momentarily and the FDD for several seconds. After that, nothing. I was expecting a lull for a minute or two, followed by heavy FDD access, followed by a beep and a reboot. None of that happened. I powered down the system after ten minutes or so, and tried again. Still nothing.
Hoping against hope, I powered down the machine, moved the configuration jumper to Normal operation mode, and powered up the system. Mirabile dictu, the Intel BIOS boot screen came up, followed by a box congratulating me on completing Express BIOS Update successfully. I have no idea what happened, but I'm not going to question it.
After that, I finished installing all the basic drivers, service packs, and so on, followed by the half-dozen essential utilities. I then started thinking about applications. I decided I might as well look at Mozilla Firebird and Thunderbird, since they're the future of Mozilla. I downloaded Firebird 0.6 and installed it. Well, unzipped it, really, since there is no installation procedure. It simply runs from the location it's unzipped to. After playing with it for a few minutes, I wasn't particularly impressed. There's a lot broken and a lot missing. Not that I have anything against the product. I suspect I'll use it once it's officially released, but the current beta is a bit raw for me. I didn't even bother to download Thunderbird, which I believe is even less finished than Firebird. I just installed Mozilla 1.4, which is plenty good enough for my purposes until the Firebird and Thunderbird products are ready for prime time.
After that, I installed Microsoft Office 2000 along with the accumulated patches. I still need to install a bunch of other applications, including OpenOffice.org and Cartes du Ciel. I'm debating about installing Encyclopedia Britannica 2000. Barbara doesn't much like it, although I find it very useful at times. It's three years old, though, and beginning to show its age badly. I just wonder if the current version has fixed some of the problems in the 2000 version, notably some missing features in the interface and the ability to browse articles rather than simply search for them. I'd really like it if they did two things. One, put the entire article on a single page (or at least offer that option) rather than breaking it down into little MacNuggets. Two, provide an alphabetized list of articles that one could browse and click.
8:25 - Windows Update runs fine on my den system, but I'm beginning to wish I'd never used it. One of the updates it downloaded was a "monitor driver" for the 17" NEC monitor I use in the den. I thought that was odd when I saw it listed, but I went ahead and downloaded and installed it. The only thing it accomplished as far as I can see is to eliminate the 85 Hz refresh setting for the 1024X768 resolution with 32-bit color depth. Now 75 Hz is the fastest refresh I can choose from Windows List All Modes. Ordinarily that wouldn't matter much, but I have an air purifier sitting a couple of feet from the monitor. When the monitor was running at 85 Hz, the fan in the air purifier didn't disturb the image much. Running at 75 Hz, the monitor shimmers when the air purifier is running.
I checked the Intel graphics settings page, and 32-bit color at 1024X768 at 85 Hz is listed as a supported mode. I knew that, because that's what I had been using. But with the new "monitor driver" from Windows Update, I can no longer choose that setting. That seems to me to be a gratuitous change.
Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004 by Robert Bruce Thompson. All Rights Reserved.