Daynotes Journal Week of 2 June 2003 Latest Update : Friday, 06 June 2003 11:30 -0400

Monday, 2 June 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

10:22 - This is a truly crappy way to start the week. I went over to visit my mother at the nursing home this morning. I happened to mention that La Toya, one of her favorite aides, would be back from vacation this week. Mom told me that La Toya had been taken ill during her vacation and was now in intensive care. Mom didn't know any details, but said something about a blood disease. She also mentioned that when she'd first started living at the nursing home she'd thought La Toya was new, but La Toya had told her that she'd worked there for a while but had been in the hospital recently for a long stay. Mom wondered if the two hospitalizations were related.

On my way out, I stopped to talk with one of the senior staff members. She told me that La Toya had a serious problem with platelets, that they kept giving them to her, but "she couldn't keep them". As a result, she bleeds uncontrollably. That sounded like hemophilia to me, but the staff person told me it had some long name that she couldn't remember. When I said something about hoping that they would get La Toya feeling better soon, she told me that La Toya wasn't expected to live.

La Toya is a wonderful aide and a wonderful person. She's always stopped in frequently to check on mom and make sure she was comfortable, even when mom was not one of her assigned patients. She'd bring little presents for mom, like a box of Cheezits, that cheered mom immensely. If there is such a thing as an ideal aide, La Toya is it. To make things even worse, La Toya is raising a 3-year-old daughter by herself. I'm not good at guessing the ages of young women, but I'd guess she's only in her mid-20's. This just sucks.

Of course, the staff aren't telling the residents anything much. Mom told me that she kept asking about La Toya, and the staff would just tell her they didn't know anything more. I'm sure the management told the staff not to give any details so as not to upset the residents. But I thought mom deserved to know, so I called and told her what I'd found out. She's very upset, of course, but I'm a firm believer in not trying to hide things from people. I'm sure the contrast strikes mom. Here she is, 84 years old and in good health other than her arthritis and fragile bones. Here La Toya is, 60 years younger and apparently not expected to live much longer. Mom has often said that she's lived long enough and never really wanted to live past 80, so I suspect she would trade places with La Toya if she could.

What a crappy way to start the week.

14:15 - I was about to build Barbara's new system this afternoon when I ran into a problem. I'm using a Pentium 4/2.8 processor with a 533 MHz FSB in her new system. It makes no sense to use PC3200 memory on a 533 MHz FSB, so I checked my supply of memory for PC2700 modules. I found only two Crucial modules, one a 256 MB and one a 512 MB. I'd thought I had two of each, but then I remembered that when I built my new 845-based den system I'd used one 512 MB and one 256 MB PC2700 module.

With the advent of the Intel 865/875-series chipsets, I'm going to have to get myself back to thinking in pairs when it comes to memory. I should have thought about it when I built my den system, but it just never occurred to me that I'd need matched pairs the next time I built a system. So now I need to tear down my den system, pull the 512 MB module, replace it with a 256 MB, and stick both the 512 MB modules in Barbara's system. My den system won't be much the worse for having "only" 512 MB instead of 768 MB, and Barbara's system can use the full gigabyte. It's overkill for now, no doubt, but Barbara tends to use her systems for quite a while before we get around to replacing them. Her current system, for example, is a Pentium III/1.0G.

Note to self: don't break up matched pairs of memory modules without a very good reason.

[Top]

Tuesday, 3 June 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

10:43 - John Dvorak is generally clueless, but his latest column, Killing Linux, reaches a new peak in cluelessness, even for him. Anyone who was not familiar with the case and read only Dvorak's blatherings might think that SCO had the whip hand here.

It's pretty clear from reading his column that he either didn't read or didn't understand any of the main documents that pertain to the SCO case. The public statements made by SCO have been confused and self-contradictory, sometimes within a single document, which Dvorak fails to mention. Dvorak doesn't even understand what SCO had been selling, as he makes clear by writing "Now if we look at SCO in this melodrama we see a company that went from 100-percent Linux sales to 95-percent Unix sales ..." Although he mentions Novell's statements in passing, he ignores many of the most pertinent apparent facts, including that SCO itself distributed any code in question under the GPL and that SCO licensed Lindows to distribute the code, which it subsequently did under GPL.

Nor does Dvorak comment on SCO's refusal to reveal the supposedly infringing code. Their reasons for refusing to do so are laughable. SCO first claimed that if they gave details about the supposedly infringing code, all the Linux bad guys would hurry to rewrite it and conceal the supposed theft. Right. As though there aren't millions of CDs out there with that Linux source code visible to anyone who cares to look. SCO later changed their story and came out with a confused statement that seemed to say they couldn't reveal the source code in question either because contractual terms prevented them from doing so (a contract with whom? I thought SCO was claiming all rights in Unix) or because of trade-secret concerns (if stolen code is in fact in Linux, then it is already public and not a trade secret).

Furthermore, Dvorak shows a complete lack of understanding about OSS software in general and Linux in particular. Even if SCO wins completely, which is so unlikely as to not be worth considering, that would have very limited impact on Linux in the US and no impact on Linux outside the US. Any code found to infringe SCO's rights would simply disappear from Linux in a very short time, replaced by pure GPL code, and Linux would continue its growth.

Another issue that Dvorak fails to consider is that if Linux is found to contain code similar or identical to code in SCO products, that does not prove that Linux has stolen SCO code. If common code is found, it is far more likely that SCO code contains code stolen from Linux rather than the converse, and SCO may have a very hard time persuading a court otherwise. Linux code is open, documented, and traceable. Anyone can easily establish when particular code was incorporated in Linux, whereas SCO with its closed code may have a very hard time proving that it in fact produced any code in question.

My own guess is that SCO is going to lose, lose big, and go down in flames. It's unfortunate that those damaged by SCO's claims will have little recourse, because SCO will have no assets against which claims can be made. Once IBM and Novell's lawyers finish with SCO, it will be nothing but an empty shell. And Linux will just keep chugging along.

11:07 - Oh, yeah. Here's something to put on your calendar: Federal Trade Commission Announces Accelerated Roll-out Schedule for National “Do Not Call” Registry Set for July. On or about 1 July, you'll be able to put your telephone number on the federal do-not-call list via their web site. You can also do it by telephone, starting around 1 July for western states and a week or so later for eastern states. You'll want to get your telephone number listed as soon as possible, because the first consolidated list goes out to telemarketers in September, and is to be incorporated in their lists on or before 1 October.

Having your number on that list should pretty much eliminate all calls from the telemarketing weasels. Unfortunately, there are some gaping holes in the new law. It exempts non-profits and political callers, for example. The reasons for that are stupid. Most people no more want tele-weasel calls from non-profits than they do from for-profit companies, but the non-profit lobby succeeded in getting themselves exempted. Of course, the politicians exempted themselves as well, even though many people hate being bothered by political calls more even than sales calls. The politicians claim that they had to exempt political calls for First Amendment reasons, which is a crock. The politicians have a First Amendment right to present their views, of course, but nothing in the First Amendment says I have to listen to them.

There's also an exemption for companies with whom someone has a pre-existing relationship, which I think is also a crock. Simply because I've bought something from L. L. Bean doesn't mean I should have to accept tele-weasel calls from them. There are way too many holes in this new law, but it's better than nothing.

As for me, I'll continue to use my current method. When I answer the phone and someone asks to speak to "Mr. or Mrs. Thompson" or says, "Mr. Thompson?" in a questioning tone of voice, I always say, "Hold on a minute." and put down the phone. A minute or two later, I'll hear the obnoxious phone company re-order tone that tells me the caller has hung up. At least I cost them some time. If everyone did the same, the tele-weasels would find it unproductive to bother people.

[Top]

Wednesday, 4 June 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

11:40 - The A-10 "Warthog" is the modern equivalent of the B-17. It takes a licking and keeps on ticking. My dad flew B-17's over Germany during WWII. I remember the stories he told of severely damaged B-17's that brought their crews back. One of his stories was personal. My dad's crew took a brand spanking new B-17 on its first mission over Germany. When they returned, their crew chief was beside himself because of the battle damage. The plane that had left in pristine new condition now had more than 2,000 holes in it (they counted). For all of that, it brought my dad and the rest of the crew home safely. What was truly amazing was that none of the crew even had a scratch.

By all accounts, the A-10 Warthog is equally tough. The Air Force won't risk its F-16 and other expensive air-superiority fighters doing ground attack. They're too vulnerable to ground fire. A so-called "golden BB" can bring down a $30 million warplane, let alone a hand-held SAM. The A-10 is a different story. Its pilot sits in a titanium "bathtub", and every possible system has redundant redundancies. The A-10 was designed from the start to do close air support, getting down and dirty. If you asked the Iraqis who fought in either Gulf War which warplane they feared most, they'd almost universally tell you the A-10 was their nightmare.

When I was in college, I knew a guy who'd fought with an SS panzer division on the Eastern Front. I remember him talking about the Soviet IL-2 Sturmovik, which scared the hell out of them. He told me of his first encounter with a Sturmovik. As the Sturmovik rolled in on a strafing pass on his convoy, he and his buddies turned their MG-42 machine guns on it and started hosing it down. Cumulatively, they were firing more than a hundred rounds a second of armor-piercing incendiary tracer at the Sturmovik. He said they could see the sparkles on the body of the aircraft as their rounds struck it, but nothing happened. The Sturmovik scared the Germans because of its reputation. You could hit it, but you couldn't hurt it.

The A-10 Warthog has much the same reputation among our enemies, which is one good argument against the recent decision to decommission the A-10. When A-10s show up over the battlefield, the bad guys keep their heads down. Bad guys who are hiding aren't shooting at our guys. The A-10 is a very good thing. So of course the Air Force is getting rid of it. I suggested to Pournelle that the best solution would be to take the A-10s away from the Air Force and give them to the Army. Army Aviation knows what to do with them. While we're at it, we should build another thousand A-10s and sufficient spares stocks to keep them flying for the next several decades.

Which brings me to an interesting article about a warthog that got its pilot home. As inspiring as the tale is, I have to wonder about the long-term survivability of a culture that puts its young women on the front lines. In a biological sense, young men are expendable, but young women are not. Certainly there are situations in which women have to fight, but it disturbs me that we are intentionally putting young women in the line of fire. Women have no place in combat, period.

I don't dispute that a young woman might be an excellent combat pilot. On average, she may have faster reaction time than a young man, and she is able to resist high g-forces better. I'm even prepared to believe that a few young women may possess the high level of aggression required of a combat pilot. As a poster I remember from college that showed two buzzards sitting on a branch had it, "Patience my ass. I'm gonna kill something." But it still makes no sense biologically to put young women at this level of risk when it can be avoided. It is the job of men to protect young women like Kim Campbell and Jessica Lynch, not to sit back and watch them risk getting their asses shot off to protect us.

Spam check:

Monday - I received 317 emails overnight, of which 85 were spam, or 26.8%. Of those 85 spams, SpamAssassin caught 85, or 100.0%.
Tuesday - I received 310 emails overnight, of which 77 were spam, or 24.8%. Of those 77 spams, SpamAssassin caught 75, or 97.4%.
Wednesday - I received 345 emails overnight, of which 113 were spam, or 32.8%. Of those 113 spams, SpamAssassin caught 111, or 98.2%.

For the week to date in overnight mail, 972 emails, of which 275 were spam, or 28.3%. Of those 275 spams, SpamAssassin caught 271, or 98.5%.

I've been doing a bit of benchmarking on the 845, 865, and 875P boards. I won't bore you with the numbers, since they don't really matter much anyway, but my conclusions are as follows:

• The 865 dual-channel boards show a noticeable improvement in memory performance over the 845 single-channel DDR boards. Memory performance is only a small part of overall performance in most applications, so you probably won't see a great deal of difference between an 845 system and an 865 system in normal operation. If your application is memory-intensive, the 865 board may show a marginal performance boost over the 845. I can't imagine that many people would find upgrading a 845 DDR system to 865 to be worthwhile. The 865 is the way to go for new systems, however.
• The 875P board does show a marginal memory performance increase relative to the 865. As far as I can see, the 875P is simply a speed-binned version of the 865 that allows using slightly faster memory timings. The only other benefit of the 875P relative to the 865 is that the 875P supports ECC memory, which may be an issue if you need gobs of memory. Otherwise, go with the 865.

15:06 - I've just sent the following message to subscribers:

-------- Original Message --------
Subject: Critical Microsoft security flaw
Date: Wed, 04 Jun 2003 15:01:12 -0400
From: Robert Bruce Thompson
To: TTG Subscribers

Today Microsoft announced yet another have-your-way-with-me critical security flaw in Internet Explorer. For details, see:

<http://www.microsoft.com/security/security_bulletins/ms03-020.asp>

and

<http://www.microsoft.com/technet/treeview/?url=/technet/security/bulletin/MS03-020.asp>

Note that Microsoft rates this problem "critical" for all versions of Internet Explorer from 5.01 through current running on any Microsoft operating systems except Windows 2003 Server, for which severity is rated "moderate". Presumably versions of Internet Explorer earlier than 5.01 are affected as well, but Microsoft does not support or comment on such problems.

The "critical" rating means Microsoft considers this problem to be of the utmost severity. A problem rated only "important" is described as:

"A vulnerability whose exploitation could result in compromise of the confidentiality, integrity, or availability of users’ data, or of the integrity or availability of processing resources."

A problem rated "critical" is described as:

"A vulnerability whose exploitation could allow the propagation of an Internet worm without user action"

This is another of the Microsoft security flaws that requires nothing more than viewing or previewing a malicious email or visiting a malicious web site. Applying the patch sounds like a good idea.

[Top]

Thursday, 5 June 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

[Top]

Friday, 6 June 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

[Top]

Saturday, 7 June 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

[Top]

Sunday, 8 June 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

[Top]

Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004 by Robert Bruce Thompson. All Rights Reserved.