Home Daynotes Home Week of 9 February 2004

Photograph of Robert Bruce Thompson Daynotes Journal

Week of 9 February 2004

Latest Update : Friday, 13 February 2004 13:19 -0500


Click Here to Subscribe Buy PC Hardware in a Nutshell, 3rd Edition: [Amazon] [Barnes & Noble] [Bookpool] Visit Barbara's Journal Page

Monday, 9 February 2004

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]
{Five Years Ago Today]


8:23 - As I was saying before I was so rudely interrupted...

Our Roadrunner internet service was down most of the day last Monday. It was intermittent Tuesday morning through Wednesday afternoon. On Wednesday evening, it died completely for all intents and purposes. We finally got service back about 5:00 p.m. yesterday. I posted this on the messageboards last night:

This has been the week from Hell. Our Roadrunner connection died completely early Monday morning, and wasn't fixed until late Monday. There were intermittent problems Tuesday and Wednesday. Wednesday evening, service died for all intents and purposes again. The "cable" light was still on, but I was seeing 70% to 99% packet loss. It took literally five minutes to load a normal web page, if it didn't time out, which it usually did. I couldn't retrieve my email.

I reported the problem Thursday morning, and was assured it was only me who was affected and it'd be fixed that day. First thing Friday morning, it still wasn't fixed, and I spent hours on the phone talking to Roadrunner tech support locally and in Reston, VA. A local tech support lady swore to me that it was only me affected and that they'd have someone out that day without fail. Come 5:30 p.m. Friday, I called and told them I'd been waiting all day and no one had shown up. They said they had no record of a service call scheduled for Friday, but had me down for 3:00 p.m. to 5:00 p.m. on Sunday! I hit the roof, but the best they could do was advance the appointment to 10:00 a.m. until noon on Sunday.

Saturday afternoon another lady called from the local Roadrunner office and asked me to check to see if I was up. She told me that they'd had systemic problems with the Rural Hall node, which we're on, and had had a bunch of people down since Wednesday evening. Somehow that didn't surprise me. They're a bunch of liars. Just me, indeed. She said they'd fixed the problem and I should be okay. I wasn't. I asked if she could get someone out that day, but she said she couldn't. She verified that I was scheduled for a 10:00 to noon appointment the next day, Sunday.

I spent 10:00 a.m. until noon today waiting, and then called to say no one had shown up. They informed me that they had no record of a service call at 10:00 to noon, but did have the 3:00 to 5:00 one on their records. So I waited, gritting my teeth. Finally, two guys showed up about 4:30. They replaced my cable modem, and all is well.

As best I can guess, the problem was systemic. On Monday, they'd had a severe problem with the Rural Hall segment. From what the techs said, it sounds to me as though they replaced some equipment as a stop-gap on Monday, had continuing problems Tuesday and Wednesday, and finally replaced the equipment with new equipment on Wednesday night. That killed everyone with an early model Toshiba 1100 cable modem, which is what we have. My guess is that the firmware in the equipment they replaced was incompatible with the firmware in older Toshiba 1100 modems. They didn't confirm that, exactly, but neither did they disagree. I asked them, "Geez, don't they test this stuff first?", to which they could only shrug. I then asked if they updated firmware transparently across the cable modem connection. They said they did do that sometimes, but apparently hadn't done it this time. Oh, well. I end up down for the last five days out of seven. Everyone has been real nice, but I'm glad my life doesn't depend on Roadrunner reliability.

So I'm well behind on my writing. Not only did I spend a significant amount of time on the phone, but I was much less productive than normal while I was trying to write. I haven't lost a full week of productivity, quite, but it does amount to at least three days' worth, probably more. What I could get done in one hour with Internet service took probably two or three hours without. It's amazing how much extra time it takes me when I can't simply look something up on the Internet.

So it's back to work for me. I'll try to post here periodically, but I'm far behind.

13:56 - I just posted another chapter, Building a SOHO Server, to the subscribers' page.

 

[Top]


Tuesday, 10 February 2004

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]
{Five Years Ago Today]


9:18 - I've been watching the details of the SCO case over on Groklaw for months now. It's kind of like watching a train wreck. At this point, SCO's case seems to be hanging by a thread. The only thing standing between them and disaster seems to be the possibility that a court will accept their twisted definition of a derivative work. SCO has abandoned the trade secrets claim that was the basis of their original suit against IBM. They're now betting everything on convincing a judge that they, SCO, should have ownership rights in all software that anyone has ever added to the System V codebase, regardless of where that software originated.

For example, say that IBM developed a filesystem for use with OS/2. They then ported that filesystem to AIX, the kernel of which is based on System V code. If I understand it correctly--IANAL--IBM's position is that that filesystem is their code to do with as they please. They are not by the terms of their contract allowed to give AIX source code away, but they are entitled to re-use that filesystem code elsewhere if they so choose, including donating it freely to Linux under the GPL. SCO's position, on the other hand, is that once IBM incorporates that filesystem code in AIX, SCO gains ownership rights to that code and can prohibit IBM from using it elsewhere. That has to be the most tortured definition of a derivative work I've ever heard of. And people called the GPL viral! By SCO's definition, they also have ownership rights to OS/2, and probably to every other operating system on the planet.

My own opinion is that the Groklaw denizens are overly optimistic about the outcome of this mess. Most of them seem to think it's a foregone conclusion that IBM will emerge victorious and SCO will go down in flames. But that assumes the court will rule rationally, and as we all know that's anything but a sure bet. I remain optimistic, though, but only because it's IBM that SCO is going up against. SCO has the proverbial tiger by the tail. They can't let go, or they'll be eaten. Nor can they hold on, or they'll be eaten. One way or another, I think SCO is going down, down, down. But it's because they chose to annoy a 900 pound gorilla, not because of the relative merits of the case.

I just ordered some networking gear that should solve several problems. First, I have a Home Theater PC in the den, but no convenient way to connect it to our network. It needs Internet access to do things like download program guides, as well as local network access to do things like share recorded programs with other systems.

My original plan was to run some UTP over to it, but it's on an exterior wall of the den, which makes running cable a pain in the butt. I have some older Intel 802.11a stuff around, but it's not quite fast enough to stream DVD video even if the connection is running at full speed and dedicated to the video stream. So I'm going to use some 802.11g 54 mb/s components to link the HTPC to the rest of the network.

I think I'm also going to put an 802.11g node in the bedroom. I'd like a small "satellite" HTPC back there so that we can watch stuff in the bedroom that we've recorded on the main HTPC. I'm still thinking about how to configure that satellite. I want it to be on all the time, which means it has to be small and quiet. It doesn't need much horsepower, since it won't be encoding. I think I may use an Antec Minuet case, perhaps replacing the standard SmartPower power supply with one of Antec's quieter TruePower power supplies, and a Mini-ITX motherboard. With a large Seagate Barracuda ATA drive, a DVD-ROM, and an 802.11g adapter, that should be all we need to watch recorded videos and DVDs back in the bedroom.

10:36 - The Inquirer has posted an interesting article about the Athlon 64, which quotes PNG says that AMD is "eating its own children" by shipping inexpensive fast Athlon 64 processors that will cannibalize sales of its more expensive models. Peter says that this is a Good Thing in the long run, because it will take sales away from Intel. I'm not so sure. The new Prescott-core Pentium 4 is Intel's ace in the hole. Intel has already announced that they intend to ramp up Prescott quickly, with speeds of 4 GHz by year-end.

I'm not privy to the details of Intel's P4 production, but I suspect that they could easily ship 4 GHz Prescotts today if they needed to. Intel is in the same position now vis--vis Athlon 64 that they were a year ago with Northwood vis--vis the Athlon XP. In both cases, Intel has a core with lots of headroom going up against an AMD core that's pushing its limits. AMD can ship faster Athlon 64 processors, certainly, but each small speed increment is a strain for them, particularly as they are transitioning to a new process. Intel, on the other hand, can ship faster Prescotts any time they want. There are heat issues, certainly, but I'm confident that Intel will address those, if only with the new BTX form factor.

The Northwood P4 caught up with the Athlon XP at about 2.4 GHz, and passed it like it was standing still. I suspect Intel plans the same for Prescott versus the Athlon 64. Stay tuned. You may see 4 GHz Prescotts a lot sooner than you think. Intel's strategy against the Athlon XP was to ship a mid-range and mid-price Pentium 4 that beat AMD's fastest Athlon XP, with still faster P4's available at premium prices. That way, Intel got the best of both worlds, keeping AMD in the trenches of low average selling price (ASP), while keeping the premium-price sales to themselves. I suspect we'll see the same this year with Prescott. Intel will ship whatever speed Prescott it takes to slightly beat AMD's fastest Athlon 64, and price it at Athlon-like levels. They'll then ship faster Prescotts at very high prices

 

[Top]


Wednesday, 11 February 2004

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]
{Five Years Ago Today]


8:22 - I've just sent the following message to subscribers:

Microsoft yesterday announced what may be the worst security flaw ever discovered in Windows. This flaw affects all NT-based Windows versions from Windows NT 4.0 through Windows Server 2003 64-bit Edition. The problem was reported to Microsoft in July 2003, but it has taken Microsoft more than 200 days to issue a patch for the affected DLL!

Microsoft rates this problem as "critical", and says:

"An attacker who successfully exploited this buffer overflow vulnerability could execute code with system privileges on an affected system. The attacker could then take any action on the system, including installing programs, viewing data, changing data, deleting data, or creating new accounts with full privileges."

This flaw is particularly critical because ASN.1 is so deeply embedded in Windows. The ubiquity of ASN.1 allows this flaw to be exploited via numerous attack vectors, rather than the single vector typically exposed by buffer overrun flaws. This flaw potentially exposes Windows to an attack at least as serious as MS-Blaster. Microsoft recommends, and we agree, that all system administrators should immediately download and apply the patch.

For more information, see this page:

 

 

[Top]


Thursday, 12 February 2004

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]
{Five Years Ago Today]


8:52 - Fred Reed strikes again. Although I don't agree with Fred's positions on some things--he's a conservative with libertarian tendencies--he's always worth reading. Fred is to Political Correctness as the kid was to the Emperor with No Clothes. Fred points out what should be obvious to anyone who has eyes. We've come to a sorry pass when the mere act of pointing out the obvious is so fraught with peril that only the most courageous men dare do so. Fred is one of those men.

In one sense, Fred's article misses the point. Fred is actually speaking out on behalf, not of white European males versus everyone else, but of competence--the ability to do and make useful things--versus incompetence. His point remains valid in the sense that competence as measured by the ability to create useful concepts and material things is in fact largely a white, male, European trait. But Fred is neither racist nor sexist. Fred would be the first to acknowledge the contributions of women like Marie Curie and blacks like Percy Julian. Fred does not claim that white men should be valued above all others, but that competence should be valued above arrogant incompetence. And on that we can agree.

So much for the 4 to 6" (10 to 15 cm) of snow that the weather liars claimed we'd get today. It was to start overnight and continue through the afternoon. So far, all we've seen is a cold rain. Barbara was to take her father to the doctor for a follow-up visit today, leaving me alone for dinner, but that will be canceled if we get any frozen precipitation. Most of the forecasts are still calling for a couple inches of snow through afternoon, but I'll believe that when I see it.

Back to work doing heads-down writing.

12:16 - Tomato soup, when heated in a microwave oven, possesses a characteristic I shall deem blurpishness. (I have to keep up my habit of creating new words and phrases for Google to index. That's about a dozen so far, with Heimatsicherheitshauptamt doing better than all the others combined, although Dishwasher Darwinism did make the Wall Street Journal.) If you heat the soup too long, or fail to stir it so that localized hot spots develop, it blurps all over the inside of the microwave. Women cover the bowl with wax paper to prevent blurps, but we men are made of sterner stuff. I set the microwave for three minutes on high, stuck the bowl in, ordered it emphatically not to blurp, and powered up. Sure enough, three minutes later I had a steaming hot bowl of tomato soup and no blurps.

And still no snow.

 

[Top]


Friday, 13 February 2004

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]
{Five Years Ago Today]


9:42 - Friday the 13th falls on a Friday this month...

SCO versus IBM has been described as a mouse versus a 900 pound gorilla or David versus Goliath, but it seems to me it's more like an ant versus a Tyrannosaurus rex and a bunch of his buddies. One of those buddies is Novell, which over the last couple of days seems to have effectively destroyed not just SCO's case against Novell, but SCO's case, period.

IANALTG, but as I understand it SCO has abandoned the trade secret claims that were the original basis of their suit against IBM and instead bet the farm on purported copyright infringements. Novell seems to make a convincing case that SCO does not and never did have the copyrights they claim to have. I do wonder about one piece of evidence that hasn't been submitted as far as I can tell. If you'll remember back when the case began, Novell claimed that SCO had sent them several letters that asked Novell to transfer the copyrights in question to SCO. It seems to me, perhaps in my non-lawyerly ignorance, that Novell could produce those letters as absolute evidence that SCO did not itself believe that it owned the copyrights that are currently in dispute.

So, SCO has abandoned the trade secret claims. All evidence I've seen suggests that they do not own the copyrights. They have never claimed to own any patents that pertain to this case. So what's left? It is a commentary on my mistrust of our legal system that I am not entirely convinced that SCO will lose this case. I certainly hope they will, and I believe they probably will, but I'm not nearly as certain as the folks over on Groklaw appear to be.

I think what has made me nervous in the past is that IBM seems to be going too slowly. I'm now convinced that rather than go for a quick win that might leave questions unanswered, IBM is instead planning to completely, utterly, totally destroy SCO and everyone who had anything to do with this lawsuit. If I'm right, SCO--and, if IBM can pierce the corporate veil, perhaps The Canopy Group--will end up as a grease patch on the sidewalk outside the courthouse. I hope that's true.

I confess that at first I believed that IBM and Novell were simply defending themselves against SCO, and that any benefit to OSS and Linux were simply coincidental. I've changed my opinion, though. It's pretty clear to me now that both IBM and Novell are doing what they're doing not just to defend themselves but to raise a very high defensive wall against anyone, including Microsoft, who might in future attempt to attack OSS and Linux through the legal system. If that's true, IBM and Novell are proceeding slowly because they want to expose all of the issues in question, including the validity of the GPL. I think it is true, and therefore I expect that once IBM and Novell have disposed of the suits filed against them by SCO, they'll continue by pursuing SCO for violating the GPL. It's going to be an interesting year.

13:19 - Here's an article worth reading about e-books and copyright. I copied it to my site because it's public domain and because I'm tired of stuff I've found on the Internet disappearing. Not that that's likely to happen with this document, but redundancy is the key to keeping data safe and available.

 

[Top]


Saturday, 14 February 2004

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]
{Five Years Ago Today]


 

 

[Top]


Sunday, 15 February 2004

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]
{Five Years Ago Today]


 

 

[Top]


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