Home » Daynotes Home » Week of 20 June 2005

Photograph of Robert Bruce Thompson
Daynotes Journal

Week of 20 June 2005

Latest Update: Friday, 24 June 2005 14:00 -0400
Free Speech Online - Blue Ribbon Campaign

Paying for this Place
Visit Barbara's Journal Page

Monday, 20 June 2005
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

08:42 - Some discussion over on the messageboard about drunk driving. I repeated the comment I've been making for years, which is that I think drunk driving should be legal, but drunk wrecking severely penalized.

I say that not because I'm in favor of drunk driving, but because I think my suggestion, if enforced, would be more effective at eliminating it. As things stand, in most jurisdictions, one doesn't risk much by driving drunk. Fines, increased insurance rates, and other civil sanctions, certainly. But the possibility of jail time is pretty small, and that's likely to happen only if one kills or seriously injures someone.

Under my proposal, if you drive drunk and aren't involved in an accident, nothing bad happens to you. But if you cause an accident, you are in deep, deep trouble. Serious jail time trouble. Perhaps even death penalty trouble.

One of the problems with the current law is that it is inequitable because it uses blood alcohol level as presumptive evidence of drunk driving. In North Carolina, as in most states, a 0.08% BAL is presumptive evidence of drunk driving. In other words, if you blow a 0.08%, you're guilty of drunk driving, whether or not you're impaired at that level. Contrary to common belief, in nearly all jurisdictions, you're not necessarily off the hook if you blow a 0.07% or lower. The presumptive level is just that: if you blow a 0.08% you're presumed guilty, but you can still be convicted even if your BAL is lower, perhaps much lower. But in practical terms it doesn't usually work that way. If you blow a 0.07% or lower, you're generally acquitted, if indeed you're even charged in the first place.

But people differ. I drink maybe one or two beers every year or two. Even though I weigh 240 pounds, I can usually feel the effect of one beer, and certainly the effect of two. I wouldn't even consider driving after drinking two beers, because I know that I'm significantly impaired, even though my BAL is low enough, perhaps 0.02%, that I'd be considered legal in any US jurisdiction. But a serious drinker at a much higher BAL would be much less impaired, if at all, at four times my BAL.

Years ago, there was an interesting demonstration put on by MADD or some similar organization. They got NASCAR Winston Cup race driver Kyle Petty to volunteer. They set up a slalom course with traffic cones. Petty started out cold sober, and ran the course in his race car as fast as he could, trying not to knock over any cones. Then he started drinking shots of bourbon and blowing into the Breathalyzer machine.

The whole idea was to demonstrate that even an immensely skilled driver would become significantly impaired as his blood alcohol level climbed. They expected his speed through the slalom to decrease, and that he'd knock over more and more cones as he got drunker. Unfortunately, it didn't work out that way. The drunker Petty got, the faster he made it through the slalom, and the fewer cones he knocked over. He passed 0.10% and just kept getting better at running the slalom. Eventually, they called it quits, not having gotten the results they were hoping to get.

My point isn't that drinking improves your driving skills, although that was clearly true for Mr. Petty. My point is that BAL is a very poor predictor of driving ability. In the days before the widespread availability of Breathalyzer machines, cops did ad hoc tests that in fact were much better evidence of impairment. Balancing on one foot with one's eyes closed, extending one's arms and then bringing them in to touch one's nose, that sort of thing. Obviously, those types of tests wouldn't always be practical after a serious accident. But they could measure BAL, reproduce it later once the suspect was fully recovered physically, and then prosecute (or not) based on the results of those tests.

I'm hard at work on the new book. As usual, I'll be posting draft chapters on the subscribers' page. Speaking of which, thanks to all of you who've subscribed or renewed recently. If you haven't yet subscribed or if it's time to renew, see here.


Tuesday, 21 June 2005
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

09:55 - Our poor UPS guy delivered a whole bunch of boxes yesterday evening, including a bunch from Antec. I have a Sonata II, a P180, and a TX1088, along with various new Antec power supplies, including the Phantom 500. I've requested an Overture II as well, but it hasn't arrived yet.

Obviously, I haven't had a chance to do anything with the new cases yet, but they're very nice looking. Interestingly, the Sonata II substitutes a 450W SmartPower unit for the TruePower 380 in the original Sonata. The additional 70W of power is nice, as are TAC and the other enhanced features in the Sonata II, but I'm wondering why Antec chose to bundle one of their mid-range SmartPower units rather than their premium TruePower. Probably a matter of keeping the Sonata II at a reasonable retail price point, I suspect. Not that the SmartPower units are bad by any means. They're solid mid-range units, and the new models have the same 80,000 hour MTBF as the TruePower units. The main difference is regulation, which is much tighter than the spec requires on TruePower units--typically 3%--versus the 5% on the SmartPower units.

But that's a minor nit. Overall, these new cases appear to build on the foundation that Antec has established, with numerous upgrades for cooler and quieter operation. In a technical sense, the most interesting new Antec case is the P180, which isn't, as many people might assume, an upgraded version of the P160. The P180 has a steel chassis rather than aluminum, and uses the composite panels that Antec introduced with the Aria, which sandwich a plastic sound-deadening layer between two layers of thin metal.

I'm also looking forward to building a system around the Phantom 500 power supply. The original Phantom 350 was (and is) dead silent, since it contains no fan. But the 350W output limited it to supporting slow to mid-range systems, while the Phantom 500 should support the fastest components available. I plan to build a fast, silent system with this power supply, quiet drives, and a quiet/silent CPU cooler from Zalman or ThermalRight.


Wednesday, 22 June 2005
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

11:19 - I asked one of my contacts at Antec about their decision to use a SmartPower power supply in the new Sonata II rather than the TruePower power supply used in the original Sonata. She confirmed my guess, that Antec wanted to price the Sonata II the same as the original Sonata and because the Sonata II case itself costs more to produce they substituted the less expensive SmartPower power supply. She emphasized, though, that their new SmartPower 2.0 units are significantly upgraded in terms of reliability and noise level, and that they're quite pleased with the combination.

One interesting thing, though. If I'd had to guess, I'd have said that the product design folks and engineers probably argued for continuing to use the TruePower power supply and pricing the Sonata II higher than the original Sonata, and the marketing folks argued for keeping the same price and using the SmartPower unit. In fact, it was the converse. The marketing folks really wanted to have the TruePower in the Sonata II. It was the product development folks who argued the case for using the SmartPower. What an interesting company Antec must be. It appears that their engineers also think about marketing issues and their marketers also think about engineering issues. That's as it should be, of course, but it's almost unheard of.

And, speaking of power supplies, this from Chris Christensen:

-------- Original Message --------
Subject:     Power supplies
Date:     Tue, 21 Jun 2005 10:38:58 -0500
From:     Christensen, Chris (Aspen Research)
To:     Robert Bruce Thompson

Bob: I know we've gone around on this before, but in the absence of any data to the contrary, SPCR's claim that "Our most powerful P4-3.2 Gaming rig drew ~180W DC from the power supply under full load -- well within the capabilities of any modern power supply." http://www.silentpcreview.com/article247-page1.html has to stand.

Unless you're going to use Intel dual core chips, that ~180 watts should hold up well, with an AMD processor using less.  At any rate, the argument is moot without any instumentation.  Consider the usefulness of having over-all power consumption data for systems: http://www.p3international.com/products/special/P4400/P4400-CE.html these cost about $30.

I don't, as a rule, shut my home computer off.  If my system (monitor, computer, printer) averages 200 watts power consumption, that's 1750 kilo watt hours annually, or ~$125 at my current incremental power bill rate ($0.0723/KWH).  If/when the petroleum situation explodes in our face, we could easily see that bill rate do a 4x, or more.

That sounds about right, but it's not in conflict with what I said.

In the first place, most power supplies are rated at 25C, when their actual operating temperature is 40C or even 50C. At those higher temperatures, power supplies may deliver 50% or less of their rated wattage. Many power supplies rated at 500W actually deliver less than 250W when they come up to temperature.

In the second place, power supplies are both inefficient and less reliable when they're operating at a high percentage of their rated output. A power supply that's delivering 50% to 70% of its rated output delivers more stable voltage, runs cooler, and lasts longer than one that's operating at 90% or more of capacity.

In the third place, their gaming rig isn't really a power hog. Driving a fast Intel processor at full load can draw 130W or more, and a high-end video card can easily draw another 90W or more. Add in current consumption for memory, drives, etc. and you can easily exceed 250W DC draw.

I do have testing equipment, by the way.

And some frustrations with the latest Xandros release:

-------- Original Message --------
Subject: Disappearing hard drive under Xandros (redux)
Date: Tue, 21 Jun 2005 17:31:13 -0600
From: Captain Ronal B Morse USN (ret)
To: Robert Bruce Thompson


Well, it's happening again.  Xandros can't find the primary drive on IDE0 (otherwise known as HDA) approximately 3 out of every four boots.  On those occasions when it does find the drive it acts like it was never missing.  Except for that Xandros BE is running like a charm.

IDE0/1 aka HDA is a two year-old Seagate PATA120.  It hosts XP on NTFS and a FAT32 partition in addition to LILO in the boot sector.

I don't think it's a hardware problem.  I removed LILO and redid the boot sector and the machine booted 10 consecutive times into XP (both warm and cold boots) as expected without error.  CHKDSK is clear.

The drive in question is a two-year old Seagate PATA 120. It passes the IBM IDE diagnostic and there are No SMART warnings.  POST always detects and properly configures the drive during IPL.  The Knoppix live-CD detects the drive every time. 

Driving me a bit nutz...I've even changed the data cable and the CMOS battery (you do things you know can't possible help when you're pushed to the edge) in addition to verifying the SMP protocol setting that appeared to fix the issue before. 

I've sent a note to Xandros support, but will keep you posted...I have a feeling it's a kernel issue.

Ron Morse

I have a feeling you're right. I'm not at all convinced it was a good idea to go to the 2.6.11 kernel. I'm not sure what that fixed compared to 2.6.9, but I think it broke some stuff. In particular, the SP2/kernel upgrade for X3 Standard/Deluxe/OCE has apparently caused a lot of problems. The Xandros forums are full of posts about it.

I'm running X3 Deluxe with the SP2/kernel upgrade on my main system, and I'm noticing some odd behavior. For example, Xandros File Manager sometimes locks up when I do my regular backup of my working data to another network volume. Also, K3b has turned a bit hinky, despite the fact that I'm running the .24 release.

Barbara is running X3BE on her main system, and hasn't reported any problems, but she doesn't hammer on things like we do. I've hesitated to upgrade my main system to X3BE because I don't want to lose Mozilla Suite just yet, and it's no longer available from XN. I could install it from other sources, but I really don't want to do that.

As far as your own problem, have you used hdparm to check how the interface and drive are configured?

15:04 - I see that SCO has released OpenServer 6, the latest version of their flagship software. (I almost typed "lastest version", which might turn out to be true.) Their announcement was greeted by the collective yawns of the trade press and industry analysts. No one cares about SCO any more, except that a lot of people are looking forward to watching them crash and burn.

One has to wonder whether anyone will buy a copy. After all, this is a company known for suing its own customers. I have to think that SCO will sell new licenses only to committed SCO shops, and probably not many of those at that. They might also sell some upgrade licenses to current customers who desperately need whatever enhancements OS6 provides, although I suspect most of those users are migrating to Linux just as fast as they can.

SCO has burned all its bridges. There's no going back, and no going forward. They're toast.


Thursday, 23 June 2005
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

08:52 - I'm cranking away on the new book. I'm using StarOffice rather than MS Office, of course. SO has its peculiarities, but at least it doesn't munge long, complex documents, which MS Office does with some regularity. Unfortunately, O'Reilly still doesn't have a template for SO, which means I have to do a lot of manual formatting. O'Reilly does have a template for MS Word, but that won't work with SO because it uses MS-specific macros and so on. Oh, well. On balance, I'm happier using SO and knowing it won't eat my documents.

O'Reilly has this to say:

O'Reilly books cover the cutting edge of software and technology, and have always championed the Open Source movement. So it is no small irony that the majority of O'Reilly manuscripts are authored using Microsoft Word. However, given its ubiquity, in conjunction with it's status as de facto word processor for most authors—inside or outside the tech publication world—we are faced with a choice: convince authors to use other tools, some of which may be highly complex, poorly documented or technically unproven, or allow (even encourage) authors to use Word1, a tool with which they're already familiar, in the interest of keeping them focused on writing, not learning a new word processor. We've chosen the latter. Of course, we couldn't resist tinkering a bit.

Structure vs. Formatting

Regardless of the tool being used to author a manuscript, the actual task of turning the content into a printed book falls on the publisher, transferring the semantic structure of the manuscript into the format and standards of the particular line of books.

In an ideal world, you, the author, would simply apply semantic markup to your manuscript, and we could wave a magic wand and convert your marked-up text into a handsomely bound edition with a funny looking animal on the cover. And, to a certain degree, we can (sort of) do that. But, there's a problem. In order for the Magic Wand Approach to work, the semantic markup must be comprehensive and unambiguous.

The quintessential tool for providing comprehensive and unambiguous markup is, of course, XML (and in the case of technical books, DocBook XML, to be specific). There's one big obstacle to widespread adoption of XML as an authoring tool—it's too comprehensive and unambiguous. Well, for human beings at least. The process of applying XML markup to text, and then clearing up all the ambiguity (by validating the XML), is just too much for all but the bravest soul.

On the other end of the spectrum, there are authoring tools, like Word, that favor formatting over structure. If you want something emphasized, you highlight it, and press the "Italic" button (or the "Bold" button for that matter); if you need a bulleted list, press the "Bulleted List" button, and voila! This is how 99%2 of Word users format their documents.

But how do you distinguish a bulleted list that's in a sidebar, from a plain ol' bulleted list? That handy button isn't very handy for that. Of course, Word does provide some very nice support for semantic structure (in the form of Styles). It just favors formatting over structure. It's pretty easy to highlight a line of a paragraph, select a big point size from the Formatting Toolbar, than click on the "Bold" button to make a heading. It's much more sensible, however, to modify the "Heading 1" Paragraph Style to suit your needs, then use that to format all of your headings. That way, you get formatting and structure, with no difference in the final look of the document.

1 A tool that is highly complex, poorly documented and proven to be technically unstable.

2 According to the highly scientific research method known as "guesstimation."

Which really is ironic, given that SO enforces styles very strictly compared to MS Word, so much so that many reviewers have panned OOo/SO for its rigorous use of styles. Word formatting versus OOo/SO styles reminds me of the old days when people were used to typewriters and just getting started with word processing. A lot of them used the space bar to make things line up instead of setting tabs in the document. If Word formatting is the equivalent of the space bar, SO styles are the equivalent of setting tab stops.

So, the irony is that O'Reilly isn't encouraging the use of a perfectly good FOSS, style-oriented, cross-platform word processor that looks and works pretty much like Word, but is much superior for technical writing. I wish they'd do something about that. Perhaps when OOo 2.0 ships they'll reconsider.

09:28 - O'Reilly is nothing if not responsive. Fourteen minutes after I posted the comments above, I got this:

-------- Original Message --------
Subject: OpenOffice Template
Date: Thu, 23 Jun 2005 09:06:33 -0400
From: Brian Jepson
To: Robert Thompson

Here it is--they've asked me to not call this a template until it's got all the same macro goodness in it as the Word template. But it should be ideal for what we are doing--much better than using the Word template. Here's what Keith Fahlgren had to say:

> Internally, the word "template" carries a lot of baggage. We do not yet
> have an OOo "template." However, if your author is going to write in
> OOo and understands how to use templates in it, they should use the
> attached document as a template. Using the attached is much better than
> any other alternative and allows the author to use the Stylist on
> "Custom Styles" to use/view only the O'Reilly-approved styles (all of
> the ones currently in the Word template). Please let me know if the
> author would like other tips on writing efficiently in OOo (or has tips
> to share with me).
> If they're no already familiar with how to use templates in OOo,  they
> provide some decent documentation here:
> http://documentation.openoffice.org/HOW_TO/word_processing/HowTo_Work_with_Tem
> plates.pdf




Friday, 24 June 2005
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

08:45 - Periodically, I check out the statistics on my web server, including the most popular searches. I did that yesterday, and sent the following message to my friend Mary Chervenak:

-------- Original Message --------
Subject: You're a popular search topic on my site
Date: Thu, 23 Jun 2005 17:14:48 -0400
From: Robert Bruce Thompson
To: Mary Chervenak
CC: Paul Jones,  Barbara Fritchman Thompson

Month-to-date, searches for "Mary Chervenak" are in a tie for 19th place, just behind "armed squirrel" and just ahead of "sock drawer".

Very strange. I'm not sure why the run on your name and the armed squirrel thing. I know the sock drawer hits come from this page, where I wrote:

"Sock and underwear organization, indeed. Life is too short to organize socks and underwear. For that reason, all of my socks are identical, and I don't worry too much about organizing the drawer they live in. When I get dressed, I just grab one white one and two gray ones. If it's dark, I just grab one big one and two small ones. No need to worry about matching socks. The gray socks go with all the shoes I normally wear, which is to say tennis shoes and boots.

This is an actual unretouched photograph of my sock and underwear drawer soon after I finished doing the laundry yesterday. Note the clever organization. I keep a package or two of new underwear at the upper right, where they can easily be swapped in for failing units. I actually do have a few pairs of black dress socks, one of which is visible at the lower left. But the important thing is that takes me literally only seconds to reinstall my underwear and socks in the drawer after I bring up the clean laundry.

Barbara asked me to point out that this disorganization scheme is completely my idea, and she had nothing to do with it. But it doesn't bother her, either. It would have bothered a girl I used to date not long after college. I walked in one day to find her ironing my underpants. With spray starch, yet. There must have been an appropriate word, but my vocabulary failed me for once. "Are you deranged?", I asked her. That relationship didn't last long. Starched underpants. Geez."

To which I received the following replies:

-------- Original Message --------
Subject: Re: You're a popular search topic on my site
Date: Thu, 23 Jun 2005 18:20:23 -0400
From: Paul Jones
To: Robert Bruce Thompson
CC: Mary Chervenak, Barbara Fritchman Thompson

I think it is pretty clear that the standing on computer photo has boosted sales.  I'm sure if you paid Mary a 25" Obsession she'd call it even.

-------- Original Message --------
Subject: Re: You're a popular search topic on my site
Date: Fri, 24 Jun 2005 07:34:09 -0400
From: Mary Chervenak
To: Paul Jones, Robert Bruce Thompson
CC: Barbara Fritchman Thompson

I'm not sure whether to be flattered or annoyed.  I really hate squirrels.

On another subject entirely, I have an obsessively tidy sock drawer. Actually, I have two sock drawers -- one for white and light-colored socks and the other for black and navy blue socks.  All my socks are matched and folded together into pairs and the toes all point the same direction in the drawer.  I can send a photo, if you wish.

I also have two other drawers reserved for tights (one for light and one for dark), which, of course, are totally different from socks and therefore merit their own drawer(s).

Hmmm.  I clearly like socks much better than squirrels.

So that's it, then. We're waiting on the promised photos of Mary's sock and underwear drawers. As to the 25" Obsession, I'm afraid Mary's Obsession with her sock drawer will have to do. Sorry, Paul. 

14:00 - I'm working right now on the "Working on PCs" chapter for the new book. It's based on the chapter from PC Hardware in a Nutshell, but simplified and cut down, oriented more towards regular people who want to upgrade or repair a PC than to PC technicians. Less text, more pictures.

So, my question is this: I'm writing the section on software utilities for diagnosis, etc. So far, I have Knoppix, and I'll probably write briefly about things like the migration utilities bundled with hard drives. What other utilities should I mention? Which are your favorites, at least the ones you consider indispensable? Post your comments on the Daynotes forums. Thanks.


Saturday, 25 June 2005
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

00:00 -


Sunday, 26 June 2005
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

00:00 -


Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005 by Robert Bruce Thompson. All Rights Reserved.