TTG Home » Robert Home » Daynotes Journal Home » Journal for Week of 2 April 2001

photo-rbt.jpg (2942 bytes)Daynotes Journal

Week of 2 April 2001

Latest Update: Friday, 05 July 2002 09:16
 

Search Site [tips]


Order PC Hardware in a Nutshell from Fatbrain.com

Monday, 2 April 2001

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[TTG Messageboard]  [HardwareGuys.com Messageboard]

Thanks to everyone who has subscribed to the site. If you've already sent me email to subscribe, you made it in under the "early bird" deadline, even though some details may still need to be worked out. If you haven't subscribed yet, please consider doing so. We even accept Visa and MasterCharge now, via PayPal's secure server. Instructions for subscribing are on the subscription page.

I ran web stats this morning, as usual, and my traffic is up a bit over the last couple weeks. This site got something like 13,000 page reads last week, which is better than it has been recently, although still well down from the former traffic levels, which were typically in the 17,000 to 18,000 range. What's also interesting is that the "distinct hosts served" line item is down from its former typical 4,500 to 5,000 per week into the 2,800 range. Much of that loss, I suspect, is due to the declining number of "casual" browsers I'm getting. The back issues of my journal were getting a surprising number of hits from search engines, but now those back issues are in the password-protected subscribers-only area, and thus no longer available to the world at large.

I've commented on this before, but the pervasiveness of the Internet is one more nail in the coffin of languages other than English. I keep seeing articles that state the dominance of English on the web is waning. By some accounts, English language web pages now make up "only" 70% of the total. That may be true, but it misses the point. The web is global in its reach, and anyone who wants other than a provincial web site has little choice but to use English, whether as the only language or at least as an optional language. English has become the new lingua franca, and web sites that are available only in a language other than English just don't cut it any more for anything other than purely local purposes. Just as at one time all diplomats spoke French, today all web sites must speak English.

But there's much more to it than just the web, and it's in these non-web Internet areas that the dominance of English is beyond question. I thought about this the other night while I was reading one of the many amateur astronomy mailing lists I subscribe to. All of them, without exception, are in English. Why? There are certainly many amateur astronomers whose native tongues are other than English, and who do not speak English as a second language. They're left out in the cold, because it's impossible to achieve a critical mass on a specialized topic if the language is other than English. And the more specialized the topic, the more certain it is that only English language resources will be available. All of these amateur astronomy mailing lists have members from all over the world. All of them read and post to the list in English, because they have no choice.

Just from curiosity, I went looking for specialized mailing lists in other languages, including French and Spanish, which I can usually kind of puzzle out the meaning from, and German, which I used to more-or-less speak. There are plenty of mailing lists in languages other than English, but they're all on topics of broad interest. If you want to find a mailing list on rock music, for example, you can find one in just about any language you choose. But on more specialized topics, which is where the value of mailing lists really lies, it's English or nothing. I saw the same thing over and over. An English language mailing list on a particular narrow topic might have several hundred or several thousand members. If I could find a French or German mailing list on the same topic, it usually had a dozen members.

Willy Sutton once famously didn't say that he robbed banks because that was where the money was, and the same holds true here. If you're a specialist in some arcane (or not so arcane) topic, you'd better learn to read and write passable English. Otherwise, you're cut off from resources that are available to everyone who's taken the trouble to learn English.

The world has always been divided one way or another into "haves" and "have-nots". I wonder how long it will be before it becomes obvious that the ability to speak, read, and write fluent English is one of the new measures of have-hood.

Here's a factoid from the Thompson Testing Laboratories®. I'd meant to mention this at the time I did it, but I kept forgetting. Several months ago I was cleaning out old stuff. Among the old stuff were a bunch of antique IDE hard disk drives. And when I say old, I mean old. Stuff like 80 MB Maxtors, a 120 MB Seagate, a 200 MB Western Digital, and so on. They all still worked, but they were all too old, too slow, and too small to bother with, and I was about to throw them out. Before I did that, though, I wanted to check just to make sure there was nothing on them I cared about. (I still have some very old documents, for example, that I know I saved, but I can't figure out where). So, at any rate, I decided since I had a naked test bed running on my credenza I'd connect each of them up in turn and see what was on them. That varied from nothing to an incredibly old BSD installation to Windows 3.11 for Workgroups to just old and useless data.

After I'd finished checking them out, I was about to pitch them in the trash when a cunning plan came to mind. I wondered what would happen if I smacked each of them with a mallet. So, I took them down to the workbench, laid each one flat on the bench, and smacked it hard with a soft-faced mallet. After that I carefully carried them back upstairs (so as not to damage them) and tried reconnecting each to the test-bed system. The results, as best I can recall, were as follows:

  • Maxtor - I had a couple of these drives, 80 and 200 MB as I recall. Both survived and started up normally.
  • Seagate - I had several of these, ranging from about 80 to 250 MB. All survived and appeared normal.
  • Western Digital - I had three of these, I think. I could almost swear that one was only 40 MB, but I'm not sure they ever made an IDE that small. I think the biggest was 400 MB or thereabouts. One of the WD's survived the beating and seemed to start and run normally. One of the others was just completely dead. The last one tried valiantly to start, but was making nasty grinding sounds.
  • Quantum - I think I only had one of these, and I think it was 500 MB or thereabouts. It died horribly with a loud clatter.
  • IBM - again, only one of these, 200 MB as I recall. It seemed to survive and run normally.

So, there you have it. An exhaustive and scientifically rigorous test of the robustness of various kinds of hard drives. And, yes, it really is fun to take a mallet to a hard drive. Brightened up my day, anyway. And I do still have a stack of old drives in case I ever want to repeat the experiment.

Oh, yeah, I forgot to mention. Don't try this experiment at home. Among the stack of drives I carried downstairs for the mallet treatment was a Maxtor. It wasn't until I had the mallet raised and about to descend that I realized I was about to smack the 20 GB main drive from the test-bed system.

Click here to read or post responses to this week's journal entries

Click here to read or post responses to the Linux Chronicles Forum

[Top]

Tuesday, 3 April 2001

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[TTG Messageboard]  [HardwareGuys.com Messageboard]

Several people commented that the expiration of the "early-bird" subscription offer took them by surprise. And I was late in getting the credit card mechanism implemented. So, by popular demand, I've decided to extend the "early-bird" offer through the end of this week, which is to say through 23:59:59 EDT on Sunday, 8 April. If you haven't subscribed yet, here is your opportunity. All subscriptions received through the end of this week qualify for the extended expiration of 30 September 2002. I realize that some people outside North America have problems transferring money here. If you're in that position, just send me email to let me know your intention to subscribe and I'll honor the "early bird" terms. We can worry about the details later. For instructions on subscribing, click here.

I hit a milestone yesterday. For the first time, the amount of hard disk storage in my house exceeded one terabyte. I remember when I added the second floppy drive to my PC and was thrilled to have almost an entire megabyte of on-line storage. Now I have considerably more than a million megabytes. I wonder if mine is the first residence ever to exceed a thousand gigabytes. Probably not, but I suspect there aren't many. I don't think even Pournelle's Chaos Manor is over a terabyte yet.

Which reminds me that I really do need a name for this place. Jerry Pournelle has Chaos Manor. Chris Ward-Johnson (AKA Dr. Keyboard) has Chateau Keyboard. I need a name for this place so that I can stop saying "this place" or "my house". If you have any suggestions, please let me know. (All suggestions become the property of Robert Bruce Thompson; your mileage may vary; do not spindle, fold, or mutilate; do not operate heavy machinery while using this product).

UPS showed up yesterday with a pre-pub copy of fellow Daynoter Bo Leuf's new book, The Wiki Way: Quick Collaboration on the Web. I made it through the first hundred pages or so last night. Very nice. Well written, as you'd expect from Bo, and a fascinating topic. If you have any interest in setting up an interactive, collaborative web site, do yourself a favor and grab a copy of this book once it starts shipping. If you don't know what a Wiki is, visit Bo's web site. He runs it using Wiki.

Barbara and I spent yesterday afternoon tearing down an old Dell Pentium/200 system. Upgrading that system will form the basis of a new chapter in the next edition of PC Hardware in a Nutshell, currently titled This Olde Computer. Our old Dell will become famous.

Click here to read or post responses to this week's journal entries

Click here to read or post responses to the Linux Chronicles Forum

[Top]

Wednesday, 4 April 2001

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[TTG Messageboard]  [HardwareGuys.com Messageboard]

Thanks to everyone who's sent suggestions, publicly or privately, for a name for this place. I still haven't come up with one that really grabs me, but I'm thinking about it.

Barbara and I worked yesterday on tearing down the project upgrade system for the new edition. People sometimes ask me how long it takes me to build a new system. The answer is that it depends. If I'm just building a personal system from a stack of new components, it might take half an hour to an hour. If I'm building a project system for the book, it might take several days. That's because each step has to be documented, both in words and pictures. And because Malcolm always insists on helping.

HardwareDog.com

malcolm-computer-1.jpg (230163 bytes)

Barbara was about to install the IDE cable backwards,
until Malcolm pointed out her error.

malcolm-computer-2.jpg (367844 bytes)

Barbara and Malcolm take a break
during the system rebuild.

malcolm-computer-3.jpg (603245 bytes)

Malcolm, with both his stop and caution signals lit.

One of my readers reports that two significant discoveries have been made recently in the area of human pheromones. First, it's apparently now established that men can detect a pheromone released by women who are ovulating. Perfume companies are burning the midnight oil trying to replicate this pheromone. Second, it's now apparently established that women can detect a pheromone released by men who've bought something or done something without permission. Researchers have deemed this one the "guilt pheromone." So, if you've wondered how your wife and/or girlfriend always seems to know when you've bought something or done something you didn't want her to know about, that's how.

The David and Goliath battle between tiny Kyro and behemoth nVIDIA for the low-end gaming video adapter market continues. nVIDIA released a document, the obvious intent of which is to discourage manufacturers and OEMs from using Kyro products. There's a link to the original document in PDF format and a point-by-point refutation available here, which makes for interesting reading. Could nVIDIA be running scared? Surely not.

Click here to read or post responses to this week's journal entries

Click here to read or post responses to the Linux Chronicles Forum

[Top]

Thursday, 5 April 2001

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[TTG Messageboard]  [HardwareGuys.com Messageboard]

Earlier this week, Intel joined the distributed computing revolution by kicking off the Intel Philanthropic Peer-to-Peer Program. The first project will use distributed computing to process data in search of a cure for leukemia. Obviously, a worth goal in itself. But I wonder about the mechanics of it all, and where the money trail leads. I've read through all the material on both the Intel and UD sites, and I still don't have an answer to the cui bono? question. There is one comment on that issue as follows:

Who will own the patent for this discovery and which commercial enterprise will profit? 

All results will be released to the public within a year. The organizations receiving this data - the University of Oxford* and the National Foundation for Cancer Research*, are both non-profit organizations. Although they will own the Intellectual Property on any discovery they find, all information will be available for free.

but that strikes me as possibly weasel-worded. Why a year? What significance does the ownership of the IP have? Will drugs developed as a result of our cumulative contributed effort be generic from the beginning, or will some drug company patent them and generate huge revenue from our contributed effort?

We're all being asked to donate millions of hours of computer time, worth millions of dollars. In exchange, UD offers incentives like Sony Playstations, etc. I fear that the expected value of winning one of these prizes--probably a few cents per participant at most--is intended in a legal sense to recompense us for our effort and computer time. If that's the case, I have two questions:

1. Is UD reselling this computer time at a profit to the scientists who are doing the actual work, and, if so, at what rate are they reselling it? If they are reselling it for a sum substantially greater than the minimal amount that they are paying for that time, I want to be paid something more in line with the actual value of the computer time that I'm being asked to donate. That might be something on the close order of $1,000 per CPU-year. Otherwise, we're all simply giving UD a valuable service that they're reselling to the scientists at a profit.

2. Assuming that all this effort and donated computer time results in a pill that cures leukemia, who has the rights to that drug?

If UD is running a commercial operation, I want to be paid the full value of any CPU ticks I contribute, which is to say something around $1,000 per year per machine. Conversely, if UD is doing this on a non-profit or nearly non-profit basis, I'd be happy to donate spare CPU ticks to a good cause. If, that is, Oxford and the NFCR intend to make the results of all this contributed horsepower freely available, i.e. any patent they hold on the results is freely licensed to anyone who wishes to make the new drug. That is, the new drug must be made available as a generic product initially, and no one, whether a drug company or other organization, is making profits as a result of the contributions of me and millions of other good-hearted people.

I hope my concerns are groundless, but I fear that UD will be making large amounts of money by reselling the computer time we all donate. The old lawyer's watchword is cui bono. Before I undertake this myself, and certainly before I recommend it to my readers, I want to know the answer to that question. Who benefits?

But at least the client downloads quickly...

There are times when I really like my Roadrunner cable modem service.

Several people have asked me, privately or on the messageboards, what the differences are between the original D815EEA ("Easton") and the D815EEA2 ("Easton 2") motherboards. Herewith the official answer from one of my contacts at Intel about the new features of the D815EEA2:

  1. Supports 1.13GHz P3 processor
  2. Lower price (I don't have figures at the moment)
  3. Intel Active Monitor (hardware management ASIC) on all skus
  4. Integrated AC'97 audio with SoundMAX 2.0 on all skus
  5. Four back panel USB ports (not all skus have front panel USB)
  6. ATX board is 1/2" shorter (8.2" x 11.5") than D815EEA board (8.2" x 12")

As far as I'm concerned, this is a minor rev, and not worth worrying about. I suspect that eventually the EEA2 will replace the EEA in the channel, but for now at least the two seem to be co-existing. I wouldn't have any hesitation about getting an original EEA, and recommend that you not hesitate either, particularly if your chosen vendor has discounted the EEA model to below the price of the EEA2. My comments on the above follow:

  1. Who cares? The Pentium III/1.0G is readily available. The 1.13 GHz model is impossible to find, and provides little additional speed anyway.
  2. That may eventually be true. Right now, EEA boards seem to be selling for less than EEA2 boards, but that's probably an inventory clearance issue.
  3. The hardware ASIC is a nice thing to have, but I don't regard having it or not as a major issue.
  4. The original EEA came in two versions, with Creative ES1373 sound or Analog Devices AD1885 sound. Now that the SoundMax 2.0 drivers are available, I think the AD1885 is a better choice, and that's what all EEA2 boards ship with. But it's not a big deal either way.
  5. If you plan to use USB, having four root ports on the back is nice. But I normally connect no more than two USB devices, one for each of the two ports on the back of the existing board. (Some EEA models have header pins for a second set of two on the front panel). Again, not a big deal either way.
  6. Well, I suppose the half-inch shorter EEA2 might be useful in some cases, but I've never found a case that the original EEA didn't fit.

Click here to read or post responses to this week's journal entries

Click here to read or post responses to the Linux Chronicles Forum

[Top]

Friday, 6 April 2001

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[TTG Messageboard]  [HardwareGuys.com Messageboard]

It appears that my concerns about the Intel Philanthropic Peer-to-Peer Program may be justified, or at least that there is no evidence to suggest that they are not. The Register posted this article this morning, which reinforces my fears. As I mentioned yesterday, The University of Oxford retains the IP rights to whatever is discovered as a result of our collective donation of computer time. What is unclear is how they will exercise those rights. 

On the one hand, they could license that IP freely (as in beer). But I'm afraid that's not what's going to happen. What I fear will happen is that The University of Oxford will license those rights for a large sum to a major pharmaceutical company, perhaps via auction. That pharmaceutical company will then patent and produce the new wonder drug, and sell it at a very high price to recoup their payments to The University of Oxford.

I regard that as completely unacceptable. As far as I'm concerned, if these parties solicit volunteers to donate computer time, the results belong to everyone equally, not to Oxford University, and certainly not to some pharmaceutical company. Everything produced as a result of volunteer effort should be GPL'd, in other words. Although no one has said it in so many words, the implication is that these scientists are sitting on an almost certain cure for leukemia, and all they need is computer time to produce that cure. I don't think that's the situation. If it were, The University of Oxford wouldn't be soliciting donated computer time. They'd be buying computer time, and any number of pharmaceutical companies would be standing in line waving money at them to get in on the ground floor.

I think what's going on is that there's a small but realistic chance that there's a cure buried in all this data. Call it a 0.01% chance. The University of Oxford thinks that it's worth going after that 0.01% chance, and they may be right. From an economic perspective, I'd guess that the cost of the computer time required is too high to make it worth risking spending millions on commercial supercomputer time against a 0.01% chance of a payout, albeit a huge one. In other words, the expected value of the solution is less than the cost of finding that solution. 

But that equation changes dramatically if the computer time is all donated. In that case, the 99.99% probability that all the effort invested will in retrospect be wasted has next to no economic cost, so there is no downside to pursuing the solution. And the 0.01% probability of a successful solution still has the huge payout. In effect, The University of Oxford are gambling with our money. If they lose, we pay. If we win, they collect. Or so I speculate. I hope I'm wrong, but I fear I'm right.

Click here to read or post responses to this week's journal entries

Click here to read or post responses to the Linux Chronicles Forum

[Top]

Saturday, 7 April 2001

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[TTG Messageboard]  [HardwareGuys.com Messageboard]

Well, Pacific Gas & Electric, California's largest utility, is now bankrupt. What really annoys me is the slant the news media put on this failure. They all blame it on "California's disastrous experiment with deregulation". Excuse me? The failure of PG&E was caused by the California government capping the price PG&E could charge consumers for their products while placing no limit on how much producers could charge PG&E for the energy they are forced to buy now that PG&E are no longer energy producers but only energy distributors. So, let's see. If California forces PG&E to sell energy for $1 that they must pay $2 to acquire, where is the mystery? Make no mistake. This is not a failure of deregulation. It is a failure of regulation. 

The Indiana legislature supposedly once enacted a law that declared the value of pi exactly equal to three. I don't know if that's true or not, but if it is it doesn't surprise me. Lawmakers have little grasp on reality and a greatly exaggerated opinion on their own powers. Wishing something so doesn't make it true, and passing laws that fly in the face of reality is no more effective. California legislators and regulators have destroyed PG&E by their ridiculous actions, and attributing the predictable results of those actions to a failure of deregulation is simply contemptible.

California now faces a long, hot summer, all thanks to their lawmakers and regulators. If Californians want to do something about the problem, I'd suggest tarring and feathering their legislators and regulators and riding them out of town on rails.

By definition, all spammers are stupid, but I got a spam yesterday from one that makes other spammers look like rocket scientists:

From: lawn@netrover.com [mailto:lawn@netrover.com]
Sent: Friday, April 06, 2001 4:33 PM
To: thompson@ttgnet.com
Subject: Spring Lawn Care

Introducing Paul's Weed & Feed
Your Lawn Care Specialists since 1975!
Servicing the Newmarket and surrounding area.
Spring Special
Lawn care programs from $120.00

contact us at lawn@netrover for your FREE ESTIMATE!!!
905-836-6221

________________________________________________________
If you wish to be removed from this advertiser's future mailings, please reply
with the subject "Remove" and this software will automatically block you
from their future mailings.

I looked up the area code, and found that this business is located in Ontario, Canada. Since he specifically asks people to contact him, and since he doesn't explicitly forbid calling collect, I suggest that everyone who reads his message place a person-to-person collect call to Paul and ask him for a quote. Let's help spammers prove what they've claimed all along--that spam is an effective way of developing new leads. In particular, I'd encourage my readers from outside North America to call this guy collect and get him to provide a free estimate on doing their lawns. Let's all do our part to make sure that this guy gets a good response to his spam.

Spring has really arrived in Winston-Salem. Outside temperatures yesterday reached nearly 80F (26C), and Barbara tells me today's forecast is for temperatures in the mid 80s (30C). The inside temperature yesterday got up above 76F (25C), and we considered turning on the air conditioning. Instead, we just turned on the fan to circulate the air, which made things comfortable last night. 

But with sunny skies and the expected highs today, we went ahead and turned on the air conditioning this morning. I suspect we'll need it. Particularly in my office, which is often a good 15+F (10C) degrees warmer than the rest of the house because of all the computer equipment running there. As I turned on our air conditioning, I thought about those poor folks in California, many of whom won't be allowed to use their air conditioning this summer. With all those computers that Jerry Pournelle keeps running, it's likely to get a bit warm at Chaos Manor.

I have a lot to do, so I'd better get to work. Things might be a little sparse around here for the next couple of weeks. I have chapters I want to get finished and sent off to my editor, and I also have to do my state and federal income tax returns. Ugh.

Click here to read or post responses to this week's journal entries

Click here to read or post responses to the Linux Chronicles Forum

[Top]

Sunday, 8 April 2001

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[TTG Messageboard]  [HardwareGuys.com Messageboard]

Late start this morning. Pournelle sent me a preliminary draft of his new column overnight, so I spent half an hour or so reading and commenting on that before I got started on anything else. Then there were dogs to take out for their morning constitutional, laundry to get started, email to read and respond to, web sites to visit, etc., so I didn't even get started on writing my journal entry until well after 9:30, which is when I usually publish it.

Well, late start in that respect, but early start otherwise. Malcolm has taken up waking us shortly after dawn. Duncan used to do that, too, and it made me want to kill him. Fortunately, I was able to restrain myself until he'd outgrown his self-assigned alarm clock duties. But at least Duncan used to wake me up just by snouting me. Malcolm has a different method. He comes over and slurps my face. Note to self: do not sleep with mouth open.

Speaking of Malcolm, there's a continuing struggle between Malcolm and us humans. He thieves, we try to stop him. But he's fast, crafty, and always remembers where the good stuff is. He's so fast and crafty that if I open a closet door to get something he can sneak in behind me without me seeing him, steal something, and then sneak back out as I close the door. I keep wondering if he could be trained to do that with a bank vault. Malcolm particularly likes used socks and underclothes, hats, gloves, and towels, but his absolute favorite is tissue paper of any sort--toilet paper, Kleenex, paper towels, whatever. A toilet paper roll after an encounter with Malcolm is not a pretty sight. I wouldn't mind so much if he'd just eat the toilet paper, but instead he shreds it. Most people are not aware that, when shredded, a single roll of toilet paper can cover the floor and all furniture in an average size room

So Barbara decided to up the ante. Mouse traps have always been used for training Border Collies not to thieve. When they disturb one and get snapped on the snout, it stings enough to make sure they'll avoid them in future. But the problem with mouse traps is that one needs an awful lot of them to protect against a larcenous Border Collie. They're very good at avoiding the traps and still getting what they want. What Barbara bought is a modified mouse trap. It's a standard trap, but with a large plastic paddle that snaps onto the main snapper part of the trap. When the trap goes off, the plastic paddle slows the snapper enough to prevent the dog from getting a painful smack on the snout. It also has enough air resistance that it sends the trap itself flying, And, of course, the plastic paddle itself becomes part of the trigger mechanism, making it much more likely that the dog will set off the trap.

So far, so good. We demonstrated the trap to Malcolm, and it terrifies him. Unfortunately, it doesn't terrify him enough to overcome his thieving tendencies. Barbara keeps her office door blocked with a baby gate, and moved the laundry basket for used towels in there some months ago to protect it from Malcolm. So, as a test, she set up one of her new snappers on top of the used towels and left the baby gate down. Minutes later, I saw Malcolm walking down the hall with a used towel in his mouth. I went back to check and found the trap still set and undisturbed. I don't know how he does it, but he does.

So much for my idea of sleeping with one of the snappy traps strapped to my forehead.

Click here to read or post responses to this week's journal entries

Click here to read or post responses to the Linux Chronicles Forum

[Top]

 

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