TTG Home Robert Home Daynotes Journal Home Journal for Week of 3 September 2001

photo-rbt.jpg (2942 bytes)Daynotes Journal

Week of 3 September 2001

Latest Update: Friday, 05 July 2002 09:16
 

Search Site [tips]


Click Here to Subscribe

Visit Barbara's Journal Page

Monday, 3 September 2001

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

10:40 - Labor Day. I was going to highlight that in bolded red letters, but for us it's just another day. The self-employed don't take holidays, or if we do it's on our own schedule. I do plan to take a break from PCs today, though. I think I'll work on our poisons e-book.

It will be short shrift around here all week. I have several chapters nearing completion, and I'd like to get at least one of them off to my editor. Preferably two. There's still a lot of work to do before the first draft of the next edition is complete, and then a lot more work to get from first draft to final book, but I can finally see the train at the end of the tunnel.

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, 4 September 2001

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

08:45 - Barbara and I took some time off yesterday. She watched some of the US Open, and I worked on the poisons e-book, glancing up from time to time to watch part of a match. I used to enjoy playing tennis, but I've never been one to watch sports on TV. Still, if something has to be on, I'd just as soon it be tennis or golf. 

The seedings this year seem deranged in both the men's and ladies' draws. A seeding that puts Sampras, Rafter, and Agassi in the same half--let alone the same quarter it did as this year--is ridiculous on the face of it. And how on earth do they justify top seedings for Martina Hingis, who hasn't done anything in the majors, and for Kuerten, who doesn't belong on the same court with Sampras, Rafter, or Agassi? 

Seedings used to be about who was expected to beat whom. No longer. Now they're based on computerized rankings, which appear to be calculated using a defective algorithm. Otherwise, how can Hingis be seeded first when Venus Williams, Davenport, and Capriati have dominated the Grand Slam events for last year and this year? Of course, to some extent, the seedings may be self-fulfilling. If you force the really good players to fight it out amongst themselves in the early rounds and leave your number one seeds a clear path to the finals, well, they're likely to get to the finals. And they may even win. But that doesn't mean they deserved the top seeding.

I see that HP is acquiring Compaq, although it's being represented as a merger. It seems that Carleton "Carly" Fiorina wasn't satisfied with butchering one large, respected computer company. Now, as CEO of the combined HP/Compaq, she'll have the opportunity to butcher another. This all reminds me of the merger years ago of Burroughs and Sperry to form Unisys Corporation. They had a stupid slogan, wherein they claimed that the joining of the two companies didn't double but "raised to the power of two". I remember emailing them at the time to explain that one doubled was twice as much whereas one "raised to the power of two" was still one. They never replied, although they did stop using their stupid slogan, so perhaps they did listen. 

I suspect that this merger of HP and Compaq will be another instance of taking what used to be two pretty good companies and raising them to the power of two. Actually, it's probably worse than that. If you assume that both Compaq and HP are each half the companies they used to be, this merger by raising them to the power of two will give us a quarter of what we used to have. That sounds about right.

Back to work on PC Hardware in a Nutshell today. I'll be working on the Memory chapter.

12:45 - Hmmm. I was doing one of my regular xcopy backups, which copies all changed files to other machines on the network with multiple redundancy, when I got an error message. It seems that orion, which is the Win98SE machine that hosts the scanner and inkjet printer, ran out of disk space. It only has a 10 GB hard drive, so keep copies of my archive directories there was bound to cause a problem sooner or later. So I moved that stuff off orion to ursa, which is my den system. That box has a 20 GB hard drive, of which 4 GB is the Windows NT4 boot/system volume and the remaining 16 GB was an empty D: volume. The stuff that was on orion is now moved over to ursa, and everything works again. I have several 100+ GB hard drives on the shelf, so perhaps I'll install one of them somewhere and devote it to archived data. More likely, I'll install two or three on different systems and use that redundancy as my primary backup method. I will pull a set of tapes of the archived stuff periodically, and probably a set to optical media at some point, but it's not a high priority.

While I was looking at orion, I noticed that one reason for the disk-full error was an 80 MB file living in the root directory. It's a log file generated by my scanner, of all things. All it contains is repeated examples of the same four log lines, telling me that a certain register on the LM8341 (or something like that) has been read. Those same four lines over and over again for 80 MB worth. Every few minutes, whether or not I'd used the scanner, that same four lines would be written again to the file. I'd imagine it'd zip down to about 200 bytes if I cared to try it. But I just deleted it instead. That was an hour or more ago, and I'd kind of expected it to reappear, but it hasn't so far.

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, 5 September 2001

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

08:50 - It took 4:07 to copy just over 10 GB of data from theodore and thoth to ursa, a rate of about 730KB/s or about 58% of theoretical 10BaseT throughput (ursa has a 10 Mb/s card). At best, 10BaseT cards in real-world applications generally provide 85% or so of theoretical throughput, but that is further reduced in this case by filesystem overhead. In other words, I was copying thousands of small files, and it takes Windows NT time to create each of those files on the destination drive. Even if I'd had a 100BaseT card in ursa, it would have probably taken about 2 hours to complete the copy. 100BaseT transfers data ten times as fast as 10BaseT, but there's still filesystem overhead to consider.

I keep reading about shark attacks. It seems to me that we need to eradicate the things, which are after all nothing more than sea rats. Unlike whacko environmentalists, who hate the species of which they are a member, I prefer to look at things from a species-centric viewpoint. What's good for homo sapiens sapiens is what counts. And sea rats are no more useful than cockroaches, mosquitoes, or rats. Granted, the things are just about brainless, but even they should be smart enough to realize after some experience that beaches where humans swim are a very bad place for them to hang out. They're easy enough to kill. I've done it myself with my .44 Magnum and a 30-06. One round right at the base of the dorsal fin has amusing results.

There is often a sighting before the attack, but the people in the water don't have time to reach the shore before the sea rat reaches them. So perhaps it's time to begin equipping lifeguards with scope-sighted rifles. Either that, or M79 grenade launchers, which would work even better. And on a larger scale, we should probably be dropping poisoned bait where sea rats tend to congregate. Perhaps using strychnine would be a good idea. I understand that a thrashing sea rat is often attacked and eaten by its own kind. We have to teach these things that people are not food.

Back to work on the chapter. The Weather Channel claims it'll be clear and cool tonight, so we take the telescopes up to Bullington.

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, 6 September 2001

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

10:05 - The Register posted some leaked information from Intel about their projected processor sales for the next couple of months, and it makes interesting reading. I'll repost the graphic here, since it's apparently not copyrighted and The Register site has been very hard to get to for the last couple of weeks. The fast rampdown of the Pentium III, from 7.5 million units in July to 2.5 million units in October, was expected, with the push that Intel has been giving the Pentium 4. The 33% increase in sales of the Coppermine128-core Celeron, from a flat 3 million/month in July and August, to 4 million units in October, probably represents replacement sales for the rapidly disappearing Pentium III. That Celeron line should flatten for a month or two and then begin dropping rapidly as Tualatin-core Celerons (the red line at the bottom) ramp up. 

Also interesting is the line for the obsolescent Socket 423 Pentium 4. Despite the rampup of the "real" Pentium 4 (the Socket 478 version), the Socket 423 Pentium 4 sales continue to increase, reaching nearly 2 million units in October. I'm not sure Intel is right about this one, unless they're also planning on producing 2 million Socket 423 Pentium 4 motherboards, because the third-party motherboard makers have essentially stopped making Socket 423 motherboards. They don't want to get stuck with a bunch of unsellable boards. My sources tell me that anyone looking for Socket 423 motherboards for the rest of this year is going to find the pickings slim. Similarly, the rampup in P4 sales has temporarily benefited Rambus, greatly increasing demand for RIMMs. That won't last long, though, because few people will be buying Rambus-based P4 motherboards once Intel's and other maker's Socket 478 SDRAM and DDR-SDRAM motherboards are shipping in volume.

The tiny sales for the 0.13 Tualatin-core Pentium IIIs are surprising, until you realize that Intel has basically killed Tualatin in the PIII segment by pricing it out of contention relative to the P4. Tualatin will live on, but only in mobile PIII variants and the "Celeron 4". The last thing Intel wants is a full blown Tualatin-based PIII with 256KB of L2 cache and running at ~1.5 GHz competing with the P4. The benchmarks of the Tualatin P3 versus the P4 might be embarrassing for the P4.

But what I find really striking is the total sales, 11 million total units in July versus more than 13 million in October. Granted, system sales usually increase in the fall over the dead summer period, but I was expecting them to remain flat until later in Q4 or perhaps Q1/2 of next year.

And here are the numbers upon which the graph is based:

  Jul Aug* Sep* Oct*
Pentium 4 (Socket 423) 440 940 1,510 1,920
Pentium 4 (Socket 478)  0 260 1,850 4,080
Pentium III (Tualatin) 110 220 320 400
Pentium III (Coppermine)  7,480 6,490 3,580 2,530
Celeron (Tualatin) 0 0 0 400
Celeron (Coppermine128) 2,970 3,080 3,560 3,990
Total 11,000 10,990 10,820 13,320
*projected sales.
All numbers in thousands

We did go up to Bullington last night. Bonnie Richardson showed up about 9:00 p.m., but she hadn't missed much. There weren't many clouds, but the haze was terrible. So bad that at times Mars completely disappeared. We never were able to see Sagittarius as a constellation, although we could spot a few of its stars from time to time. We were mostly limited to viewing more or less straight overhead. That's a problem with a Dobsonian-mounted telescope. The area within a few degrees of straight up is known as "Dobson's Hole" because it's nearly impossible to aim the telescope or, having somehow aimed it, to track the objects you have in view. 

Dobsonians are altazimuth mounts, which means you aim by moving the scope up and down (altitude) and right and left (azimuth). But when you're pointed straight up or nearly so, those directions have little meaning. At 70 or 75 degrees elevation, a Dobsonian starts getting difficult to maneuver. At 80 to 85 degrees, it's just about impossible. So we basically just let the scope sit unused, and the three of us sat around in our lawn chairs using binoculars to look at objects high overhead. We found some neat stuff, including the Coathanger Cluster, which looks just like its name, the Double Double, a bunch of other double stars, and so on. The moon rose over Pilot Mountain about 9:30, so that gave us something else to look at. Overall, it was a disappointing night as far as objects logged, but we still enjoy just being up there under the night sky.

Fortunately, we're headed into the good part of the year for observing. Summer in the Carolinas is miserable for observing. The Fall and Winter months will be a lot better. Clear skies, cool temperatures, and a lot less haze. 

I sometimes point out when I was right, so it's only fair that I point out when I was wrong. At some point, I declared IEEE1394 a moribund technology in the PC environment, saying that while it was all well and good in the consumer video segment, there was little point to it on PCs and that USB 2.0 would eat its lunch. Well, that was right as far as it went, but things have changed. There's an ongoing bloodbath in PC sales, and everyone is desperate to do something to jumpstart them. The current thinking is apparently that putting IEEE1394 ports on PCs will encourage millions of home-video enthusiasts to buy new PCs with 1394 ports. I don't see that happening, but whether it does or not is pretty much immaterial. USB 2.0 has stalled, and all the big guys are now pushing IEEE1394.

In one sense, that's good, because IEEE1394 actually has some advantages over the simple-minded USB. But in another sense, it's bad, because IEEE1394 is a more expensive, proprietary solution than the royalty-free USB.

Interesting news from reader Holden Aust:

-----Original Message-----
From: Holden Aust
Sent: Thursday, September 06, 2001 2:12 AM
To: Robert Bruce Thompson
Subject: Tyan Dual-Athlon AMD-Chipset $233

Hi Bob,

Did you know that Tyan has just come out with a new AMD-chipset-based Dual-Athlon (or Dual-Duron) motherboard which uses a standard ATX 300 watt power supply (it has to provide 30A at 5V) for $233. This is the Tiger S2460. Four 64-bit (!) PCI slots, 2 32-bit PCI slots, AGP slot. Crucial has 256 Meg DDR ECC Registered DIMMs for it for $43 each. One of these motherboards is on it's way to me now. This should be fun!

Here's some info:

Tyan's page:

http://www.tyan.com/products/html/tigermp.html

Price on motherboard from www.tccomputers.com

Let's see, two 750 MHz Durons for $36 each or two 950 MHz Athlons for $62 each or should I "splurge" and get two 1.4 G Athlons for $109 each? Choices, choices, choices...

Tom's Hardware says you don't have to use the MP Athlons, that regular Athlons or even Durons will work fine and that was confirmed recently by an AMD techie who told me that you could indeed run regular Duron or Athlons in those motherboards, although I don't think either Tyan or AMD officially support that. Then again, MP Athlons are only $161.

--- 
Holden

Which pretty much eliminates my objections to dual-Athlon systems. Tyan is not my favorite motherboard maker, but their boards are generally solid. My main qualm about Athlon/Duron systems has always been their dependence on VIA chipsets. AMD chipsets are solid, and I'd have no hesitation running an AMD760MP-based system as my main personal system. 

At this point, I think if I were building an SMP system, I'd base it on the components you mention rather than Intel processors. My one remaining concern is heat. AMD processors pull more current than Intel processors, and putting two AMD processors (especially fast Athlons) in a system means you need to pay particular attention to cooling. I'd be inclined to use at least two or three 120mm supplementary fans, or the equivalent in smaller fans. At least the Palomino-core Athlons have thermal diodes, which has always been a glaring deficiency in AMD processors until now. 

Also, I think I'd be inclined to use a better power supply than a generic ATX 300W unit. Even if it supplies the required 30A at 5V, it'll be running its components at or over capacity. I think I'd be inclined to stick in a high-end PC Power & Cooling unit, which delivers its rated capacity while running its components at 50% or 70% of their ratings instead of 90% or 100% (or even 100+%).

And that's enough for now. I need to get to work.

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, 7 September 2001

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

10:17 - My comments on sea rats seem to be drawing a lot of feedback via private mail, including many from people I've never heard from before. Just to clear up one issue, many people seem to assume that I'm advocating eradicating all sharks. I didn't say that, and that's not what I meant. There are many species of sharks, only a few of which, like the Great White and Hammerhead, are dangerous to people. Eradicating those species--or at least teaching them that places where humans congregate is not a good place for them to hang out--would be a Good Thing. I also think that eradicating cobras would be a Good Thing, but that doesn't mean I'm in favor of wiping out garter snakes or black snakes. In fact, I rather like them.

As to the people who've commented that sea rats attack only a few people, I find that a contemptible argument. One is too many. How many of their own family would they regard as acceptable losses to sea rats?

At least one ISP, Blueyonder, has taken the obvious step of cutting off service to customers who are infected with CodeRed. It's about time, I say. My cable modem is still flickering constantly as CodeRed-infected machines probe my gateway box. Anyone who was infected with CodeRed in the first place was irresponsible, given that a patch was available for a month before the worm appeared. Anyone who is still running a server infected with CodeRed goes so far beyond irresponsible that I don't know if there's a word to describe their behavior. It's easy enough for service providers to determine which of their customers are flooding the Internet with CodeRed probes, and under any likely terms of service agreements, those providers are entitled to cut off service to those customers. So why haven't they done so? Probably because they want their money.

Barbara named this place the Thompson Techno-Grotto, and that's pretty apt. I've cut way down on the number of computers in my office. I have only half a dozen or so running right now. But even so, with all the UPSs, KVM switches and other ancillary equipment, there are bunches of electronic components active in my office. One of them is emitting a high-pitched beep sporadically, and we can't figure out which one is doing it. It woke Barbara in the middle of the night. I was actually crawling around on the floor yesterday trying to figure out which component was beeping, but without success. Wherever I was, the beep seemed to be coming from somewhere else.

And these comments about dual Athlon systems from reader Steve R. Hastings:

-----Original Message-----
From: Steve R. Hastings
Sent: Friday, September 07, 2001 12:20 AM
To: thompson@ttgnet.com
Cc: jerryp@jerrypournelle.com
Subject: Hot chips: SMP with AMD and Intel

Dear Mr. Thompson,

In your journal you commented that an AMD-based SMP system would be hotter than an Intel-based SMP system. I got curious and looked up some numbers.

All numbers came from the PDF data sheets on amd.com and intel.com.  Intel only publishes one number for the chip, which I take to be the maximum; AMD publishes both "maximum" and "typical" numbers.

To build an SMP box with Intel chips, you would want the .13 micron process version of the Pentium III. Running at 1.2 GHz, this chip dissipates 30 Watts (maximum).

In comparison, the AMD Athlon MP running at 1.2 GHz will dissipate 49 Watts (typical) or 55 Watts (maximum).  But the Athlon is faster than the Pentium III.  You could use the 1.0 GHz version of the Athlon MP, which will dissipate 41 Watts (typical) or 46 Watts (maximum).

By the way, although it is not an officially supported configuration, you could run two 1 GHz Duron chips in an SMP Athlon motherboard.  But the thermal specs for the 1 GHz Duron are identical to the thermal specs of the 1 GHz Athlon MP, so it would make no sense to go with the non-supported and lower-performing Duron for an SMP system.

Anyway, let's now compare with the Intel Pentium 4.  (IIRC there is no SMP for the Pentium 4 at this time, so you can't do SMP with it anyway, but never mind.)  The latest version of the Pentium 4, running at 1.5 GHz, will dissipate 58 Watts (maximum).  At 2 GHz it will dissipate 75 Watts (maximum).

It appears to me that with respect to heat, the Athlon is actually better than the Pentium 4: for most applications the Athlon is faster than the Pentium 4 at equivalent clock rates, and a Pentium 4 that is about equal to an Athlon will be hotter than the Athlon.  Both Athlon nor Pentium 4 run rather hot.

Both the Athlon and the Pentium 4 will be getting a die shrink to .13 micron, which will make them both a lot cooler, but I imagine the Pentium 4 will still be hotter than an equal-performing Athlon.

P.S. What I would really love would be an inexpensive SMP motherboard with Transmeta Crusoe chips. At 800 MHz, a Crusoe TM5800 dissipates 6 Watts (maximum), with a claimed typical dissipation during DVD decoding and playback of only about 1 Watt! For email, Internet, desktop applications and the like, such a system would be more than powerful enough; and with that little heat dissipation it should be possible to make such a system very quiet, perhaps silent. I've seen sub-notebooks and rack-mount servers with the Crusoe, but no ATX or microATX motherboards (either SMP or single-proc).

--
Steve R. Hastings             "Vita est"
steve@hastings.org
          http://www.blarg.net/~steveha

Well, you make an unwarranted assumption about the relative performance of the Athlon and Pentium III. In fact, those two processors are about evenly matched. Clock-for-clock, the Athlon wins some benchmarks, the Pentium III others, the Athlon generally has slightly faster floating point performance, and the Pentium III slightly faster integer performance, but overall the two processors are about identical in performance. Of course, we're comparing the Athlon with DDR-SDRAM versus the Pentium III with SDR-SDRAM, so that gives a slight advantage to the Athlon. A very slight advantage, since neither processor benefits greatly from the additional bandwidth of DDR-SDRAM. The Pentium 4 is not a consideration, because as you've observed there aren't any dual Pentium 4 motherboards available.

So what we're comparing is dual Pentium IIIs against dual Athlons. Using your own figures, the dual Pentium III configuration dissipates a maximum of 60W, whereas the dual Athlon dissipates 110W. That is a huge difference. Just imagine the difference between a PC case with nothing in it versus that same case with a 50W incandescent light bulb running all the time. That's the difference in thermal load between those two configurations, and the delta remains similar whether the processors are running maxed out or loafing.

But the real point, of course, is that system cooling is critically important in any SMP box. Even the dual Pentium III system dissipates as much heat as a 60W light bulb, and that's non-trivial. If you build any kind of dual processor system, pay very close attention to system cooling.

10:42 - Road Runner has turned into Road Food this morning, so I have no idea when I'll be able to publish this. When our service died, I immediately called Time-Warner tech support, having learned my lesson previously. They've already cut off our service twice for non-payment, even though we've never been late paying our bill in the 18 years we've had cable service. 

Each time, it took only seconds to get service restored, once I got through to someone who knew what was going on. I thought this morning's problem was more of the same. When I called their automated service, it had me enter our phone number. I did that. It told me that no outages had been reported and suggested I press 9 to speak to a human. I did that, and she told me there was in fact a small outage--the state of North Carolina is down. She'd gotten a message just as I got through to her, telling her that the Raleigh operations center was down and working on the problem.

I was hoping that someone had tripped over a cable in their Raleigh data center, but the service has been down long enough now that it's likely there's a more severe problem than that.

Oh, well. I'll publish this when service comes back up.

11:14 - Road Runner is back up and the beeping mystery is solved.

When the beeping started again, Barbara ran into my office to try to locate it. I started shutting down machines, thinking that perhaps one was overheating and the BIOS was generating a warning beep. That didn't help, but we eventually localized the beep to, of all places, my storage closet. There, I found an old APC 1000VA Smart-UPS system, sitting there beeping away. It wasn't connected to anything, mind you, but it had decided to come to life on its own. I remember having some strange problems with it a few months ago. I have numerous UPSs, so the easy thing to do was disconnect it and put it in the closet.

I think what happened was the battery had self-discharged to the point that the UPS was crying for help. So I plugged it in to give it a charge, but it now sits there beeping even when it's connected to mains power. I think it's a defective unit, which is surprising. Of the hundreds of APC UPSs I've bought over the last 20 years, very few have failed. But I think this one is the exception.

Turning it off doesn't help. It comes to life on its own and starts beeping. Hmmm. I have two alternatives here. Disconnect the battery, or take it out in the back yard and shoot it.

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, 8 September 2001

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

9:22 - A year ago, I bought Barbara an AT&T Prepaid Wireless phone. We'd been with a cellular company for five years or more, and I got tired of paying $50 every month and using only 10 or 15 minutes of air time. What really set me off was that after us paying $50/month for five years and using a total of perhaps 15 hours of airtime over that period, they charged Barbara something like $10 for one two-minute phone call that was just outside our home service area. She didn't even realize at the time she was roaming, and so ended up being charged something like $1/minute for airtime plus an outrageous per minute long-distance charge, plus something like $6 for a daily roaming charge. At that point, I decided they could go to hell, and signed up for AT&T Prepaid Wireless.

That turned out not to be ideal. Instead of the phone having its own phone number, you needed to call an 800 number and then enter a five-digit PIN to ring the mobile phone. Also, placing calls was a pain in the butt. Instead of just dialing the number you wanted to call and pressing Send, you had to go through a complex rigmarole to get the call placed. 

About six months after we got the Prepaid Wireless phone, AT&T announced that they were changing plans as of September 30th of this year. Any minutes you had would be lost, but the good part was that they offered two plans, local and national, with the local much cheaper per minute. Also, they'd assign an actual phone number to the phone, and make outbound calls as simple as with a standard cell phone. Fine.

This morning, I decided to get Barbara's phone changed over to the new plan. I visited the AT&T Wireless web site, and it told me that there was no service available in 27106 or any other Winston-Salem zipcode. That's bizarre, since Winston-Salem is a major town. After some searching around the web site, it became obvious that we were going to have to drive to Greensboro, a 60 to 70 mile round trip, just to get the phone changed over. 

So I decided to see what alternatives were available locally. While I was on the Alltel website, which is pretty pathetic, I found a local dealer, Wilson's Electronics, not five minutes away from us. I called them and asked about Alltel prepaid wireless. During the conversation, I commented on AT&T, and the guy told me that they were an AT&T dealer and could do the change-over to the new AT&T Prepaid Wireless plan right there. So we drove over there.

When they called AT&T to change our phone over, AT&T told them they didn't do that by phone, and the only choice was for us to go to a direct AT&T center. The guys at the local store argued with them, and told them they'd done this before. No dice. AT&T refused to do anything, and insisted that we had to drive to Greensboro. So screw them.

The guys at Wilson's Electronics were apologetic, but there was nothing they could do. He said that about the only prepaid wireless they sold much of was Cingular, and that Cingular was a much better deal--flat $0.35/min pricing, larger service area, etc. The only downside, he said, was that we'd have to buy a new phone. Barbara hated that, because her existing phone is only a year old and has had next to no usage. But I'm not emotionally attached to plastic and chips, so I asked him what the deal is. He said that Cingular had a deal using the same phone we have now, a Nokia (but with a different chip). The phone costs $110, and that includes 100 minutes of prepaid airtime.

That meant that, counting the cost of driving to Greensboro and back, it wouldn't cost us any more to get a new phone with Cingular service than it would to reactivate our existing phone under AT&T's new service plan. And we'd end up with two phones, two batteries, and two chargers. The guy said our current phone couldn't be used with Cingular because the chips were different, but I'd be willing to bet that the chips are interchangeable. So if Barbara ever damages her new phone, I should be able to pull the chip from it and stick it in the old phone and have everything work. I may need to change the ESN or something, but that shouldn't be too hard.

Here are a couple of images of Malcolm in a snarly/fangy mood as I'm petting him. Still pictures can't convey the true level of threat he exhibits. For that, you'd need video showing the lip curling up and down, the fangs snapping together, and audio of the rumbling growls from deep in his chest. Anyone who didn't know Malcolm would swear that he was on the edge of a full-blown savage attack. But, in fact, he's not the least bit aggressive toward people. He loves all people, all the time. The aggression is directed against Duncan, whom Malcolm regards as a deadly threat.

malcolm-fang-1.jpg (203691 bytes) malcolm-fang-2.jpg (226830 bytes)

When Malcolm was young, Duncan used to go after him while Malcolm was getting attention from us. So Malcolm, reasonably enough, decided that he was subject to attack while getting attention from us. So he growls and shows his fangs, just to let Duncan know not to mess with him. It's actually very strange to pet Malcolm while he's in a snarly/fangy mood. Except for the growling and snarling, he acts just like a normal dog. If you stop petting him, he reaches over with his paw to pull your hand back, demanding that you continue petting him. When he wants his tummy rubbed, he'll roll over on his back, just like a normal dog, but growling and showing his fangs the whole time.

If I put my face or hand close to his snout when he's snarly/fangy, he licks me through his fangs. His tongue just comes out and slurps away just like a normal dog, but he's growling and showing his fangs the whole time. And what's really strange is when he combines snarly/fangy mode with his happy-dog face. That combines the typical smiling dog face, tongue hanging out, with his lips drawn back in a snarl. Very strange. But Malcolm is still the most affectionate dog I've ever had, even when he's snarly/fangy.

14:54 - Be very afraid. 

If you thought the DMCA was as bad as it got, there's worse on the horizon. Thanks to reader Miguel Bazdresch for this link. After you read it, let your blood pressure settle down, and then write your congressman. 

Abraham Lincoln once famously described the US Government as a government of the people, by the people, and for the people. That concept is rapidly being replaced by a government of the large corporations, by the large corporations, and for the large corporations. This new measure is to the DMCA as a tornado is to a summer breeze. One thing they're right about, though. It is time to change the copyright laws. We can start by repealing the DMCA and rolling back the duration of copyright protection to 14 years renewable for another 14. Retroactively.

These people need to be stopped, and they need to be stopped now. Write your congressman. Call your congressman. And STOP BUYING THEIR PRODUCTS. Every audio CD or DVD movie you buy from them simply contributes to their war chest. Every movie ticket you buy helps them pay for the chains they'll use to bind you. Every monthly payment you make for HBO or another premium service subsidizes their efforts. Make no mistake. These people are scum. They want to control you, and when you buy anything from them you're helping them do that.

Stop helping them. Please.

 

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, 9 September 2001

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

9:41 - House cleaning and laundry this morning, as usual.

I re-read an excellent book last night. The Red Thumb Mark by R. Austin Freeman was the first "scientific detection" book, published in 1907, and is probably one of the greatest first novels ever written. Some of the Sherlock Holmes books used science and forensics peripherally, but not as the main thread of the story line. Dr. John Evelyn Thorndyke, Freeman's protagonist, was a forensic physician and toxicologist, and the plots of Freeman's Thorndyke novels are based on forensic detection. They are also the first examples of the "reverse mystery" wherein the reader knows (or strongly suspects) from the beginning of the novel who committed the crime, but not how it was committed. The relentless application of logic, spiced in Thorndyke's case with detailed descriptions of forensics tests used at the time, allows the great detective to discover the actual criminal and prove the case against him, in the process freeing an unjustly accused man.

There were great similarities between Arthur Conan Doyle and Richard Austin Freeman. Both wrote exquisitely crafted mystery novels set in Britain around the turn of the last century. Both were physicians, and both abandoned practicing medicine in favor of writing mystery novels. Both were influenced by real forensic scientists, Dr. Joseph Bell in the case of Doyle and Dr. Alfred Swaine Taylor in the case of Freeman. Although written as fiction, the work of both authors was used by Scotland Yard and other police organizations to train their own personnel in crime scene investigation and other aspects of detection that were new developments at the time. 

Freeman was unique in that he practiced what he wrote about. Each of the forensic twists featured in his novels he reproduced in his own laboratory. The science and forensics are accurate for the time. More than accurate, because Freeman himself became a respected authority on many aspects of forensics, including questioned documents, toxicology, firearms and ballistics, and so on.

If your library doesn't have it, you can buy a $7.00 reprint of The Red Thumb Mark from Dover Publications. If you enjoy Sherlock Holmes and have never read John Thorndyke, I encourage you to do so.

 

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.