Daynotes Journal

Week of 12 March 2001

Latest Update: Friday, 05 July 2002 09:16
 

Order PC Hardware in a Nutshell from Fatbrain.com

Monday, 12 March 2001

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

In general, things should be pretty transparent for subscribers. When you try to access a document that's subscriber-only, you may be prompted for your username and password. If so, enter them. If you mark the "Save this password in your password list" check box, you should be prompted only once per session, and the name/password should be filled in for you. All you need to do is click OK to access that document and any other restricted links you click on during that session.

For non-subscribers, things may be a bit frustrating. If you click on a restricted link, you'll also get the password dialog. The only way to gain access to that document is to subscribe, which is the point of the whole thing. I've set up the search engine to index the entire site, so searches may turn up documents that you can't access. Sorry.

As of this week, I've decided to shift to the Pournelle Method of updating. In the past, I've typically posted an update each morning around 9:30 local time. If I had more to write about during the day, I'd add it to the next day's entry, but not publish that until the following morning. I did that because FrontPage 2000 won't publish through my proxy server, which means that I have to publish from meepmeep, the Roadrunner box. But that's a minor inconvenience, so from now on, I'll publish whenever I've written something. That might mean I'll update this page twice or more in a day, but it also may mean it'll sometimes be a while between updates.

Also, in the past, when I did do a second update in one day, I'd move the "current" bookmark to the beginning of the new update. Several people have told me that they found this annoying, so from now on I'll leave the current bookmark at the top of the current day no matter how many updates I've done that day. You should be able to tell if there's new material based on how your screen behaves when you refresh.

This morning I start heads-down work on updating PC Hardware in a Nutshell, Chapter 2, Working on PCs. A quick read through tells me that it's in pretty good shape as it stands, but in need of some minor updating. One thing a couple of people have commented on is that I should consider adding some pictures, so I'll look at the chapter to see where it makes sense to do that. If you have any comments, criticisms, or suggestions about Chapter 2, now is the time to post them.

Speaking of which, I've posted my first draft of Chapter 1, Fundamentals, on the Subscribers' Home Page. The file is a zipped Word document file of 600+ KB, with embedded images. If you've ever wondered what one of my chapters looks like before the editor gets to it, now's your chance to find out. I've also created a new topic in the Subscribers Only forum of the HardwareGuys.com messageboard, where you can post and read comments about the chapter. Because this is only a first draft, I will have the chance later to make changes/additions to it based on your feedback.  If you do have any comments or suggestions, please make them on the messageboard rather than sending me private mail (or, worse yet, marking up the document and sending it back to me). That way, all comments are in one place and it'll be a lot easier for me to consider them.

Just a reminder. This document is copyrighted and confidential, so please do not distribute it. I am posting it solely for the purpose of soliciting your comments with the intention of improving the chapter in a later draft.

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, 13 March 2001

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

All day today is devoted to working on Chapter 2, Working on PCs.

More later, if there's anything worth writing about.

Now, this is interesting. One of the most aggravating things about Outlook 2000 not exiting properly is that I frequently swap back and forth during the day between working at my desktop system in my office and my notebook system in the den. I'll close Outlook 2000 on my desktop system, go have lunch, sit down on the sofa, and fire up my notebook. When I start Outlook 2000, I'll get an error message, at which point I have to get up, go into my office, fire up Task Manager, and kill the OUTLOOK.EXE process (I'm still not sure why the process shows up in all upper case, but there it is). Or vice versa. I'll shutdown Outlook 2000 on my notebook system, go back into my office, try to open Outlook, and find that as far as Microsoft is concerned my PST file is still locked by Outlook running on my notebook.

So, this morning (as usual) I checked my overnight email on my notebook system, read a few web pages, and so on. I then closed the cover on the notebook (I leave my Compaq Armada E500 plugged in all the time and simply put it to sleep by closing the lid when I'm through using it) and went back to my office. I opened Outlook there, and used it normally until lunch time. After lunch, I fired up the notebook. When it awoke from suspended animation, it was still displaying a web page that I'd left up on screen when I'd put it to sleep. But I heard the characteristic chime of an Outlook reminder.

Sure enough, when I minimized the web site, there was an Outlook reminder message up on my screen, along with a new mail icon down in the task bar. When I clicked to close the reminder message, Outlook popped up an error message saying that it couldn't write to the data file. It was only then that I noticed that I'd completely forgotten to close Outlook this morning. It was on the Task Bar as a minimized process, where it'd been sitting since early this morning. In fact, when I maximized it, it was still showing the mail messages from this morning that I'd long ago responded to and moved to folders. So I closed Outlook. Again, I got an error message about not being able to write to the data file, and eventually I had to go in and kill OUTLOOK.EXE in Task Manager.

What I can't figure out is how I was able to access my Outlook data on my desktop system, since the file was already open on the notebook, albeit in slumber mode. That's never happened before. Ordinarily I'd get an error message on the desktop, but this time I was able to open and use Outlook normally. Very strange.

And, speaking of very strange, my screen saver problem is cured, kind of. I finally grew so aggravated about the screen refusing to blank that I did something I almost never do--re-enabled power management in Windows 2000. I set the monitor to power down after one hour of inactivity, and that works fine. What's really odd is that the software screen saver also works fine. It was set to blank the screen after ten minutes of inactivity, and it does that now. And 50 minutes after that the power management kicks in and puts my monitor into sleep mode.

Okay, I figured, maybe something happened to fix the screen saver (I hate not knowing what Windows 2000 is doing, but there it is). So I disabled power management, assuming (or at least hoping) that the software screen saver would continue to work. No dice. With power management disabled, I'm back to no screen saver at all. With power management enabled, both it and the screen saver work as expected. I hate Windows.

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, 14 March 2001

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

Much of what Microsoft does is designed to lock users into Microsoft products forever. If you doubt that, just try getting your data out of Outlook and into some competing product. Microsoft designed Outlook as a Hotel California application. They provide all sorts of nice import capabilities, but support for export is pathetic. Once you check in to Outlook, you can never leave. 

And I think they plan the same thing for Linux support in Dot-Net. They'll provide all sorts of nifty tools to lure Linux sysadmins into Dot-Net, AKA the Microsoft Cunning Plan for World Domination, and let them get comfortable with it. There'll be service packs and updates to add useful features, all intended to lure users deeper and deeper into the Microsoft Way and get Microsoft's hooks deeper and deeper into their core operations. Then, a few years down the road, Microsoft will suddenly discover that there's not enough demand for Linux support to continue providing it. And all those poor foolish Linux admins will find that their only real option is to get rid of Linux and start running a Microsoft OS.

Count on it. Microsoft isn't stupid. They realize that Linux is a deadly threat in server space right now, and a developing threat on the desktop. One of Microsoft's top priorities must be to kill Linux, whatever it takes. Lobbying legislators with sanctimonious concerns about intellectual property. Inviting Linux admins to share the benefits of Dot-Net. Whatever it takes. Come into my parlor said the spider to the fly.

No, thanks.

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, 15 March 2001

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

Here's something that I find unutterably disgusting. Hewlett-Packard used to be one of my favorite companies. That was before Carly Fiorina took the helm. I don't know if it all comes down to Ms. Fiorina, but in the last few years HP has slid from being one of the best technology companies around to one of the worst. In particular, their Windows 2000 driver support (or lack thereof) has been a nagging issue for a lot of users, including me. 

About 18 months ago, not realizing that HP's slide from grace was imminent, I bought an HP 6200C scanner based on HP's reputation and my prior good experiences with HP gear. That scanner has both USB and SCSI interfaces. I've never gotten it working very well under USB. I did at one point have it running on a Windows 98SE box, but when I attempted to relocate it to another Win98SE box I never could get it running under USB. Nor was I able to get it running via USB under Windows 2000. For a while, I had it working on a SCSI system, but I'd really like it to be connected to Barbara's main system, which runs Windows 2000. That has SCSI but the external connector is of the wrong sort to use the cable I have with the scanner. I suppose I could just buy another SCSI cable, but I'd really like to get it working under USB. 

So I visited the HP web site. Right there on the main page is a big box marked "Drivers". Great. I clicked on that and entered ScanJet 6200C in the search box. A page popped up immediately that listed drivers for Windows 95/98/Me and Windows NT/2000. The problem is, most of them were missing. Under the Windows 2000 section, for example, there were exactly two links. One for the "HP Copy Utility Update for the ScanJet 6200/6250C Series" dated 3/5/00, and a second link that pointed to a CD ordering page. That, at least, was better than the NT4 section, which had only the link to the CD ordering page.

So I clicked on the link to display the CD ordering page. HP informs me that "The drivers for these products will be available on the web in the future, until this process is complete this software will only be available in CD form through your regional distribution center." What they failed to mention was that these drivers used to be available for download. I know, because I downloaded them, and still have them in my downloads directory. But if you don't have them and need them, apparently your only option is to call a non-toll-free telephone number and give them your credit card information. They charge $9.95 for the CD plus some unspecified amount for shipping. I suspect they charge even for basic shipping, because FedEx is listed as an option.

Obviously, HP has decided to turn drivers into a profit center. And if you doubt that, consider that it costs 50 cents to make a CD. Anyone can make a good profit selling a 50 cent item for $9.95 plus shipping, especially with no 800 number to pay for, and most especially when one has a captive audience for the product. If this isn't a case of HP ripping off users for nickels and dimes, what is it? If that's not the case, and if the drivers are available on CD, why wouldn't they post those drivers for download as any other company does?

I no longer recommend any HP product, and I certainly won't buy one myself.

Here's an idea that just came to me out of nowhere. In exchange for the protections that software vendors enjoy under copyright and trademark laws, UCITA, DMCA, etc. etc., how about we modify those laws slightly? Any time a software vendor releases a new version which is other than free to licensed users, the immediately preceding version of that software--including source code, internal documentation, file storage formats, etc.--becomes public domain. To avoid software makers playing games by updating only some modules and so on, we'd have to specify that updating any module and charging for that update counts as updating the entire product of which that module is a part, that source code for all libraries must also be supplied, and so on.

The more I consider all the implications of such a change in the law, the more I like it.

Here's an interesting article about the Rambus trial. It appears at least possible, and perhaps even likely, that the ruling, due as early as this morning, may void the claims of Rambus to dominion over SDRAM. It also appears possible that some current and former employees of Rambus may be charged with criminal fraud and, if convicted, may do jail time.

I am not a lawyer, and I have no special access to the documents and facts in the continuing dispute between Rambus and memory manufacturers. But from reading publicly available documents and listening to the dueling allegations, it has always seemed to me that the memory manufacturers were in the right and that the Rambus patents in question should be voided. 

As I understand it, the other memory manufacturers claim that Rambus was a participant in a JEDEC conference years ago. The memory manufacturers got together with the idea that each would pool its own proprietary memory technologies with those of the other members of the group, and that all members of the group would be entitled to use all of those pooled technologies. The story is that Rambus sat in on the conference and then promptly ran out and patented a bunch of technologies that had been contributed by others with the idea of them being freely available to all group members. That, presumably, is where the criminal fraud charges arise. If that's all true, I wonder if the RICO statutes apply. 

If all of that happens, RDRAM will have to stand on its own against DDR-SDRAM. Let's see, RDRAM is much more expensive, has slower throughput, and has much higher latency than DDR-SDRAM. Duh.

It looks to me as though Intel had better get on the stick and start shipping a DDR-SDRAM chipset for the Pentium 4 soon if they expect to sell many Pentium 4s.

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, 16 March 2001

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

Actually, I'm surprised that European nations haven't taken stronger steps to avoid the spread of the disease across borders. Until the problem is isolated and eradicated, I think the authorities should be quarantining affected areas. I understand that some countries are making arriving airline passengers walk through pools of disinfectant and spraying automobiles with disinfectant at border crossing points. Those seem reasonable steps, but I'm not sure why the unaffected countries haven't simply closed their borders to travelers from affected countries. I've been told that the potential losses from a widespread hoof and mouth outbreak in the US are in the hundreds of billions of dollars, and the danger always exists that hoof and mouth could become endemic in the wildlife population, resulting in continuing re-infections of domestic livestock. 

With that at risk, a ban on European meat and dairy products seems a very moderate reaction, and a ban on travel from afflicted areas would also be a reasonable step. That would certainly inconvenience business and vacation travelers from the UK and Europe, but on a cost/benefit basis it'd be hard to argue against such a step. Apparently, the hoof and mouth problem has already spread to South America, whether by contaminated imports or on the shoes of airline passengers no one is sure, so there's no doubt that the casual spread of the disease is a very real risk.

One has to have a great deal of sympathy for the Brits. First, BSE, and now hoof and mouth. I think the US should do everything possible to help the Brits once the outbreak is contained. But until that happens, protecting our own livestock has to take first priority.

Another big article in the paper this morning about the failure of advertising on the Internet. Click-through rates have fallen to nearly zero, and no one can figure out why Internet advertising doesn't work. But they're looking at it the wrong way. They believe that, for some inexplicable reason, Internet advertising doesn't work as well as TV or print advertising. That's not true, of course. Internet advertising works just as well as TV or print advertising, and probably better. Which is to say that it doesn't really work very well at all. The problem is that Internet advertising is being held to a higher standard than traditional advertising. With Internet advertising, it's immediately obvious that it doesn't work very well. Traditional advertising doesn't work, either, but that's not so apparent because there's no means to track it.

Think about. When Gateway runs a television ad for their new computers, does anyone at all turn off the TV, jump up, run out the door and head for their nearest Gateway dealer to buy a new computer? Or do they get out their credit cards, pick up the phone, and call Gateway to order a new PC? Of course not. They just ignore the commercial, if they even see it. Chances are, they got up to use the bathroom or go to the kitchen to make a snack, so they probably didn't see the commercial anyway. Or they're watching the show on tape or Tivo and fast-forwarded through the commercial. So Gateway wasted the money they spent running that commercial. Some people will order Gateway computers, of course, so the ad agency can claim that those buyers bought because they saw the commercial, or at least that Gateway sold more computers because they ran the ad. Not provably true, but not provably false, either.

With Internet ads, on the other hand, it's easy to track exactly what happened. Or, what didn't happen, more like. For every 200 or 300 times that banner ad for Gateway computers is viewed, there might be one click-through. A pathetically small percentage in one sense, but it begs the question of what percentage of people who "view" a TV or print ad take any action based on that. One in a thousand? Almost certainly not. One in ten thousand? I suppose that's possible, but I doubt it. 

If my speculations are correct, Internet ads are actually something close to or exceeding an order of magnitude more effective and efficient at generating action on the part of potential buyers. That's not considering the cost of running an Internet ad versus a TV or print ad. If that's taken into account, Internet ads are probably two or three order of magnitude better. But even at that, they're not cost-effective. So how much less cost-effective are traditional ads, and how long will it be until companies realize that?

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, 17 March 2001

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

Barbara is off this morning for a day-trip bus tour with her parents and sister to Charlotte, where they'll attend Phantom of the Opera. She won't be home until late this evening, so food arrangements and dog management are up to me for the day. I tried to convince her to take Malcolm along--he's never been to an opera--but Barbara vetoed that idea.

I've gotten several private responses to my diatribe at Hewlett-Packard the other day. These came from HP employees in both the US and Europe, and what's interesting is that they strike a common chord. All of them sympathize with the problems I've had and agree that this type of thing is happening more as of late than it should be. All of them mention that HP had become huge and unwieldy, that Carly Fiorina is doing her best to cut out deadwood and rationalize operations, and that perhaps all of this is proceeding so quickly that some balls are being dropped. Interestingly, all of them seem fairly upbeat about expectations for the "new" HP. We'll see. But if the new HP is to be anything like the old HP as far as customers are concerned, HP must refocus on customer satisfaction.

The latest wife/significant other/concubine to join the ranks of the daily journal keepers is Ann Dominik, the wife of John Dominik. She's just gotten her main page up, with a daily journal to come, but it's obvious just from her main page that she'll be worth reading. Unfortunately, it appears that she is a liberal Democrat, and I'm pretty sure she'll disagree with my position that giving women the vote was a mistake. But check out her site and see what you think.

My notebook system has begun behaving strangely. I keep it on all the time, flipping the screen closed to put it to sleep while I'm not using it. Until yesterday, that worked without problems. Yesterday morning, I opened the notebook to rouse it to life, clicked on the Outlook icon, and got an error message because the notebook couldn't connect to the network. I clicked the Intel 2011 Wireless LAN icon in the task bar, and the applet told me that the system wasn't connected because it was out of range. That was odd, since the notebook is probably 30 feet from the Access Point. So I rebooted the system and everything worked normally, with the signal strength flipping between good and very good. Hmm. I passed that off as a momentary glitch.

Then, yesterday afternoon, I again sat down on the sofa, picked up the notebook, and fired it up. Once again, no signal. I didn't want to reboot the notebook, because I tend to "bookmark" web sites by minimizing them. I typically have anything from half a dozen to perhaps 25 instances of IE active in the task bar, each with a web site that I wanted to "keep" for one reason or another. So this time, I fired up the load/eject applet and told the system I wanted to eject the 802.11b wireless network adapter from the notebook. I did that, and then reinserted it. The network connection came up fine at 11 Mb/s, again with signal strength varying between good and very good. 

Throughout the day, last evening, and this morning, the same thing kept happening. Every time I put the notebook to sleep, it'd wake up having lost its wireless network connection. I could live with ejecting and re-inserting the wireless network card, except for one thing. When I eject it, the notebook emits a loud beep-boop. When I reinsert it, the notebook emits a loud boop-beep. Both of those noises, particularly the boop-beep, scare the hell out of the dogs. That's not so bad for Duncan and Malcolm. They're young and agile. But it's really pathetic watching 13-year-old Kerry scrabble around trying to get to his feet so that he can flee the boop-beep monster at his best rate of speed.

The way I see it, this problem could be caused either by the Intel wireless networking hardware/software, or by Windows 2000. Guess which one I suspect. 

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, 18 March 2001

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

That means there now are at least two countries, Red China and Germany, who don't trust Microsoft to the extent that they have banned use of their software, at least in some situations. Red China has its own Linux development effort, and presumably Germany won't be far behind. 

Ann Dominik takes strong exception to my well-reasoned position that it was a mistake to give women the vote. Not to mention allowing them to serve on juries. Or, heaven forbid, as judges or in political office. O. J. Simpson would have been tried and hanged within a week had it not been for the soft-hearted influence of women on the justice system. Women simply can't be trusted to make the tough decisions. Nor, apparently, to engage in reasoned arguments. Instead, for some reason, Ann chose to rebut my arguments by spitting on her husband's monitor. But, despite her delusions about women's proper role in society, Ann is smart, funny, and writes well. If you enjoy reading daily journals, check her page out.

Having on two occasions won the coveted Bob Walder Internet's Most Boring Photo trophy, I feel obligated to keep my hand in. Yesterday as I was walking the dog around the front yard, I noticed how lush and green Barbara's lawn appears. Seemingly only yesterday it was brown and dead, but now it looks healthy and happy. So, without further ado, here is my current entry in the Internet's Most Boring Photo competition. Yep, you guessed it. Grass growing.

Next week, paint drying.

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]