Monday, 9 September 2002
9:30 - I know this is sacrilege, but the den system just wasn't fast enough for Linux, at least for my taste. I would have thought a Pentium III/750 with 256 MB would have been adequate, but I really want something faster in there. Programs were slow to load, and the speed problem was particularly apparent when I was trying to edit large documents, which is a big part of my work. So I plan to build a replacement system for my main Linux workstation, probably a fast Pentium 4 and 512 MB or more. For the time being, I needed a functional system in the den, so I re-installed Windows 2000 Professional, which runs quite fast.
Windows 2000 is about the only Microsoft software on that system though. I was about to install Office 2000 when I thought better of it. Instead, I installed Mozilla 1.1 and OpenOffice.org 1.0.1, along with WinZIP, WebWasher, and several other utilities. I may not have a Microsoft-free system, but at least I have a Microsoft-application-free system.
Now I need to figure out how to setup Mozilla Mail to allow accessing the same mail store from multiple machines, if that's even possible. Right now, I have Mozilla Mail working as my main mail client on my office system. It stores the mail several directories deep under Documents and Settings on the local drive. I'd like to be able to put the mail store in my directory on the file server and access that same mail store from both my office and den systems. I played a bit with Profile Manager, which looks as though it allows you to specify a different location for mail files, but I'm not at all sure that relocating my mail files to the server, setting up a profile on my office system that points to that directory, and then attempting to do the same from the den system will work. I'm concerned that I'll corrupt configuration files or whatever that weren't intended to be accessed in shared mode among multiple clients. If anyone knows the answer to this, I'd appreciate you letting me know. A search of the Mozilla site didn't turn up any information about sharing a mail store between clients running on two systems.
I'm off to visit my mother this morning, and then I'll come back here to begin some heads down writing all week. Updates may be short and sporadic.
Tuesday, 10 September 2002
9:11 - Today is our 19th anniversary. It's hard to believe it's been 19 years. I remember back before we got married I had friends who'd been married 10 years, 20 years, or more. That seemed like forever then. Now it seems like no time at all.
Barbara stopped by Best Buy yesterday to check out the Sony Clie PEG-SL10 PDA. She decided the screen display was fine, so I told her to just buy it on the spot if she wanted it. She said she wasn't in any hurry so I should just order it on the web. After I checked around, the best price I found from a reputable vendor was about $149 including shipping. Best Buy wanted $149 plus tax, so I called Barbara back and told her for a $10 difference she might just as well grab one at Best Buy.
She did that, and it took me only a few minutes to get the software installed on her system and her Outlook data synched to the new PDA. The unit itself has a little flip-down padded cover for the screen, which seems adequate given that she carries it in a padded compartment of her purse. Barbara said Best Buy had cases for $40, which seems outrageous for something that Sony probably has produced in China for $2 each.
At any rate, Barbara now has exactly what she wants--a basic PDA with a
backlit mono screen and minimal gimmicks. The battery life is rated as 20 days
on two AA alkalines, although Barbara should get much more than that. The rated
life is based on using the unit for 30 minutes a day, and Barbara probably
wouldn't average more than five minutes a day. So a set of batteries ought to
last her something like three or four months. I told her to carry a spare pair
of AA alkalines in her purse, just in case.
Wednesday, 11 September 2002
8:46 - There's nothing I can write that hasn't been written. There's nothing I can say that hasn't been said. Please take a moment this morning to think about those thousands of Americans who were murdered by state-supported Arab terrorists a year ago today.
The US government has done nothing to avenge those victims or to punish those persons and nations responsible. The US hasn't even declared war, a year after the fact, and that is simply outrageous. Instead the US government punishes US citizens by running roughshod over the Bill of Rights. It is time and past time to acknowledge the real enemies here, starting with Saudi Arabia. If the US is to be an empire, it should act like one. The US should declare war on Saudi Arabia today, rain devastation on its cities and people, occupy it, and pump it dry. Using the former Saudi Arabia as a base, the US can then destroy those other nations, starting with Syria, Libya, Iran, Iraq, Afghanistan, and Pakistan, that support terrorism against the US and its allies and clients.
Make no mistake. Islam is the real enemy. As a symbolic response on this, the one-year anniversary of Islam's attack on America, I think it would be appropriate for the US to nuke Mecca into a lake of molten glass. At least I can hope.
Thursday, 12 September 2002
11:33 - I emailed the following warning to subscribers yesterday. I meant to post it here as well, but I never got around to it:
Speaking of mailings to subscribers, I got a couple of fatal bounces. I've pretty much gotten that list cleaned up, but two addresses still bounce consistently. If one of these addresses belongs to you, please send me a working address. I've removed these non-working addresses from my subscriber mailing list.
I talked to my editor yesterday about the third edition of PC Hardware in a Nutshell. He'd really like the page count to go no higher than it already is, which means I have to do some trimming if I'm to fit two new chapters, one on modems and one on networking. So I fired up Open Office and built a spreadsheet to show the statistics for each chapter in the 2nd edition and the changes in the TOC for the 3rd edition.
Basically, I'm consolidating two chapters into one in two places. The chapters on Floppy Disk Drives and High-Capacity FDDs are combined, as are the chapters on CD-ROM drives and CD writers, with probably some minor space savings. That leaves the final chapter count at 28, which is fine. I'll be cutting NT4 coverage dramatically, as NT4 is starting to fade now and will be relatively unimportant by the time the third edition hits the bookstores. That saves quite a bit of space, but balanced against that is that I want to add at least some Linux coverage, to the extent practical. I'll probably make major cuts in the Serial Communications chapter (migrating the cut information to the web site), and probably make minor cuts in Parallel Communications. I'll also be cutting the Tape Drives chapter by removing the stuff about developing a backup strategy and so on (again, moving it to the web site). I'll be removing all the pre-Pentium II stuff from the chapters on Motherboards and Processors, but balanced against that is the new stuff that'll need to be covered. To say the least, adding two substantive new chapters while maintaining the same page count isn't going to be easy.
At any rate, here's the table I made. It'll probably wrap into something ugly on your screen, but it's the best I could do.
I've had lots of requests for things that I simply don't have room to cover, or that aren't really on-topic for the book. For example, lots of people asked about covering stuff like Internet connection sharing, home firewalls, and so on. Much as I'd like to write about that stuff, there's simply no room to do so, and my editor would almost certainly insist it be cut anyway.
At least there will be a third edition. The way the computer book market has tanked, that wasn't a given. Computer book publishers, including O'Reilly, are hurting badly, and the pain has flowed down to computer book authors. A lot of them are making a career change, going back into consulting or getting day jobs. Frankly, I'm not sure I'll even pursue doing a new title. Before the crash, I would have already proposed a new book or two to O'Reilly, and probably had it/them under contract. Nowadays, though, the number of new computer book titles has shrunk dramatically, and publishers are cutting advances and royalty rates to the bone.
My agent told me yesterday that advances for new books have plummeted into the $4,000 to $8,000 range, and there's simply no way that an author can write a book for that amount and still be able to eat. Also, the Open Source phenomenon has spread into computer book publishing. Some (by no means all) OSS folks think money is a dirty word. So we now have some OSS people willing to write books for literally zero advance and royalty rates a small fraction of the norm, which kills the market for professional writers. Some publishers are jumping on that, of course, but in the long run I think it will hurt them. One early wake-up call for publishers is that many of these OSS authors are insisting that their books in effect be GPL'd and made available freely for download.
Although it may superficially seem that the cooperative OSS model should work just as well for writing books as for writing software, I think that will turn out not to be the case, as a quick survey of the documentation typically provided with OSS (or indeed any software) shows. People who are good at writing software are seldom good at writing documentation, and the cooperative model doesn't work very well for writing books (as a glance at a typical Frankenbook shows). Although there are obviously exceptions, including many O'Reilly books, a good computer book is typically written by a person who is a full-time writer. If full-time writers are driven from the market, the next round of computer books that hits the stores is likely to be of much lower quality. We'll see.
I'll be working pretty intensively on stuff related to the third edition, so updates around here are likely to be sporadic and brief.
Friday, 13 September 2002
8:31 - I started Red Rabbit, the new Tom Clancy, yesterday. I'd gotten 50 pages into it, not believing how bad it was, before I went over to Amazon to check the reader reviews. With 199 reviews in, the average rating was 2 stars. That's almost unprecedented for a big-name author like Clancy, considering that many loyal readers give favorable reviews to a bad book whose author they like. A three-star average rating for a major book by a big-name author is disastrous. A two-star rating beggars belief.
I knew Clancy was going down the drain when his first Op-Center book came out. I should say the first Op-Center book with his name on it. When big-name authors attempt to parlay their clout by putting their names on books they didn't write, expect the worst. In the case of Red Rabbit, Clancy should have been ashamed to put his name on a book he presumably did write. I didn't think much of the last couple of Clancys, but this one hits a new low. Don't waste your time reading it. Don't even bother buying it in paperback. Life is too short to waste it reading 600+ pages of this garbage.
I was already considering bagging the book before I read those reviews on Amazon. The consensus seems to be that not only does the book not get better as you get into it, but it actually gets worse. That was enough for me, so I bagged it. That's something I seldom used to do. Once I'd read more than a few pages of a book, I tended to finish it regardless. No more. Life is too short. Barbara has a good idea of which authors I like, so she reserves them for me. I'm going to tell her to take Clancy off my list, at least until he does another book worth reading, which may never happen.
Fortunately, Barbara also brought me The Apprentice, Tess Gerritsen's latest, from the library. I got about half of it read before I finally decided to call it a night last night, and it seems to be up to Gerritsen's usual standard. Tess Gerritsen writes the kind of books that Patricia Cornwell should write. Technically, Cornwell is one of the worst, sloppiest big-name popular authors out there, although her Kay Scarpetta novels remain popular. Gerritsen is a skilled writer, and her background as an internist allows her to write credibly about forensic medicine. If you like Cornwell and haven't read Gerritsen, do yourself a favor and try one of her books. I think you'll like it.
The Apprentice is one of what has become almost a sub-genre, the police-procedural-forensic-serial-killer novel. Barbara tells me that these books are as popular among women as men, if not more so, which surprises me. If I were a woman living alone, I don't think I'd want to read something like this right before I turned off the light for the night. The fiend always breaks in while the victim is asleep and does unspeakable things to her. The moral of all of these stories is that to avoid becoming a victim, one should have a dog (or dogs) and a heavy-caliber handgun (or, better yet, a short-barreled shotgun). If a serial killer ever came for Barbara in the middle of the night, he'd have to deal with Malcolm and Duncan (fanging hazards), Kerry (tripping-over hazard), my riot gun, and (finally) Barbara's .357. Unless he made the mistake of coming in through the basement, in which case we'd hear a scream when George the rattlesnake got him. Oh, well. I guess our home is not a serial-killer-friendly environment.
There was an interesting article in the paper this morning about the geographic distribution of the word used generically to describe soft drinks. "Soda" predominates in the Northeast and California, "Pop" in the Midwest and West, and "Coke" (regardless of the actual drink) in the South. There's even a web page where you can register your preference and view a map of the results. I grew up in Western Pennsylvania, which is definitely in the "pop" camp. Barbara says she grew up (in Winston-Salem) using "soda", which tallies with the "coke/soda" mix shown for this area. The only real anomaly on the map is that Milwaukee and St. Louis, both firmly in the "pop" area of the map, instead use "soda".
9:29 - Just out of curiosity, I went over to Amazon.com and counted the 5-star and 1-star reviews on Clancy's new book. There are now 200 reviews. Of those, 11 are 5-star. Clancy could put his name on the front of a telephone directory and get 11-of-200 5-star reviews, so I discount those. What's really impressive is that there are 71 1-star reviews, which is to say that more than a third of the people who read the book and chose to write a review gave it the lowest possible rating. That's bad news for Clancy, because people who hate your book usually don't come back for the next one.
There's also bad news for people with Fujitsu hard drives. According to The Register (articles here and here), Fujitsu is recalling 300,000 drives made between September 2000 and September 2001. If your system has a Fujitsu drive that's between a year and two years old, it'd probably be a good idea to make sure your data is well backed up. Well, obviously, that's always a good idea, but in this case it seems a particularly good idea.
Saturday, 14 September 2002
9:37 - At first the reports said that three terrorists had been captured on a Florida Interstate highway. Later reports said it had been a prank, that they weren't really terrorists after all Still later reports make it pretty clear that the woman who reported the problem, apparently in good faith, had overreacted when she reported the conversation she claims to have overheard among three men of Middle Eastern origin. Her son, who was sitting with her at the time, didn't see anything to get excited about. Nor did the waitress nor any of the other customers. But on the basis of one unsubstantiated report by a person of unknown credibility and the fact that one of the suspect vehicles had run a toll-collection booth, the police shut down the main East-West connector for the state of Florida for nearly a full day, at incalculable cost.
Did the police overreact? A CNN Poll says that something like 80% of respondents say they didn't. I'd say they did overreact, and badly. And that's not 20/20 hindsight speaking. The police examined the suspect vehicles and found nothing suspicious other than that bomb-sniffing dogs alerted. But those dogs alert on even a residue of explosive, including things as innocuous as fireworks. Having examined the vehicles and found no large amount of explosives, it should have been clear to the police that those vehicles presented no danger to the traveling public, and they should have re-opened the road quickly.
Perhaps it's time we start equipping our state police cruisers with shoulder-fired anti-tank missiles and train officers to use them. That way, the cops could have stood back several hundred yards and detonated the suspect vehicles. After that, they could quickly bulldoze the carcasses off the road, sweep up, and re-open the road to traffic. That would be a quick, cheap way to get the road open and traffic flowing again.
Instead, we allowed a false alarm to close a major traffic artery for something like 18 hours. We are taking counsel from our fears. We've already let the Islamic bastards convince us to gut our air transportation system ourselves. Letting them convince us to close our highways is simply ridiculous. Knowing only what the cops knew at the time, would I have driven past those stopped suspect vehicles? You bet. Even if they had been packed with explosives, I'd probably have been safer simply driving past them on an open road than driving on a normally crowded Interstate.
8:57 - The usual Sunday tasks this morning, doing laundry, visiting my mother, doing the weekly full backup, and so on.
Pournelle posted a piece (here through tomorrow, here thereafter) in which he explains that he'd erroneously used "causus belli" rather than "casus belli" for years, and thanks me for pointing out his error. I thought it was a typo, and was surprised that he brought out his crow-eating graphic. Our phone conversations frequently include one of us correcting something the other believed to be true, and it's more often Pournelle correcting me than the converse. The corrections aren't often for vocabulary or word usage, but that does happen sometimes. Most recently, Pournelle pointed out that I was using the word "fungible" to mean something entirely different than what it really means. I didn't make a public point of correcting my error, and was surprised that Jerry made a point of correcting his publicly.
Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004 by Robert Bruce Thompson. All Rights Reserved.