Monday, 13 May 2002
9:45 - I'm about ready to shut down my account at pair Networks. All of my domains are now moved over to Rocket, the co-located server that Greg Lincoln maintains, and which also runs LinuxMuse. With the DNS changes propagating, my server at pair Networks has dropped from an average of 10,000 hits per day down to 283 hits yesterday, so nearly everyone is now resolving the new server.
The mystery of my light mail traffic yesterday is solved. With pair Networks, each of my domains was assigned an IP address so that, for example, mail.ttgnet.com was physically distinct from mail.hardwareguys.com. In each case, I logged on to the appropriate mailserver with the same account name, thompson. That's no longer the case on Rocket. All of my domains share one IP address, and the account name I use to log on to the mailserver distinguishes one POP account from another.
When I was getting no mail yesterday, I was attempting to log on to different mailservers, but all using the same thompson account. In fact, each time I specified thompson as the account, I was actually logging on to the technomayhem.com mailserver, which actually didn't have any new mail. For example, when I logged on to mail.ttgnet.com and specified thompson as the account name, I was actually POPping from the technomayhem.com mailbox, despite the fact that I'd used the mail.ttgnet.com server name.
Once Greg got me straightened out, I logged on properly to the different mail servers and had something like 170 email messages waiting, which was normal for the amount of mail I'd have expected to accumulate from Saturday afternoon through Sunday morning. I'm still digging through some of that mail, so if you sent me mail and haven't heard from me, it may be a while.
Greg also told me that he's running a spam-stopper on Rocket. It uses various databases, including ones that list open relays, known spammers, and known IP address blocks used by spammers. It appears to have some beneficial effect. In Greg's case, he tells me that he went from getting 3 to 5 spams a day down to getting only 1 to 3 spams per week. In my case, I had only 18 spams when I logged in this morning, which is many fewer than usual. Of those 18, my own local mail filters had caught more than a dozen and moved them to my Deleted Items folder. I keep that folder sorted by subject, which means it takes me literally less than one second per spam to delete the bad stuff.
I've been running Mozilla 1.0RC2 since it shipped Friday, and it seems to be stable. It appears that they've fixed the bug that I found most annoying, which was Mozilla's frequent failure to accept keyboard input. That bug occurred sporadically, so I'm not entirely sure it's fixed, but it hasn't happened on any of the three machines I'm running RC2 on. Despite a few remaining bugs, I really like Mozilla and find myself using it most of the time.
Speaking of Open Source, I've also downloaded the Linux and Windows versions of OpenOffice.org 1.0. I haven't installed the Linux version yet, but the Windows version appears to be fast and stable. I haven't really wrung it out yet, but it appears to be quite usable. There are still problems reported regarding compatibility with MS Word documents, but I haven't encountered any major ones. Most of those I've seen have been pretty minor. For example, yesterday Pournelle sent me a Word document with material that he plans to use in next month's column and asked me to do a sanity check on it. As usual, I embedded comments in it using Word 2000's Comment feature. Using that, I highlight a word or phrase and choose Insert - Comment. The highlighted material is given a yellow background, and the comment appears when you put the mouse pointer on the highlight. I later called that document up in OpenOffice Writer. The comments were present, and moving the mouse over the highlight displayed the comment. The only oddity was that instead of the entire word or phrase being highlighted, the highlight appeared only as a single character following the material I'd highlighted. I can see situations in which that might be a problem, but overall it's usable.
It's back to work for me, and you won't see much here for the rest of the month. I have four major projects that require my attention, along with numerous smaller ones.
10:25 - Something I meant to mention and forgot:
No, I was quite pleased with pair Networks, and wouldn't hesitate to recommend them to anyone.
Greg offered to host my domains on his server. He was already hosting the messageboards for ttgnet.com, hardwareguys.com, and technomayhem.com, so it made sense to move the rest of my stuff there as well. It also gives me a lot more flexibility for the future.
But my decision to move is not a slam on pair Networks by any means. It occurs to me that I'd better make that clear on my journal page.
Tuesday, 14 May 2002
9:25 - All of our sites are now fully migrated from pair Networks to Rocket. I just did a final download of the web sites sitting on the pair Networks server, and then deleted the contents of public_html, so anyone whose DNS hasn't updated will get an error when they try to access any of our sites. There won't be many of those. I checked the pair Stats page, and there were only a couple of hundred hits each day for the last two days. Presumably the laggards will update Real Soon Now.
I mentioned yesterday that there's spam filtering in effect on Rocket. It seems to work, although not as well for me as for some. Just out of curiosity, I decided to leave the spams in my Deleted Items folder until I had a full day's worth. From 9:57 p.m. Sunday through 9:57 p.m. last night, I got 47 spam messages. When I downloaded mail this morning, I ended up with a total of 62, or 15 more overnight. I deleted those from 8:18 a.m. yesterday through the last one I got at 8:08 a.m. this morning, and I again had 47 spams left in Deleted Items.
So it seems that I'm getting something like 50 spams a day now. I don't know how many I was getting before, but I'd guess it might have been as high as a couple hundred a day. I had to make a few changes to my filters, but everything is set now. About 95%+ of the spams are filtered directly to Deleted Items, where it takes me an average of less than a second per spam to get rid of them. I could even improve that a bit. I still haven't bothered to create filters for some real messages that end up in Deleted Items--occasional newsletters and stuff like that. As things are now, my Deleted Items folder contains about 98% or 99% spam and the rest real messages. A few minutes spent creating filters could raise that to very near 100%, at which point I can simply empty Deleted Items without looking at it.
Oh, well. It's back to work for me.
Wednesday, 15 May 2002
8:39 - There was an extraordinary conjunction between Venus and Luna last evening. The crescent Luna, at only 2.5-days old and magnitude -6.8, was separated by only about one degree from Venus, at magnitude -3.8. Other than Sol, these are by far the two brightest objects in the sky, so having them in such close conjunction was an extraordinary event. I don't usually bother photographing things astronomical, but I hauled out the 35mm camera and 400mm lens for this one.
Barbara is off to the gym and store this morning. I'll work all day. This evening, we'll go out for dinner and then head to the Forsyth Astronomical Society meeting. After that, we may head up to Bullington to get some observing in.
10:15 - Patron Subscriber Richard Micko sends this, which is indeed important:
Thanks. I hadn't heard about the Sonic Blue situation, but that is indeed outrageous. It all goes to prove that the old adage is true: we do have the best judges money can buy. As I've said before, protesting loudly to your representatives helps, but the real solution is to strangle these bastards where it counts, by refusing to pay for their products. Don't buy CDs from the music companies. Don't go to movies, and don't rent or buy video tapes or DVDs. Books are another matter. Publishers are not as bad as the music and movie industry, although they're getting there. It used to be that authors and publishers had a balanced relationship. On average, an author and his publisher would profit about equally from a title. That hasn't been the case for years, particularly in fiction. And publishers' new contracts are increasingly simple rights grabs, designed to make authors serfs in every sense of the word. But at least authors can still make a living, albeit not a very good one.
The real problem is that the content distributors--movie studios, record companies, and book publishers--have gotten too large and too few. The markets for authors and other content providers have shrunk and continue to do so. With only a few very large publishers, movie companies, and music companies, things have consolidated to the extent that creative people have few remaining choices. And those huge distributors have political clout which individual artists and consumers do not.
Technology will be the death of those distributors, and they know it, which is why they're trying to lock everything up tight. Even though I predicted all of this years ago, even I am shocked at what the music and movie industry are getting away with. And they'll continue to get away with it, and worse, unless consumers finally cry enough. None of this is about piracy. It never has been. It's about the music and movie industry losing control of distribution. Piracy by consumers is a trivially small issue. Their real worry is that they're losing control of the reins. All of these measures that are supposedly intended to combat piracy are in fact measures to ensure that the status quo is maintained, and that the distributors can continue to skim off the lion's share of the benefit from content producers' labors. Period.
Unless a widespread boycott occurs, I don't see any solution to this problem. The content distributors will continue to kill off useful new technologies like PVRs and e-books, because those technologies threaten the death grip that distributors have on content. The only thing we as consumers can do is refuse to pay for the rope that they're using to hang us.
14:52 - The galley proofs for the new edition of PC Hardware in a Nutshell showed up this afternoon. They're in PDF form, and O'Reilly needs them back fast. I'd originally planned to post the PDFs in the Subscribers' Area, but when I mentioned that to my editor he seemed a bit hesitant. He did finally tell me to go ahead and do it if I wanted to, but it was clear to me that he'd be happier if I didn't, so I won't. Also, the real point of doing that was to get feedback from my subscribers, and there won't be time for me to incorporate any of that and still meet deadline. Finally, I haven't yet gotten the password-protected Subscribers-only area implemented on the new server yet, so there's no secure place to put the PDFs, and O'Reilly would send a hitman after me if I posted them in a public area.
I need to get these things knocked out ASAP and back to O'Reilly, because there's only a month or so left until the book hits the bookstores. As it turns out, there isn't much difference at all between the final draft manuscript I submitted (and made available to subscribers for downloading) and the final PDF version. The production editor at O'Reilly tells me that she herself did the copy edit and that she changed almost nothing except for obvious typos and similar errors. So what you see in the final book will be pretty much what we submitted.
In order to make the deadline, I have to knock out four or five chapters a day, and that means reading the PDFs carefully word-for-word and in some cases comparing what's there with what I wrote. I'll do minor updates to the text. For example, in the PDF it might say "As of March, 2002". I'll change that to "As of June, 2002" (yes, I know, a slight lie) and update the reference accordingly. I don't know for a fact that there are any major changes needed, but I suspect there are a few. Some of those won't get made, either because I don't have time to do the re-write or because the re-write would screw up the current layout too much. So, inevitably, some non-current information will slip through. But my job now is to make the book as up-to-the-minute as possible before it goes to printer. And that will occupy all my time from now until I finish, so I won't be posting any updates here until probably early next week.
If you have any interest in OpenOffice, check out Brian Bilbrey's detailed comparison of Word 2000 and OpenOffice 1.0. I downloaded both the Linux and Windows versions on the day they were posted, but I haven't had a chance to install the Linux version. I have installed and played with the Windows version. There are some minor niggles, but overall OpenOffice 1.0 Writer is very serious competition for Word 2000. As Brian says, OO is now Good Enough that the next time someone asks him which version of Office they should buy he'll suggest they look at OpenOffice.
As far as Linux becoming a viable mainstream desktop OS, the pieces are starting to fall into place, and OpenOffice is a major one.
Thursday, 16 May 2002
8:43 - I finished chapters 0, 1, and 2 yesterday, and started on chapter 3, Motherboards. I have to talk to O'Reilly this morning about whether I can add some material to cover the new 850E, 845E, 845G, and 845GL chipsets. I may be able to just expand the existing table, or I may relabel that table "First-generation Pentium 4 Chipsets" and add a second table to cover Intel's second-generation P4 chipsets. Or I may only be able to add a sentence or two that mentions the new chipsets and refers readers to a page on HardwareGuys.com. It all depends on how badly added material would mess up the layouts.
We went out to dinner and the astronomy club meeting last night, and the whole time I was thinking about how I should be back here working on the PDFs. Heads-down work today and for the next several days until I get these knocked out.
Friday, 17 May 2002
8:43 - I put in about 12 hours yesterday working on final changes to the text. I've finished chapters 0 through 7, and am part-way through chapter 8. Fortunately, Emily (my production editor at O'Reilly) has a sense of humor. I told her that I'd need to make some significant changes to Chapter 3, Motherboards, and Chapter 4, Processors, to account for the fact that this month Intel has introduced completely new lines of chipsets, motherboards, and processors. I'm still waiting to get official data on some of the new chipsets, so that I can avoid writing an "alternative history".
For example, many people expected the new 850E chipset to use the ICH4 southbridge, which supports USB 2.0 natively. I'm sure there are books out there that state incorrectly that the D850E uses ICH4. In fact, the 850E chipset uses the old ICH2 southbridge, and so supports only USB 1.1 natively. That's further confused by the fact that some variants of the new D850EMV2 motherboard include embedded USB 2.0 support. Presumably, Intel used the same workaround for these models that they did for the D845BGSE, which was to include an NEC USB 2.0 controller on the motherboard.
As much as I hate it, I'm forced to leave some stuff as is that I'd prefer to re-write. The problem is that lengthy adds or deletes would butcher the current layout, not to mention playing Hobb with the index, and probably ruining other stuff that I'm not even aware of. So I'm just doing the best I can to make the book accurate and current as of now. It's scheduled to hit the bookstores in about a month, so we're really down to the wire here.
It's back to work for me. I'm going to see how many chapters I can get cranked out today, although my weekend is also going to be spent working on the galley proofs.
Saturday, 18 May 2002
9:10 - I've now worked my way about half way through the manuscript, making changes and updating stuff. One interesting aspect of doing this word-by-word read-through is that I'm actually getting to read the book, as opposed to writing it. As a reader, I have to say that I think the authors did a pretty good job.
One minor problem I've encountered is that the editors have made some arbitrary changes to the text to make me sound Politically Correct. For example, here's one of my comments:
and, worse still:
I'm pretty easygoing about editors making changes to stuff I've written, but it does make me see red when I'm made to sound PC, let alone when a proper English sentence I'd written is changed to mixed singular and plural.
And it's back to work for me. There was to have been a public observation at Pilot Mountain tonight, but it's raining now and the forecast is for heavy clouds tonight, so Barbara and I plan to give this event a miss. I hope to have most of the final editing done by Monday, at which point I can get back to work on the other critical items on my to-do list. There's no rest for the weary.
Sunday, 19 May 2002
8:37 - I've now worked my way through the end of Chapter 21, page 586, so I have only a couple hundred pages left to review. The whole book, counting front matter and index, totals something like 800 pages, so I'm in the home stretch now. I'm going to try to finish it up today, so my editor at O'Reilly will have it all when she gets into work tomorrow morning.
Regarding my comments yesterday, C. E. Myers sends an interesting link.
But I don't want to overemphasize my objections yesterday. In fact, the editing job on this material is by far the best I've ever received. I sent the following message yesterday:
For those who've asked, the new edition of PC Hardware in a Nutshell should hit the bookstores in about a month. If you want to order a copy now, Amazon and other on-line sellers are accepting pre-orders. This is turning out to be a very good book. We only have half the page count to work with that Upgrading and Repairing PCs and similar monster PC books use, but I think we've done a pretty good job of fitting more actual useful information into that smaller number of pages.
And it's back to work for me. With only 200 or so pages remaining to be checked, I'll also do some of my regular chores today, like doing the laundry. Sitting staring at a screen and checking facts for 12 or 14 hours a day is getting old.
Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004 by Robert Bruce Thompson. All Rights Reserved.