|
Week of 13 September 1999 Sunday, 19 September 1999 09:46 A (mostly) daily journal of the trials, tribulations, and random observations of Robert Bruce Thompson, a writer of computer books. |
|
|
Monday, 13 September 1999 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Well, I know I'd planned to go to evening updates rather than morning updates, but this is ready to go now, so I'll go ahead and post it. I think what I'll start doing is updating whenever I have a spare moment, which means this page might not be updated again until tomorrow evening. Or it may be tomorrow morning. Hell, it could be this evening, depending on when I have time to work on it. * * * * * I'd intended to start an experiment this week. As I mentioned yesterday, the length of the consolidated journal page was getting too long for comfort. FrontPage has an estimated download time indicator at the bottom, and last week's page approached two minutes. Granted, that's at 28.8, and assumes no compression, so few readers would have to wait that long. Still, at almost 150 KB, that's a big page. So I broke my single journal page in separate journal pages and mail pages. Getting all the new pages and links set up took me about an hour. But in the mean time, I got two messages from readers, both of whom pretty emphatically told me to leave things as they were, or at least to ask my readers before I made any significant changes. So, I spent another hour or so last night reverting back to the old style. If you want to look at the new style, you can see it here for Daynotes, and here for Mail. * * * * * I now have yet one more thing to keep up with. My friends John Mikol and Steve Tucker came over Saturday to help with a couple of projects. John brought back my 6 foot bellhanger drill bit and guide, which he'd borrowed two or three years ago and forgotten he had. I in turn returned Steve's butt set and toner, which I'd had for the better part of a year and hadn't used. The main project was to get my generator operational. We bought it six months or so ago, and it's been sitting in the basement unused ever since. I'd been avoiding running it for the first time, because I knew that once I did that I'd need to run it every month or so to keep it from corroding. John convinced me that I should run it. The first order of business was to build a backfeed cable. For those who aren't familiar with generators, the right way to connect a generator to your home wiring is to purchase an isolation switch and pay an electrician to install it. The problem is that buying such a switch and getting it installed can easily cost more than $1,000, so very few people actually do what they're supposed to. Instead, what many do is use a backfeed cable. The power company frowns on this, to say the least, because it kills linemen. In theory, a backfeed cable is perfectly safe. When the power fails, you turn off the main breakers, which disconnects your home from the utility wiring. You then plug the backfeed cable, which is a simple male-to-male extension cord, into your generator and the 240 volt receptacle for your electric clothes dryer. When the generator powers up, it puts voltage on both legs of the 240 volt line, and you can run anything in your home from generator power, turning individual breakers on and off to prevent the total load from exceeding the generator's capacity. The problem, of course, occurs when someone connects and powers up the generator but forgets to throw the main breakers. That puts voltage on the power company's wires. What's worse is that a transformer works just as well going backwards as it does going forwards, so the 240 volts you put on the public wire can be stepped up to 50,000 volts or more. Some unsuspecting lineman, assuming that your end of the cable is dead (as it should be) can get a 50,000 volt surprise. Although I plan to install a proper isolation switch, I also wanted to have a backfeed cable just on general principles. I acquired fifty feet of heavy-duty cable and the necessary connectors. This cable looks almost exactly like a garden hose and has four 8-gauge wires in it. I'd never built a high-amperage cable, and didn't feel comfortable attempting it. John offered to build it for me, and I'm glad he did. Once he had the cable built and tested out with his meter, we decided to make sure everything fit. The three of us went to the main breaker panel and located the main breaker. We stood there scratching our heads, because the main breaker was only 60 amps. No way is that big enough to service this house. We *knew* that. Standard service is 100 amps, and most newer homes have 200 amp service. That should have been a major clue, but we ignored it. We threw the main breaker and connected the backfeed cable to the dryer receptacle. Fortunately, we were doing things in the proper order. John was examining the end of the cable intended to connect to the generator when Steve and I heard a shout. John said a very bad word. He'd gotten 240 volts across his thumb, thereby craftily saving wear and tear on his meter. Fortunately, John was not hurt badly, although he burned his thumb and it could have been much more serious. 240 volts is not something to mess around with. The problem, of course, is that our main breaker isn't really a main breaker at all. The 60 amps it provides powers everything in the house except for the heavy stuff--electric dryer, kitchen range, etc. The feed for those appliances goes directly from mains power to the appliance via a dedicated breaker, bypassing the main breaker entirely. The upshot is that I can't backfeed from my generator, which I probably shouldn't have been considering anyway. The only solution, short of installing the proper bypass switch, is to have the power company come out and install a service disconnect switch at the meter. And that's something I'm going to have done anyway. I'm not comfortable not having a way to disconnect all power from the house. * * * * * This from Hobsons212@aol.com:
You're right. I should have asked. In fact, I did ask at one time, and probably two-thirds to three-quarters of the people who responded suggested I keep it at one page. But I've recently gotten several messages from people complaining about the size and download time. That strikes me as kind of strange--"I love your page and read it every day. Please make it smaller..."--but I can see their point. At any rate, it is now too late. Immediately after posting that update Sunday morning I started working on splitting things up, and it's now a fait accompli. If I'd gotten your message before I started doing that, I would have done as you suggest. But you took at least five whole minutes after I posted that update before you responded, so the change ended up being made. I guess I'll keeping doing separate pages for at least a couple or three weeks to see how things go. If enough others feel the same as you do and let me know about it, I'll change back. The trouble is, making any change to a page draws flack one way or another. My guess is that a few people will dislike the new method enough to write and tell me about it. A few more will like the new method enough to write and tell me about it. The vast majority probably won't care much either way. So I'll try it for a while and see if a consensus develops. Thanks for letting me know your thoughts. I depend a great deal on reader feedback. Otherwise, I'm operating in a vacuum. * * * * * I guess this next counts as mail. I was reading AnandTech yesterday evening, and saw that he'd reviewed another Palo Alto computer case. I sent Anand the following: I've been following your case reviews with interest. I consider all of those you've reviewed as mid-range (the Dell/Micron OEM versions) at best, and as low-end junk cases at worst. There's a much better alternative, and I'm surprised you haven't reviewed it. PC Power & Cooling (http://www.pcpowercooling.com) makes the best cases and power supplies available, bar none. Solid construction (a tad old-fashioned in some respects), heavy gauge metal, no sharp edges, etc. What's more important is the power supply. You seem to regard power supplies as essentially commodity items, and nothing could be further from the truth. There's far more to evaluating a power supply than just looking at how many watts it delivers and how much air it moves. If you ever disassemble one of those junk power supplies that come with the cases you've reviewed and then compare it to a disassembled PC Power & Cooling power supply, the differences will be obvious to you, even if you don't know anything about power supplies. I just checked the PPC web page. Their equivalent to the Palo Alto ATCX case is their Personal Mid-Tower ($65 without p/s direct from PPC, or cheaper via mail order) and the Standard 235 ATX power supply ($59 direct, cheaper mail order). The total $124 compares favorably with the Palo Alto ATCX, and the PPC 235 watt power supply is far superior to the Delta Electronics power supply. The PPC case has six drive bays, including three external 5.25" bays, and can be upgraded for $5 to eight drive bays. There's just no comparison. As far as the power supplies, just compare such things as load regulation, line regulation, ripple, hold time, overvoltage protection, and overcurrent protection. Just look at the specs and you'll easily see which one's a high quality power supply and which one is junk. And note that this PPC model is one of their "economy" models. Their mid-range and high-end ones are better still. I'm surprised that you don't use a PPC case and power supply for your personal system. Most people I know who are really knowledgeable about PCs wouldn't consider using anything else. * * * * * This from J.H. Ricketson [culam@neteze.com]:
Oooh. Good point, although your final paragraph was a low blow. To prove I'm not like Microsoft, I'm going to spend an hour or so reverting. That is, I'd already redesigned things, created the new pages and links, etc. but I'll go back to the old way until I find some consensus among my readers as to which they'd prefer. So far, the vote is two in favor of the old way and zero in favor of the new way. * * * * * This from Robert Rudzki [rasterho@pacbell.net]:
I'm not sure where FrontPage gets its estimate either. I connect 99% of the time at 31.6, and when I downloaded last week's full page it seemed to take perhaps 30 seconds, although I didn't measure it. At any rate, enough people have mailed me to say that it doesn't take anywhere near two minutes to download the page that I'll assume that download time is not a problem. A couple of people had mailed me to complain about long download times, so perhaps they're using 14.4 modems or something. I didn't watch Braveheart last night. Barbara and I watched it when it was on Showtime a couple or three years ago. Given the level of violence I was rather surprised that Barbara was not only willing to watch it, but eager to do so. As you point out, the history wasn't the best, to say the least. Still, it was a pretty decent movie. My grandmother told me many years ago that I was a direct descendent of Robert the Bruce, but I don't really know if that's true or not. My brother was named after William James. I'm not related to Mel Gibson, as far as I know. * * * * * This from Bo Leuf [bo@leuf.net]:
The last time I heard, Microsoft was still claiming that they'd ship W2K this year, although I doubt they'll make it. Of course, they can choose to define W2K as "finished" any time they want, but I think they'd be foolish to ship it sooner than mid-2000, if not late 2000. All that doing that would buy them is meeting a date, which is surely less important than the other considerations. With Y2K and its aftermath, no one is going to be adopting W2K in large numbers any time soon. Even companies that survive 1/1/00 with few problems are going to be very busy throughout CY 2000 catching up on stuff that they had to delay to get ready for Y2K. I don't see widespread adoption of W2K happening much before 2001, and many companies are likely to wait until 2002. So what's the rush to ship a product that no one will use? W2K Professional is a lot further along than Server, which still has major problems. I haven't seen RC2 yet, but RC1 isn't what I'd call an RC. More like a late alpha or early beta. If Microsoft is smart, they'll spend the rest of this year and most of next getting W2K ready. They can't afford to ship a buggy W2K, and shipping earlier than late next year buys them nothing. Of course, the same could have been said of Office 2K, and they certainly shipped that one a year before they should have. * * * * * This from Bo Leuf [bo@leuf.net]:
Yes, I think base-60 is probably a reasonable explanation. Thanks. * * * * * This from Paul Christy Jenkins [cjenkins@ssc.wisc.edu]:
Thanks for the kind words. Yes, I'm beginning to conclude that the Frontpage download estimator is of limited value. I typically get 3.0 to 3.3 KB/s downloads, which would make last week's ~150 KB page less than a minute. That's without compression, and I often get twice or more the throughput when downloading text-based stuff like this page. So your 15 second figure is credible. * * * * * This from Scott, Terry [tscott@atinc.com]:
Don't feel too badly. The same thing has happened to me, and I'm sure to many of my other readers. The problem people seldom think about when dual-booting Windows NT 4 and Windows 98 is that the former is not PnP and the latter is. So stuff that is automatically arbitrated under Win9X isn't under NT, which can cause no end of problems. I went absolutely nuts the first time I tried to install an ISA sound card in a dual-boot system. It worked fine under Windows 9X, but I couldn't even get NT to see the card. The solution was to run PNPISA.INF, which NT requires to use a PnP ISA card. That to my mind will be one of the biggest advantages of Windows 2000--its support for PnP. * * * * * As of right now, FrontPage is estimating that this page will take 25 seconds to download, which I doubt. * * * * * Afternoon: Hmm. This ain't good. My mother just called upstairs to tell me that she was watching The Weather Channel, and that hurricane Floyd is worse than Andrew or Hugo. Sustained winds of 155 MPH, with gusts to 175 MPH. Floyd may end up being the first Category Six hurricane ever. It's expected to hit Florida Wednesday, and one of the possible paths once it comes inland is aimed right at us. We're pretty far inland, but Hugo was actually still technically a hurricane when it came through here several years ago. Although it had started to become disorganized, it still had an eye and sustained winds greater than 74 MPH. So I suppose we may get yet another hurricane, albeit a weak one by the time it arrives here. Still, weak is a very relative term when it comes to hurricanes. This ain't good. * * * * * And, continuing the eclectic nature of this page, this from Chuck Waggoner [waggoner at gis dot net]:
I'd not thought about that, but it seems reasonable. I remember studying saponification and the differing mechanisms of soaps and detergents, but that's one I'd not have come up with on my own. I'll CC my reply to Barbara, and I'm sure we'll do the experiment. * * * * * Okay, despite my best intentions, I find myself working on this page during the day. So here's what I've decided: I'm going to go to the Pournelle method, and update this page whenever I damned well feel like it. That means there may be days with several updates, and other times when the page may go a day or two without an update (probably mostly the former, because I'm a verbose kind of guy.) * * * * * Incidentally, I appreciate those of you who have been sending me messages in Rich Text or HTML format, but that's no longer necessary. If you find it convenient to send messages that way, it's fine to continue doing so, but if you find it easier to send me plain text, that's fine as well. During my short experience with FrontPage 2000, one of the nice things I found was the "Paste Special - Normal Paragraphs" option. This allowed me to cut or copy the text from an email message and paste it in without all the line breaks and so on. I was just getting ready to announce that I didn't need RTF/HTML messages any more when FP2K blew up on me and I was back to using FP98. At first I didn't realize that FP98 had a similar function, because there's no Paste-Special option on the right-click menu. But FP98 Editor does have Paste Special on the Edit Menu, and that's all I need. There's one less option for Paste Special in FP98 versus 2K, but the all-important Normal Paragraphs option is present. * * * * * One change I have decided to make is adding the "Jump to most recent update" link at the top of the page. I stole this idea from Bo Leuf, and it seems to make a lot of sense to me. Obviously, it's possible that more than one update will have occurred between your visits, so you may need to scroll back a bit to make sure you haven't missed anything. But I put up so much text here that it makes sense to me to have a quick way to get to the latest verbiage without scanning through all the stuff you've already read. I hope this change meets with everyone's approval. 16:00: There I was, working away and minding my own business, when I noticed motion from the corner of my eye. There was a spider the size of a Buick crawling up the wall behind my desk. Well, a very small Buick. About 1.5" (40 mm) in diameter, counting its legs. I'm a live-and-let-live kind of guy, but this thing looked fangish to me. I ran into the kitchen, grabbed the fly swatter, and arrived back in my office to find Mr. Spider staring at me. He hadn't moved. I took a spastic swing at him and he dropped unwounded onto my desk in among all the cables and other mess behind my systems and monitors. Figuring he wasn't too kindly disposed toward me at the moment, I used the fly swatter to move cables around carefully. No sign of the spider. I abandoned that plan as hopeless after a couple of minutes, and went looking for some Raid or something. I found a can of Hot Shot under the kitchen sink, and returned to my office to do battle. Still no sign of the spider, so I carefully sprayed a line of Hot Shot along the back edge of my desk. Within seconds, the smell became overpowering. I returned to the kitchen, reading the Hot Shot can directions as I went. The important parts said: "It is a violation of Federal Law to use this product other than as directed" and "DO NOT EVEN THINK OF USING THIS PRODUCT INDOORS" (or words to that effect.) So I'm now a federal felon with an unusable office. I shut the door to my office and am typing this on Barbara's system under FrontPage 2000. Ugh. * * * * * This from bdenman [bdenman@ftc-i.net]:
Yes, so far the vote has been unanimous in favor of the old way. In fact, I mailed Pournelle earlier today to suggest that he might consider switching to a single page. 26,400? You do have a bad phone line. I used to be at the far end of our local loop. The CO is about 7 miles from here as the crow flies, so there was a lot of copper between me and the CO. Even then, I got 28,800 about 80% of the time, 31,200 about 10%, 26,400 about 8%, 24,000 about 1%, and 33,600 about 1%. A year or so ago, the phone company installed a SLC about a mile from my house. Since then, I've gotten 31,200 about 99% of the time, with the remainder at 28,000. Strangely, I haven't gotten one 33,600 connect since that SLC went in. I've never felt the need for anything faster than 31,200, unless that something faster is a whole heck of a lot faster, like cable or xDSL. I actually have a 56K modem or two sitting around here that I haven't even taken the trouble to set up. I suppose a 50% improvement would be noticeable, but I do most of my major transfers in the background anyway. The web just doesn't seem much faster to me at 56K than at 33,600. * * * * * This from J.H. Ricketson [culam@neteze.com]:
Thanks. I probably won't explore that, because my readers tell me that my page load times aren't a problem. Even if I did become concerned, I'd not be likely to consider such a proprietary approach. * * * * * This from David_Blodgett@doh.state.fl.us:
Thanks for the links. Yes, I did actually consider going to a daily page format, and discarded the idea for two reasons. First, it would be a lot more work for me to create and build links for seven pages a week instead of one. Even more important, when I brought up the subject some months ago, the readers told me almost unanimously not to do it. * * * * * This from bdenman [bdenman@ftc-i.net]:
Yep. The one thing you can predict about a hurricane is that you can't predict anything about a hurricane. We could still use another few inches of rain to recover from the drought, but that's a pretty extreme way of getting rain. * * * * * 19:30: My friend John Mikol called an hour or so ago to tell me that he was watching the TV news and that they say we're going to get nailed by Floyd. We sit exactly in the middle of its projected track through North Carolina, and it doesn't really much matter where on the projected track it ends up going. The thing is bigger than North Carolina anyway. We're far enough inland that hurricanes usually just dump a bunch of rain on us, but this time the weather forecasters say that we're going to have sustained winds of 85 MPH+ starting Wednesday night or Thursday. I hope they're wrong, but I'm afraid they're not. We can expect lengthy power outages as well, so I may be off the air for a couple of days or more once I shut down my computers Wednesday night. It could well be much longer if we get nailed badly. The last time we had a storm this bad, we lost power for the better part of a week. I have a lot to do tomorrow to prepare, so there may not be much new posted here tomorrow.
|
|
|
Tuesday, 14 September 1999 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] The latest local forecast says we can expect 65 MPH (105 KPH) sustained winds. That's not good, but it's a lot better than the 85 MPH (137 KPH) sustained winds they were talking about last night. I'll be spending quite a bit of time today getting everything movable indoors, roping down stuff that can't be brought indoors, and getting emergency supplies organized. I wish we had storm shutters. We actually do have shutters, and they're very nice indeed. Strongly built from thick wood. The only problem is that instead of being hinged, they're screwed directly into the brick facing. A lot of good that does. I've never been able to figure out why people do things like that. If you're going to have shutters, they should work. I suppose they wanted the appearance of having shutters, but the cost of installing them with hinges was considered to be too high, so they installed real shutters that don't work. Geez. I didn't used to worry much about storms, but that was before we got nailed ten years ago. Part of our roof was ripped off, and one of the columns that supports the pediment over the front porch was torn loose and came in through the library window. The insurance adjuster agreed to pay for the roof, but balked at paying for the damage to the front porch. He claimed that it might have been "gradual damage", which wasn't covered by homeowner's insurance. Right. Our column gradually collapsed and fell through the window. They did eventually pay, though. What was ironic was that we already had a quote from a contractor friend of mine for about $1,500 to repair the damage to the front porch. The adjuster said that was way too high, and insisted that we get two quotes from contractors on the insurance company's approved list. He said they'd pay the lower of the two. We got those two quotes, the lower of which was for $3,000. At that point, the adjuster came back to us and said they'd decided to pay the original estimate that my friend had given me. I called him, and he said to tell them that that quote had expired, that he had accepted other work in the mean time, and that he was too busy to do the work now. Heh. Sure enough, the insurance company wrote us a check for the lower of the estimates from their approved contractors. We accepted the check, had my friend do the work, and pocketed the difference. I'll probably see if I can contact Barbara tonight. I suspect that we have keys to her parents' and sister's houses somewhere around here, but I have no idea where. Assuming that this storm really does arrive, I'd like to be able to check out their houses after it departs. They live on the other side of town, probably 15 miles from here, and there are likely to be trees down all over the place, but I should be able to get there in one of our 4X4s. Whether I'd be able to do anything or not is another question. * * * * * And amidst all this, I still have a new workstation to build for myself. I've accumulated most of what I need, and it now sits on the kitchen table, where it is likely to remain until the storm passes through. I'd planned to build it while Barbara was away, but that's not going to happen.
I still have a few minor items to buy. Such things as a 3.5" floppy drive with 5.25" chassis (the PC Power & Cooling full tower case has no 3.5" bays); a SCSI cable or two; a keyboard; a mouse; and so on. I have 256 MB of RAM already set aside for this system. I was considering bumping that up to 384 MB or even 512 MB, but the recent skyrocketing price of RAM put paid to that idea. Oh, well, the price of RAM should start dropping again before too much longer, and it's easy enough to add RAM later. I'm debating what to do about speakers. I have a perfectly good set of $40 computer speakers sitting on the shelf, but I'm considering doing something more than that. What I'm thinking about is buying an inexpensive home audio receiver (say, a $139 Sony) and a decent set of inexpensive home audio speakers (I note that Crutchfield is selling a pair of $300 retail AR video-shielded bookshelf speakers for $129 on special). I've never been able to figure out why people buy those $250 computer speaker sets. Even with a subwoofer, the sound is nowhere near as good as what you can get from traditional home audio components for about the same price. Well, enough for now. I'd better go start battening down the hatches. * * * * * This from Tom Syroid [tsyroid@home.com]:
Yes, the link to "most recent" has been staring me in the face on Bo's site for months now. I don't know why I never thought about doing that before. Pizza would have been nice. I had peanut butter and jelly sandwiches for dinner last night. Instead of eating them over the sink, however, I dined in civilized fashion by using a plate. Barbara is actually due back on Sunday between 5:00 p.m. and 7:00 p.m. I've already got a reminder in Outlook, but my computers may not be running on Sunday, depending on what happens with Floyd. * * * * * This from Robert Rudzki [rasterho@pacbell.net]:
Yes, yes, I know. But my desk is a 3-0 door with all kinds of equipment sitting on it. I couldn't reach the spider with a water glass unless I had literally stood on the desk. As I said, I'm a L&LL kind of guy, although I think that can be taken too far. A woman my wife knows recently reported that she arrived home to find that her husband had found a copperhead in their yard. There are numerous pets and small children in the area. Instead of getting the snake's attention with a shovel, as I would have done, her husband captured it, took it to some woods not more than a few yards from their back yard, and released it. Now, granted, a copperhead isn't really a major threat to an adult, but that guy is going to feel terrible if that snake returns to the area and ends up chomping a kid or even a pet. * * * * * This from Bo Leuf [bo@leuf.com]:
Well, it's not like I'm a Natural Born Killer. I actually like snakes, although Barbara feels differently. I generally do catch and release critters, unless there's some likelihood that they may be rabid or something. But I'm not religious about it. If I ever found a rat, I'd kill it (or feed it to a snake...) * * * * * Another from Bo Leuf [bo@leuf.net]:
I'd say there's a reasonably good chance that your domain hosting service will be down for at least a while later this week. A lot depends on the track that Floyd takes. We have a front coming in that will tend to bend Floyd's track to the north and east. If that front stalls, as it appears to be doing, Floyd's track will come right through our area. If it starts moving again, Floyd may end up tracking up the Carolina coast, which would put your hosting service at risk. Malmo is a long way from here, but Floyd is gigantic. I wouldn't be a bit surprised if you eventually see some effects from it. * * * * * This from Jan Swijsen [qjsw@oce.nl]:
Hmm. You're the first person to say anything about my site loading slowly. Of course, a major part of that is that I generally post a great deal more text than any of the people you mention. As far as splitting up the week into daily pages, I mentioned one time that I was thinking about doing that, and got pretty universal feedback that people preferred the weekly page. That's fine with me, because doing seven daily pages a week is much more work than doing one weekly page. * * * * * This follow-up from Jan Swijsen [qjsw@oce.nl]:
The extra work involved has to do with maintaining the web site structure (for the Table of Contents) and particularly with maintaining the links for Last Week (or Yesterday) and Next Week (or Tomorrow). Those have to be done manually, at least under my current system. What I do now each Sunday is copy the blank page for the following week and paste it to the page for the week after that. (I always have a blank page for the next week after the current one. That avoids having to delete all the text from the current week to make the next week's page). Once I've done that, I have to edit that new page in the HTML editor to replace all the links for things that point outside that current page. For example, next Sunday (assuming I'm still operational then), I'll copy 0913RTDN.html to lastweek.html. I'll then call up 0920RTDN.html (which is a complete page template but without any content) and copy it to 0927RTDN.html. I then have to call up that document in the HTML editor and replace all references to "last week" (which still point to 0913 instead of 0920) and to "next week" (which still point to 0927 instead of 1004). There are other similar things that need to be fixed. So, I prefer to do that only once per week instead of every day. Frankly, I was relieved when my readers told me they preferred I stick with the weekly format. There are also other subtle reasons for sticking with the weekly format. For example, the search engine I use (Thunderstone) to provide the search function on this site limits the size of sites that they'll index for free by number of pages. If I used a daily format, I'd quickly exceed that limit. There are probably other good reasons to stick with weekly format, but I can't think of them at the moment. * * * * * This from Guntis Glinavs [gglinavs@serix.com]:
Thanks. I wasn't able to access that site, probably because so many other people are attempting to hit it right now. I know that Floyd is likely to cause major disruptions in power, phone service, etc., but I wonder what effect it will have on Internet service. The last time I looked, the vast majority of the entire world's Internet traffic still went through the east coast backbone, so major disruptions here may be felt all across the planet. Even if we get through this storm without major damage, I may find myself unable to talk to remote servers. Bummer. * * * * * 13:15: This is probably the last update for today. I got all the outside stuff secured or brought indoors. Now to see if I can get some work done. Incidentally, FrontPage 98 now tells me that this page through the end of this current entry will require 67 seconds to download. * * * * * Here's an interesting thought about when Windows 2000 will ship. One of the mailing lists I subscribe to comprises almost exclusively authors who are writing Windows 2000 books. Yesterday, someone posted a message asking other readers what their best bet was as to when W2K would ship. I posted a response, saying: "June, 2000. That is, if they plan to ship Pro and Server at the same time, and expect Server to actually work (more or less)." One of the other authors on the list responded to my message, saying that, based on the responses he's gotten on his bug reports to Microsoft, he doesn't think they care if Server is usable at RTM (release to manufacturing). He says, rightly, that no sane person would even consider using W2K Server on a production system anyway, so why should Microsoft worry about fixing the gaping holes in AD and other aspects of Server when all that would really accomplish is to delay the release of WK2 Pro? According to him, after a couple of service packs have been released, Server might actually be usable, but in the mean time Microsoft can take credit for shipping the product. He thinks the Server criteria for RTM will be no bluescreens during common operations with conventional hardware. I hope he's wrong, be he may be right. I've been wondering for some time now why Microsoft keeps talking about release dates that there's no hope of making with a usable product. It seemed to me that they were just painting themselves into a corner. Based on what I've seen, a usable version of W2K Server is at least a year away. The only thing that makes me think that he might be wrong is that Microsoft simply cannot afford to ship a buggy or incomplete product. If they ship a garbage version of W2K Server, they may never recover from that. We'll see, but my strong recommendation to all of my readers is that, if a version of W2K Server does ship sooner than next summer they avoid it like the proverbial plague. * * * * * This from bdenman [bdenman@ftc-i.net]:
I'm afraid you guys are liable to get nailed worse that we are, being earlier in the path. As far as PNGs, it sounds to me as though you spoke to a complete moron. What does the server care whether it's serving GIFs or PNGs? I can't say that I've ever before heard of a phone company intentionally degrading its service. Sounds like you need a new phone company. * * * * * This from Guntis Glinavs [gglinavs@serix.com]:
Okay, thanks. I'll keep trying periodically. It sounds from your description as though that's an interesting site. And, yes, I do plan on taking a stack of 20 or so books to the basement. We can live without power, so long as we have fuel for our kerosene lanterns. * * * * * This from Bo Leuf [bo@leuf.net]:
Hmm. I didn't realize that Europe in general and Sweden in particular were much affected by hurricanes. But what you say makes sense. I suppose that people in Europe must consider US weather to be incredibly dangerous and wonder how we can live here. Each dog has its own fleas. I remember a conversation with Pournelle. He asked how I could possibly feel comfortable living in an area that was subject periodically to hurricanes and tornadoes (which we may also get from this storm.) I told him that at least one can prepare somewhat for weather. He lives in an area subject to earthquakes, which provide little or no warning. How can he be comfortable living in a state that may one day simply drop off the continent and sink into the Pacific? * * * * * This from Chuck Waggoner [waggoner at gis dot net]:
Actually, I don't think I did. So I'll CC this reply to him. If he's interested, I'll be happy to explain. * * * * * This from bdenman [bdenman@ftc-i.net]:
Well, good luck. As you already know but others may not, you're in a much more exposed position than we are. I shudder to think what it must have been like for you when Hugo arrived. It was bad enough here, still technically a hurricane, but just barely. And that was bad enough. I notice that Floyd is starting to bear more sharply northwest now, and keep hoping it'll decide to head back out into the Atlantic. But there'll probably be no such luck. * * * * * 19:30: Okay, so I lied. Here's another update. I've gotten everything outside either moved inside or tied down. While I was out with the dog a little while ago, I happened to look up at the sky. These are some very weird looking clouds. I don't know much about clouds or weather, but I assume these must be from Floyd. But surely Floyd is still too far away to be doing this? Perhaps it's just the front coming through that's giving this odd banded appearance to the clouds.
* * * * * Chuck Waggoner sent me a message earlier today that mentioned floppy drives. I wrote a three paragraph response to that message, but I somehow left out the following paragraphs, both in my first reply to him and in what I posted on this page: You're right about floppy drives, and in fact the new standards for PCs eliminate them (along with other legacy stuff like ISA slots). The only thing they're useful for is extraordinary circumstances, such as disaster recovery or flashing a BIOS, but those things to my mind are certainly worth spending the $15 or so a floppy drive costs. As far as a universal floppy drive replacement, I don't see one on the horizon, unless it's CD-R(W). CD-ROM is a replacement in one sense, because many new systems can boot from CD, but it is not writable. None of the superfloppies is a realistic replacement, for reasons of compatibility, market share, reliability, media cost, etc. I don't really see CD-R(W) becoming a ubiquitous replacement, either. It's true that the media cost has dropped and will continue to drop, but a CD-R(W) disc inherently costs more to manufacture than a floppy, unless they succeed in transitioning away from gold and even silver. But the real problem is drive cost. There's no way that volume production could ever get the cost of a CD-R(W) drive down to the $20 - $30 range, which is what it would take for them to become ubiquitous. It's true that floppy disk drives used to cost what CD-R(W) drives do now, but machines were much more expensive then, and there was no real alternative to installing one or more floppy drives. Nowadays, the need for that function has become quite limited, so my guess is that the floppy will simply go away and not be replaced by anything. * * * * * This from Bo Leuf [bo@leuf.com]:
True. De re gustibus non est disputandum. Some, perhaps most, people love the beach. I wouldn't think of going there. It has all of the dreaded 4 S's: sun, sea, sand, and sharks. * * * * * This from Robert Rudzki [rasterho@pacbell.net]:
Well, maybe I am a Natural Born Killer then, because I regard a poisonous snake in close proximity to children and pets as a candidate for the chop. I know he can't help being a poisonous snake, any more than a rat can help being a rat or a rabid skunk can help being a rabid skunk, but that's his tough luck. I would probably leave his decapitated body lying around as a warning pour les autres. * * * * * This from Svenson Sjon [sjon@svenson.com]:
Well, I can't really blame FrontPage for this problem. It does in fact automatically take care of fixing all the links internal to the page, which is all one might reasonably expect of it. It can't, after all, read my mind. In effect, I'd be asking it to be smart enough to notice that I'm doing a weekly page and automatically adjust itself to point to the new "last week" and "next week" page instead of to the existing pages that are already defined as "last week" and "next week". I can't imagine that there's any product currently available that's smart enough to do that. If there were, I'm not sure I'd want to use it. "Smart" software has a habit of being too smart for its own good, and for mine.
|
|
|
Wednesday, 15 September 1999 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] As of 9:15, we're just starting to get some very light rain. It's supposed to increase through the day, with very heavy rains and high winds beginning this evening. It's times like this that I wish our dogs were trained to use the toilet. The worst part is that they'll expect a full walk, regardless of the weather. But I'm sure not planning to walk the dogs tonight. They'll have to be satisfied with a quick out with me standing on the porch. * * * * * I woke up hungrier than usual this morning. How embarrassing. I forgot to eat dinner last night. That wouldn't be so bad, except that I don't generally eat breakfast or lunch, either. So I went all day yesterday without eating. I may have breakfast or lunch today, as well as dinner. * * * * * I called Barbara last night to find out where the key to her sister's house was, among other things. She asked after the dogs. I told her that Duncan had run away to join the circus. * * * * * Sun announced that 250,000 people had downloaded StarOffice in the first week and a half. Microsoft says they're not worried. I would be, if I were them. * * * * * More Outlook weirdness. Below are exact copies of two return-receipts I received yesterday evening. Outlook moved the first one to my Receipts folder. It left the second one in my Inbox. I have automatic processing of return receipts enabled in Outlook, but I don't have any explicit rules for processing them. Out of every 100 return-receipts I get, Outlook moves maybe one or two of them to the Receipts folder. Return-receipts arrive in a great variety of formats. Even Outlook-to-Outlook generates different types of return-receipts, depending on the format of the original message (plain text, RTF, etc.). At first, I thought that differing formats for the return receipts were causing the inconsistent processing, but that's not the case. If anyone can spot any difference between these two receipts (other than the times) I'd sure like to know about it. I hate Microsoft. Your message
was read on 9/14/1999 9:17 PM. Your message
was read on 9/14/1999 7:31 PM. * * * * * This from Robert Rudzki [rasterho@pacbell.net] concerning dangerous animals and dangerous people:
And I would treat such people just as I would treat the poisonous snake or the rabid skunk. Again, I don't really care if they can help it. The point is their very nature makes them a threat. So I'd give them the chop. * * * * * This from Bo Leuf [bo@leuf.com]:
I checked the Weather Channel when I awoke this morning, and it appears that we have a chance of avoiding the worst of it. Formerly, the center of the projected track came right through Winston-Salem. Their latest projected track puts us on the far western edge. Of course, Floyd is so huge that even if it follows the far eastern line of the projected track, we'll still probably get many inches of rain and very high winds. Our local forecast is for rain starting this morning, getting heavier throughout the day, and with high winds starting tonight. Tomorrow is supposed to be worse. I'll probably power down all my computers tonight and sleep downstairs. * * * * * This from R Hakanson [asdfsa@hotmail.com]:
Sorry. I know what causes that problem, but I have no idea how to fix it. The cause is what, for lack of a better term, I'll call "long lines." I've tried to avoid them as much as possible, for example by replacing long URLs contained in reader messages with a link to the URL instead. But I can't realistically fix the problem, because lines with embedded links also appear "long" when I view source in Notepad. I fixed one of these pages a couple of weeks ago, and it took me more than half an hour of trial and error to locate and fix the line that was causing the problem. I simply don't have time to do that routinely. If anyone has a quick way to find and repair these problems, I'd love to hear about it. * * * * * This from R Hakanson [asdfsa@hotmail.com] regarding smelly dishrags:
Yuck. I think I'll start using more Clorox *and* soap when I do the laundry... * * * * * This from Alan Donders [alan_donders@hotmail.com]:
Good point. Those of us who network our home machines tend to forget about that use for a floppy disk. And, as I said, I'm no fan of eliminating the FDD from PCs. Unfortunately, the powers that be have determined that the next generation of PCs will eliminate "legacy" items like the floppy drive, ISA slots, serial and parallel ports, etc. It will be interesting to see just how far they go in stamping them out. Not installing a floppy disk drive or serial port is one thing. Not including an FDD interface on the motherboard is a step worse. Removing BIOS support for legacy devices is likely to cause some problems for those of us who prefer to stick with the old ways. The next couple of years is going to be very interesting in PCs. * * * * * This from Gary M. Berg [Gary_Berg at ibm.net]:
I probably could, if I could write an AWK script, which I can't. But the truth is that making the weekly changeover isn't a big problem. It takes me perhaps 15 minutes to set up the new page, modify the links, add the navigation components for the TOC, etc. I don't begrudge doing that once a week for one page. But I really don't want to do it every day. And your idea about FrontPage is certainly a valid concern. When you change stuff like that outside of FrontPage, you need to do a "recalc links" to get FrontPage's internal mechanisms resynchronized. The problem is, each time you do a recalc links, it modifies many pages. They still show up with the same date/timestamp within FrontPage Explorer, but the date/timestamps in NT Explorer show that the pages have been changed. Sure enough, when you subsequently publish, every single page has to be sent up to the server. That takes a while, and is something I prefer to do as seldom as possible. I think we'll ride out the storm okay, if the latest projected track turns out to be accurate. We'll get rain starting this morning, with heavy rain and substantial winds by tonight. Tomorrow should be the worst. Looking at things on the Weather Channel, now, I'd guess we may get winds gusting up to 50 MPH or so. That's not good, and is likely to do some damage, but nothing we can't live with. I'll power down my PCs tonight, and see how things go tomorrow. Barbara was indeed smart to be a thousand miles away while this thing visits. * * * * * 12:30: The rain is starting to come down a bit harder. I'm not sure if I'll do another update today or not. Keep checking back. I'll turn off all my PCs if the weather gets really bad, and turn them on again once it passes, assuming that we still have power. Supposedly, the weather will get heavy tonight and much worse tomorrow, with the worst of it tomorrow afternoon or evening, and everything fine again by Friday. We'll see. * * * * * A couple of weeks ago, I mentioned my experiments with a field-expedient treatment for Athlete's Foot, so I thought I'd update everyone. After trying unsuccessfully to treat this problem for about six months using every commercial Athlete's Foot remedy available, I finally decided to try using Soft Scrub with Bleach, on the assumption that sodium hypochlorite kills everything it touches. And it worked. I used a small foot brush (I can tell it's a small foot brush because it's shaped like a small foot) to apply Soft Scrub with Bleach while I sat on the floor of the shower. I scrubbed very gently in a circular pattern, and allowed the stuff to work for a couple of minutes. After one treatment, the problem had just about gone away. After two, it was gone completely. I did a third treatment just to make sure, and I'll probably continue to treat once or twice a week for a couple of weeks. But the fact is, this works. I am not a physician, so try this at your own risk. Your feet may fall off for all I know. But it worked for me. * * * * * This from Chuck Waggoner [waggoner at gis dot net]:
I confess that I've always thought of Irish Setters as beautiful, loyal, friendly dogs, but not too bright. Perhaps yours was an exception. I'm sure that dogs vary just as people do. As I was working on this, my younger Border Collie, Duncan, pointed out to me that cats, which he says are too dumb to live, are commonly toilet-trained. * * * * * 16:00: The rain is actually lighter than it was this morning, and although it's a bit breezy there's no sign yet of any strong wind. We were supposed to be getting heavy rain and wind by this afternoon, so perhaps Floyd is progressing slower than expected, or perhaps its shift toward the north and east is taking us out of the worst of it. They're still saying we can expect heavy rain and strong winds tonight, so we'll see. I'll probably power down all my PCs after dinner. My main NT box takes the better part of an hour to shut down (literally), so I have to plan ahead. * * * * * Apparently, I'm not the only one who forgets to eat. This from Tom Syroid [tsyroid@home.com]:
I do eat dinner every night when Barbara is here. Some days I even eat lunch. Once in a great while, I'll have breakfast. I guess that keeping up with the dogs' needs and my mother's needs has kept me busy enough that I just forgot to eat. It's beginning to look more and more like there won't be much battening required. We're now sitting almost on the far west side of the projected track (versus dead in the middle yesterday), and not much has happened here yet. I hope it stays that way. * * * * * This from Dave Farquhar [farquhar@lcms.org]:
You may be right about USB, but I don't think they'll become ubiquitous if only because of cost. LS-120 drives could become standard, but I don't think so. The drives themselves are still too expensive, but that's a fairly minor obstacle. The real problem is the cost of disks. No floppy disk standard has become successful until the cost per diskette fell below $2, and it takes a per diskette cost well under $1 to allow a drive to become truly ubiquitous. Note that, even with IBM's strong support, 1.44 MB drives got off to a slow start because of high media cost. And those drives had the advantage of being able to use the much less expensive 720 KB disks interchangeably. Many people used 720 KB DD diskettes in their 1.44 MB drives for a long time, and many systems came with 720 KB 3.5" drives long after 1.44 MB drives were readily available, all because HD disks were too expensive. I can't see LS-120 disks getting down to $2 each, let alone $1, so I'd guess that LS-120 drives will remain a niche product.
|
|
|
Thursday, 16 September 1999 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Well, I spoke too soon. I finished work for the evening last night at about 8:00. After closing Outlook on my own PC, I decided to go back and download Barbara's mail for her. One of the messages she received was from a potential client. About 8:10, as I was sitting at her PC responding to that message to tell the person that she wouldn't be returning until next week, I heard a chirp. I was sitting in her office in the dark, so it wasn't immediately obvious to me that we'd had a power failure. A moment later I heard another chirp. I got up and switched on the overhead light. Sure enough, no power. I went downstairs and fired up the hurricane lantern to await developments. The Duke Power automated outage response system was swamped, but I did get the outage reported. My mother has one of those powered lift chairs, which she needs to get up and down. When the power failed, the lift chair happened to be in the up position, so she was stranded, unable to sit back down. I helped her over to a regular chair, and we sat in the dark for an hour or so. In the mean time, my wife and my friend John Mikol both called, and both called the main phone number, which rings only upstairs when there's no power to my phone system. Note to self: run our upstairs phone line to line 2 of the two-line phones downstairs, so I can answer calls on that line without running up and down the stairs when power fails. After the power had been out for an hour or so, I decided I needed to do something about her lift chair. I wasn't about to connect up the generator and power it on for an outage that might end within a few hours or a day, so I decided to lug a charged UPS downstairs and use it to power her lift chair. I happen to have an iEPC 2 KVA unit sitting charged in my office, so I went upstairs to lug it down. Let me tell you, carrying 80 pounds of UPS down the stairs in the dark with a flashlight in your mouth is not fun. I got the UPS downstairs and unplugged her chair from the wall outlet. As I plugged the chair into the UPS, her light came on. That was odd. It took me several seconds to realize that it wasn't the UPS powering the light somehow. The power had come back on, literally at the exact split second that I plugged the chair into the UPS. Duh. Total outage time was from about 20:08 until about 21:18. That's not bad, given the circumstances. I have nothing but admiration for the guys who go out in weather like this (or worse) to work on fallen and damaged high-voltage power lines. While the rest of us are sheltering in our basements, these guys are outside, up poles, handling wires that I wouldn't touch even on a sunny day. These guys are true heroes in every sense of the word. But the storm has passed, and all we got from it was a little rain and some gusty winds. No damage at all. I feel sorry for those folks down toward the coast, though. Some of them got 15" (37 cm) or more rain. Not what they needed, particularly following so close on the heels of Dennis. * * * * * I see on John Doucette's page that Novell has announced a CCR (Continuing Certification Requirement) for CNEs (Certified NetWare Engineers) who want to maintain their certifications. By 31 August 2000, all current CNEs will have to pass a NetWare 5 exam or their certifications will lapse. I was one of the first Enterprise CNEs in North Carolina, and (I believe) the first Master CNE. But the days when a CNE was worth something are long past, so I'm considering just letting my certifications lapse. The last time Novell had a CCR was a couple of years ago. That time, they at least had the courtesy to provide a training manual for the required test without charge, as well as a coupon for a free test. Buying the training materials and taking the test is liable to cost $300 or $400 this time, not to mention the amount of time required to prepare, drive to the testing site, and take the test. I just don't think it's worth it. Even with the troubles that Microsoft has had and continues to have with NT, Novell continues to lose market share, not only to NT, but to Linux. I think the day is not far off when Novell will cease to be a factor in PC networking, except perhaps as an infrastructure provider, as with NDS for NT. * * * * * This from J.H. Ricketson [culam@neteze.com]:
That's odd. I just typed dusenberg in the search box in the left column and it immediately found the 2/15/99 page. I found your replica address there, but I forgot to copy the URL. You can find it just by searching that page for replica. * * * * * This from Robert Rudzki [rasterho@pacbell.net]:
I confess that I think my Soft Scrub with Bleach method is grounded in science whereas I consider the snake bile thing to be witchdoctoring, but then I don't really know. As Pournelle points out, the same was thought of folk medicines that used molds to treat infections. Other examples abound. The use of certain roots in India led to the development of the reserpine-based tranquilizers. Healers have for thousands of years been using maggots to eat the proud flesh around wounds, and you'll find that that method is now used in hospitals. So who knows? * * * * * This from Bo Leuf [bo@leuf.com]:
I spent an hour or so this morning trying to solve the problem by adding a third column to the right side. I made that column one pixel wide, and then played with various settings for width of the other columns, using both percentage and absolute pixel widths. Nothing I did worked, so I've about concluded that whatever makes things scroll off the screen is not something I can fix easily. I considered changing the column layout as you suggested, but I really prefer it this way. Actually, I'm not sure what that would buy anyway. As things are now, a person with 800X600 can scroll over to put the left column off the screen to the left. The problem is, the text is still sometimes too wide to read without scrolling. Would putting the navigation column on the right solve that problem? * * * * * This from Jan Swijsen [qjsw@oce.nl]:
I don't think there's much doubt that Microsoft needs NT5 to ship both soon and stable. It looks to me as though they can't have both. Given the choice, I think they should take stable over soon. * * * * * This from Alan Donders [alan_donders@hotmail.com]:
No, that's not a normal duration. Something on my personal PC is causing NT to take forever to shut down, and I can't figure out what it is. My wife's PC, which is configured almost identically, requires only a minute or two to shut down, versus 45 minutes or so for mine. I've encountered this problem in the past on another system running Exchange Server, and I know the cause of the problem. It's a poorly written service that doesn't terminate. In other words, Cooperative Multitasking lives on in NT. NT badly needs a UNIX-like "kill, and I really mean it" command. * * * * * This from Alan Donders [alan_donders@hotmail.com]:
Yes, there are several people who maintain more-or-less daily web journals. The ones I read daily are listed in the Frequent section of my links page. * * * * * 15:45: Well, FrontPage has screwed me again. As I was editing my page this morning, I noticed that FrontPage had changed the name of 0830RTDN.html to 0830rtdn.html. It did this all by itself, you understand. It was fine last night, and renamed this morning. I'd previously learned the futility of attempting to rename simply by changing lower-case to upper-case, so I renamed 0830rtdn.html to 0830junk.html and let FrontPage save the change. I then re-renamed 0830junk.html to 0830RTDN.html and it worked properly. I then went on making my edits and adds to my current page. When I started to publish, I noticed that FrontPage had again renamed 0830RTDN.html to 0830rtdn.html. I hate Microsoft. I really hate Microsoft. Fortunately, both files, 0830rtdn.html and 0830RTDN.html now exist on my server at pair, so no links are broken. But I really, really hate Microsoft. * * * * * The Register published an article this morning, Overclocking -- just say no, that explains the downsides to overclocking a CPU, particularly the dangers of running at higher than nominal voltages as the trace widths have dropped to 0.25 microns and lower. I have always taken the position that overclocking is a sucker bet and should be avoided, but it's hard to get anyone to listen. If anything, I'm in a better position than most to experiment with overclocking. I have many Engineering Sample Intel processors around here, and ES processors are not multiplier locked. But I have never overclocked a CPU, and don't ever intend to start doing so. People who overclock, particularly if they boost the core voltage, are looking for trouble. I am reminded of a guy I used to know who handloaded his ammunition. As far as I know, the guy is still alive, but I wouldn't bet on it. Companies who manufacture components for reloading--powder, bullets, and so on--publish tables listing suggested loadings for each caliber, according to bullet weight and other variables. They always list a variety of recommended loads to achieve various bullet velocities, and they always suggest a maximum safe load, with the caveat that it should be approached with caution. The problem is that guns vary. One gun might have a slightly tighter bore than another of the same nominal caliber, and that tighter bore can cause pressures to skyrocket. Any sane handloader begins with a light load and works his way up. The fired cartridges give clues about when you're approaching the maximum safe load. Such things as raised primers (or, god forbid, punctured ones). This maniac *started* with the listed maximum load, and worked his way *up* from there. Now, granted, component manufacturers are as worried about lawsuits as any other company, and they probably build a fudge factor into their recommended maximum load. But still. The day I decided to stay completely away from him was the day I watched him loading some .44 Special rounds. The .44 Special is a very old cartridge, and is identical in dimension to the .44 Magnum, except that the case of the latter is a tiny bit longer, to prevent a .44 Magnum round from being chambered in a .44 Special revolver. The difference between the two cartridges is like night and day. The .44 Special is a rather sedate loading, developing perhaps 850 feet/s at a chamber pressure of perhaps 12,000 to 14,000 CUP (copper units of pressure). The .44 Magnum is a tiger, developing perhaps 1,400 feet/s at horrendously high chamber pressures, 60,000 CUP or more in some loadings, as I recall. This maniac's reasoning was actually pretty good. He figured that commercial .44 Special loadings are kept mild because they may be fired in very old guns. The .44 Special also originally used a balloon-head case, which is much weaker than modern solid-head cases. He figured that since he was loading modern solid-head .44 Special cases, and would fire the rounds in a modern Smith & Wesson .44 Special revolver, that he had a lot of room to play. And he was probably right. He might have been able to work his way up gradually to perhaps 1,000 or even 1,100 feet/s without much risk. But instead, he simply picked out a mid-range loading for the .44 Magnum and proceeded to load that into a .44 Special case. At that point, I walked away, after telling him that if he wanted a .44 Magnum, he should buy a .44 Magnum rather than trying to make a .44 Special into something it wasn't. I give the same advice to prospective overclockers. If you want a fast processor, buy a fast processor. Don't try to increasing core voltage to make a slow processor into something it isn't. * * * * * This from bdenman [bdenman@ftc-i.net]:
Glad to hear you dodged the bullet. Everything seemed to come together well, between Floyd's shift to the north and east and its rapid drop in intensity. If it had followed the original projected track and arrived at 155 MPH instead of 115 MPH, we'd have had substantial damage and I'm sure yours would have been catastrophic. As it is, all we got was perhaps a couple inches of rain (I brought the rain gauge indoors, so I don't really know) and a few wind gusts. We didn't even have any branches down, let alone the whole trees which I'd feared. Our power has failed a couple of more times today, but only for a second or two at a time. Now if only Gert stays away. * * * * * This from Kerry M. Liles [kerryl@allinson-ross.com]:
Hmm. That's interesting, but I sure don't know what the basis for it is. I've been buying PC Power & Cooling cases and power supplies for years, and I haven't noticed any reduction in quality over that time. The last ones I got were a Personal Mid-Tower and a Full Tower. That was a couple of months ago. I suppose it's remotely possible that PC Power & Cooling has gone down the drain quality-wise since then, but I'd bet a lot of money that they haven't. The mid-tower appears identical to the same model I bought a year or more ago, and the Full Tower appears identical to the last one I bought several years ago. If there's been a reduction in quality, I sure don't see it. As far as packaging, I'm not sure what they're talking about there either. The PPC cases are packed the same way any other case I've ever seen is packed--top and bottom Styrofoam inserts that put at least an inch or two of space between the case and the cardboard box. The only way I can see that a PPC case could be damaged in shipping is if the cardboard box itself were completely mangled or a hole was punched in it. I'm not familiar with how AOpen cases are packaged, so I can't comment on the relative quality of the methods used. I know Larry Aldridge at PC Power & Cooling, and I can say with certainty that he's not the kind of guy that would put up with cost-reducing components. In fact, every time I talk to him, he complains about one company or another doing just that. I suspect the rest of the folks at PC Power & Cooling are much the same. I have no way to know why the folks at that web site are knocking PC Power & Cooling cases. Perhaps they've decided to push AOpen cases for reasons of their own. Perhaps they've had a run of bad luck, or perhaps they did indeed get a bad case or two. No one is perfect, but I'd count on PC Power & Cooling to be as close to perfect as any company I know of. * * * * * Reading Tom Syroid's page, I came across a link to another daily journal page, this one kept by Brian Bilbrey. Actually, it's sporadic more than daily, but worth checking out nonetheless. Mr. Bilbrey maintains the site on a local Linux system. * * * * * It has been called to my attention that a statement I made earlier might be subject to an interpretation I did not intend. I said "But the days when a CNE was worth something are long past, so I'm considering just letting my certifications lapse." In case anyone took that the wrong way, I was questioning the value of the certification itself, not the value of the people who hold it. If you have a NetWare network, you definitely want the people who work on it to be CNEs. When I said "worth something" I was speaking literally. Years ago, having a CNE bought those seeking employment in the networking field a salary at least several thousand dollars a year higher than the salaries offered to those who did not have a CNE. Nowadays, there are so many people with CNEs (including so-called "paper CNEs") that the value of having a CNE is nil when it comes to salary negotiations. Unfortunately, the same is becoming true of the MCSE. Supply and demand in action, I guess. At any rate, I'll probably let my certifications lapse, simply because they're not worth the time, trouble, and expense needed to maintain them. If anyone ever requires me to have a certification (e.g. for a consulting job), I'll simply point out that I write training materials that people use to study for certifications. If that isn't good enough for them, tough.
|
|
|
Friday, 17 September 1999 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] In the past, I've frequently commented on some of Anand's more outrageous statements. When I read his editorial on memory prices, I was almost moved to compose a reply but I didn't have time to do so. Fortunately, someone else did it for me. Dean Kent's rebuttal explains what's really going on. Basically, memory manufacturer's are between a rock and a hard place with the transition from 64 Kb to 128 Kb parts. They can't sell 128 Kb parts right now, because no one wants them. Everyone wants 64 Kb parts, but the production lines have already shifted to making 128 Kb parts, and they can't be turned on a dime. No one wants to retool for 64 Kb parts, because by the time they start coming off the line, 64 Kb parts will be a drug on the market. Look for spot memory prices to remain high for the next three or four months, and then to drop rapidly again as everyone's 128 Kb parts hit a market that's finally ready to use them. * * * * * This from bdenman [bdenman@ftc-i.net]:
Yes. I've gotten a few messages from people who don't understand what all the fuss about hurricanes is. Just some rain and high winds, they say. But at the time I became concerned, Floyd was nearing Category 5 status (think of a tornado the size of Germany) and aimed straight at us. Even a Category 2 or 3 hurricane that directly impacted New York would be a disaster. A Category 5 might well take down more than a few skyscrapers. * * * * * This from Richard Sherburne Jr [ryszardsh@eatel.net]:
You're welcome. Please see my journal yesterday about overclocking. The danger is not so much running the Celeron/300 at 450 (although that in itself may indeed cut CPU life), but running it over its rated voltage. I wouldn't try for anything faster than 450 if I were you. It's just not worth the risk for the slight performance improvement. * * * * * This from bilbrey@pacbell.net:
No problem. I hope I can help drive a bit of traffic to your site. Good luck with your new page. * * * * * This from Robert Rudzki [rasterho@pacbell.net]:
You and your conventional food preferences. A friend of mine, a former Green Beret, used to invite a large group of his friends over for "Road Kill Chili". They all thought he was kidding. I knew he was serious. They don't call those guys "snake eaters" for nothing. The chili wasn't bad, either. * * * * * This from John Dougan [jdougan@acm.org]:
Thanks. I actually do have the RK installed, but my problem is that I don't know which process is causing the problem. I suppose I could use kill to kill processes one by one (asking them to halt themselves), and find out which one doesn't kill in a timely manner. What I'd really like to have is a "shutdown verbose" command for NT that would display details of the shutdown process. That would let me see what was happening, instead of just staring at a green screen.
|
|
|
Saturday, 18 September 1999 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] My mother forgot to have a couple of her prescriptions re-filled before Barbara left, so she called the doctor's office and had them phone in the prescriptions. It was up to me to go get them. When I arrived at the drive-through pickup window, some moronic woman was sitting there with her backup lights on. I stopped well behind her. Sure enough, she managed to accelerate backwards for a good two or three feet before she realized she was going in the wrong direction. How did this woman get a driver's license? When I finally arrived at the window, the woman told me that the two prescriptions totaled about $34, $24 for one and $10 for the other. I told her that my mother said they should be about $17 and $4. She said they were recorded as Regular Cash, and did my mother have insurance that paid for part. How the hell am I supposed to know? Barbara usually does this stuff. The woman assured me that they'd refund the difference if they'd charged incorrectly, so I paid the full $34 and drove home. It wasn't as though I had any choice. When I arrived home, my mother said she did have insurance that was supposed to pay part of the cost, and that that fact is on record at the drugstore. I hate incompetence, and I really hate it when someone makes his problem into my problem. * * * * * I periodically mention web site stats here, and over the months I've gotten several messages from people who are curious about what exactly they are. Each morning at 12:02 or 12:03 a.m. pair Networks drops a file with the raw data for the preceding day into the www_logs directory on my server. I download those files to my local hard drive, and use a program called Analog to massage that raw data into meaningful reports. I usually run reports once a week, on Sunday morning, and really don't pay all that much attention to them. But for those who wonder exactly what I see when I run web site stats, here's an example. Instead of running a large full report, I just used the three most recent daily log files, so this report lists activity from Wednesday, Thursday, and Friday. Incidentally, the tailing off in hits each day is pretty normal. Ordinarily, I get heavy traffic Monday, very heavy traffic Tuesday, and then gradually reduced traffic for the remainder of the week, with light traffic on Saturday, and very light traffic on Sunday. * * * * * I was standing out front last evening, talking with my neighbor about DSL and cable modems. The mosquitos were getting kind of bad. I noticed one that had landed on my arm. In line with Rudzki's advice, I decided to capture it unharmed and release it into the wild. Alas, when I slapped it to stun it into submission, it died a horrible death. Mosquitos are very delicate creatures. That got me to thinking about the relative sizes of the creatures involved. I mass about 110 Kg, or 110,000 grams. I guessed that a mosquito probably masses a couple milligrams, which means I outmass her by something on the close order of 50 million times. But I decided to check to see how much a mosquito really weighs. My first stop was at AltaVista. I used their cute plain English search by typing in "how much does a mosquito weigh, in milligrams?" (I figured "mass" might confuse it). I got a bunch of hits, many related to mosquito weight boxers. Back to the drawing board. I used the brute force method, searching for "+mosquito +weight -"mosquito weight" +milligram*" AltaVista returned a bunch of garbage. I used the same search string on Northern Light, which returned this document as hit number two. The document is even specific to North Carolina mosquitos. I can't ask for better than that. (Incidentally, an adult Aedes aegypti masses 2 to 2.5 mg). * * * * * I note that Olympus is replacing my beloved D-400 Zoom with the D-450 Zoom. Comparing the specs, there appear to be few differences between the models, so the new model is more a tweak than an overhaul. But that's fine, because the D-400 Zoom is just about perfect as it is. If you're considering buying a digital camera, I continue to recommend the D-400 Zoom, which you'll probably be able to pick up at a deep discount as the D-450 Zoom replaces it in the channel. Otherwise, take a close look at the D-450 Zoom, which I'm sure will be just as good as the D-400 in all respects and somewhat better in a few. * * * * * This from Andre I. Mel'cuk [amelcuk@polysci.umass.edu]:
Thanks for the kind words. I'm sure that Pournelle and I influence each other to some extent. I respect his opinions, and he respects mine. But in the matter of PC Power & Cooling cases, we each arrived at the same conclusion independently and long ago. All I can do is repeat my own experiences. I've been buying and using PPC cases and power supplies for what must be close to a decade, and I've never had any problems with quality, fit-and-finish, etc. I certainly don't have experience with every brand of case out there, and I'm always prepared to learn something. I periodically look at inexpensive cases. I've never been impressed with Enlight or Inwin cases--even their "high end" models--although they have a lot of market share. It goes without saying that the power supplies bundled with the cheap cases aren't anywhere near the quality of a PPC power supply, but I've also had problems with the cases themselves--sharp edges, things that don't line up properly, and so on. One inexpensive brand that I'm reasonably happy with is Antec. I have a mid-tower and a full-tower Antec case here, which I'll probably build test bed systems on. The Antec cases are built like tanks (one actually uses heavier gauge steel than the PPC case), and I've found no sharp edges or other signs of manufacturing short-cuts. I'll probably replace the power supplies with PPC units, though. The power supplies that come with the Antec cases are typical of inexpensive cases. They'll probably work okay, but I feel more comfortable with a better power supply. So what cases have you used that you regard as nicer than the PPCs? * * * * * This from Ken Scott (US- InfoSys) [KScott@us.ci.org]:
Oops. Looks like you're using an outdated links page. I found and fixed the link on last week's page a couple of days ago. It was pointing to 0913RTDN-b.html, which was the quickly aborted new page format. Try going to the home page, www.ttgnet.com, refresh it, and then click on the link for this week. That should solve the problem. * * * * * This from Andre I. Mel'cuk [amelcuk@polysci.umass.edu]:
I certainly didn't take your message as a criticism of my recommendation, and I'm sorry if my response made you think I had done. In fact, as I said, I can only report my own experiences, which have been uniformly good with PC Power & Cooling. Of course, all of my cases have come directly from PPC, and it's possible that you were sent a returned case that had been bent by the original purchaser. You point out another important factor, and that is that what's the best case for you might not be the best case for me. Different people have different priorities. I might not care about how easy it is to swap motherboards, while that might be a critical issue for you. Or vice versa. So an ideal case for one of us might be a mediocre choice for the other. And, as I've always said, I like the PPC cases, but what I really like is their power supplies. If I were operating on a very tight budget, I might buy a less expensive case, but you can be sure that I'd install a PPC power supply in it. * * * * * This from Tom Syroid [tsyroid@home.com]:
I have no idea what I used, but I think it was Yahoo maps or Mapquest. Perhaps my readers will have some suggestions. * * * * * This from Andre I. Mel'cuk [amelcuk@polysci.umass.edu]:
Not at all. We have the same problem in this industry that Consumer Reports runs into. When they test washing machines or automobiles, they test one sample of each product. As anyone who has even a passing familiarity with statistics knows, a sample of one is inadequate. This is particularly true of items, like monitors, that are subject to wide variations between samples. I suspect, and I have some data to support that suspicion, that one of the biggest differences between the expensive brands and the less expensive ones is the amount of variation between samples. You can certainly get a bad Hitachi or Sony monitor, for example, but the chances of that happening are much less than with a no-name brand. Conversely, one particular example of the no-name monitor may be superb, but the average quality will be substantially lower than competing Hitachi or Sony models, and the standard deviation will be much larger. As far as the Lovecraft Syndrome, I work very long hours and write very quickly. That's the only thing saving me so far, but something is going to give at some point, perhaps soon. I attempt to make full responses to each reader message that I post, but I may soon have to go to the Pournelle method of publishing reader mail with short replies or no reply. I hate to do that, but I understand why he has to. Boy, do I understand. * * * * *
Well, I for one wouldn't walk around in those woods without my trusty .44. Particularly since my dogs would probably be with me, and you know what dogs do: "sniff, sniff, sniff. Oh, I smell something worth following. Follow, follow, follow. Oh, my god, it's a *bear*. I'll show this old bear who's boss. Snarl, snarl, snarl. Oops, this thing is bigger than me. I'd better run back to my master for protection." I've only ever had one close encounter with dangerous wildlife. My wife and I used to camp out a lot in our younger days, back before the idiots took over all the camping areas. I was snoozing away one night when Barbara poked me and said, "There's a big, huge, monstrous dog over by our stuff." I looked out of the tent. "Dog, hell, that's a bear." I soon had my 12 gauge riot gun with rifled slugs in it pointed at that guy, but he didn't seem to be thinking about bothering us. So I just watched him, saying "Go away, Mr. Bear" in a calm voice. He snuffled around our packs for a while, but didn't find any food, so he left. All we have around here is black bear, which are pretty laid back. If it had been a grizzly, I'd have been a lot more worried. * * * * *
Yes, I've made exactly the same point on more than one occasion. He's only 17, and he works his butt off to maintain that site. I predict great things for him once he grows up. And, although I'm sure that Anand would appreciate your putting him in the same class as Tom's Hardware, he'd probably be the first to admit that his site gets 25% or less of the traffic that Tom's does. Tom claims a peak of about 1,000,000 page reads a day. Of course, that day he'd probably published one of his articles that he breaks into 30+ separate pages, so my guess is that by the ordinary definition of a page read (one article per page), he probably really averages perhaps 15,000 or 20,000 page reads a day. Even so, that's 20 times what I do, and three times what Pournelle does. Incidentally, I'm not sure why you enclosed the "Dr." in quotes. As far as I know, Tom Pabst is a real doctor (as in MD). In fact, I think he still practices. I'm not sure exactly what a crypto-Nazi is, or why you consider him one.
|
|
|
Sunday, 19 September 1999 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Now here's a duh. Shawn Wallbridge recently bought a Kodak DC200+ digital camera. I wasn't familiar with that model, so I went over to the Kodak site to look at its specs. One of those they listed was "Color: 24-bit, millions of colors". As it happened, I had one of my images from the Olympus D-400Z up at the same time, and the IrfanView title bar was showing it as 8 BPP. Now, I wondered how it was possible that Shawn's camera was capturing 24-bit images and mine only 8-bit images. I checked the Olympus web site, but the specs on the D-400Z say nothing about color depth. Thinking that perhaps there was some sort of setting on the Olympus to change the color depth it captured, I scoured the manual, finding nothing. Then I finally realized. My monitor was set to display only 256 colors (8 BPP). I changed it to display 16.7 million colors (24 BPP) and suddenly the Olympus image was 24 BPP. I guess that's really not so much of a duh after all. I was assuming that IrfanView was reporting the properties of the stored image, whereas it was actually reporting the properties of the image as it was currently being displayed. * * * * * This letter from Matt Beland [mbeland@itool.com] originally appeared on Tom Syroid's page yesterday evening, and Matt has kindly given me permission to reprint it for the purpose of rebutting it:
I want first to state in no uncertain terms that, regardless of the implication in this letter, I am not a cigar smoker. I've smoked fewer than 20 cigars in my entire life. I am a pipe smoker. By way of evidence, here's a photo of me from 1955. I'd received the pipe I'm smoking in this picture that morning as a 2nd birthday present. I wasn't allowed to have matches yet, so I had to depend on an adult when I needed a light. Back in those days, I smoked aromatic tobaccos, because my palate hadn't developed yet. As I grew more discriminating, I shifted to pure English tobaccos, probably by the time I entered first grade, and I've been smoking them ever since.
I also want to deny in the strongest possible terms that I have ever used a heavy-caliber handgun to put extra holes in a Windows 98 CD. It is true that I have considered using Office 2000 CDs as targets on the skeet range, so perhaps that's where this vicious rumor started. * * * * *
Good point. I'll try to remember to do that. It probably won't have a great deal of impact, though. Most of my traffic is current stuff. * * * * *
Sorry to hear you're having problems, but there's not much I can do about it. This page renders best when viewed at 1024 X 768 or higher. I suspect you're viewing at 800 X 600 or lower, because all of the other similar reports I've had have been from people running 800 X 600. The problem has nothing to do with Windows 3.11 or indeed with Netscape Navigator (although it has many problems of its own.) The only thing I can suggest is that you read the page on a computer that has at least 1024 X 768 resolution. * * * * *
And here I always thought of myself as rude. * * * * * Here's one of the longest messages I've ever posted...
You might want to check Palo Alto Products. I understand that they build the cases for Dell and Micron, and that those cases are now available retail. * * * * *
Sorry to hear you got nailed by Floyd. We were very fortunate. When I first posted about Floyd, it was, at 155 MPH, a Category 4 that was within 1 MPH of being a Category 5, and the center of its projected track looked to me like it bisected our kitchen and den. All of us were fortunate that it lost strength rather than gaining it. Those along the coast were unlucky that it took a track along the coast. If it had come inland as originally projected, it would probably have been worse than Hugo or Fran, both of which were catastrophic for the Carolinas. Even as it is, the folks in the eastern part of the state are suffering, and I don't just mean those actually on the coast. Floyd dumped anything up to 18" of rain. That's about a four- to six-month supply of rain in one day, and the flooding is awful. I heard on the radio yesterday that they estimated there were still more than 1,500 people sitting on their roofs or in the tops of trees awaiting rescue. As far as tool-free access, it's a $0.29 option with the PPC cases. Well, actually $0.29 times however many screws your case has. That's what they sell brass thumbscrews for. I've never much cared about tool-free access. In a corporate environment, most IT departments don't want to make it easier for users to open the PCs. In fact, my former employer considered using Security Torx screws for just that reason. In my own office, I seldom open the cases on production systems. When I do, I have a power driver handy. Test-bed systems often run with the case off, so tool-free access isn't an issue there either. The main reason I've never been a big fan of truly tool-free cases (the ones that just pop open without removing any screws at all) is that I suspect they're less rigid than traditional cases. But there are any number of truly tool-free cases available for those who prefer them.. I used to have one desktop case that was a flip-top. All you did was set the monitor aside, press down on the top, and it opened clamshell-like in two pieces. As far as the nocache meta-tag, I've not had overwhelming response to it either way. I believe that yours is the second message I've received voting against it, and I think I've gotten about the same number in favor of it. Perhaps some other people will comment. Unless I receive some more positive votes, I'll probably just dispense with it. * * * * *
* * * * * -----Original Message-----
Sure you can make a bootable CD, assuming that your ThinkPad will boot from CD. The first thing to do is run Setup on the ThinkPad and check for a setting named Boot Sequence (or something similar). Ordinarily, that'll read something like "A:, C:" If you have "CD" as an option, you can boot from a CD. Autorun has some significant limitations, but you don't need to worry about those in this case. There are complete instructions for creating a bootable CD on the Adaptec site. I don't have the exact URL handy, but searching for "bootable cd" should find several documents that describe how to do what you want to do. The one thing to keep in mind is that when you boot from the CD drive, it becomes A: That may have an impact on the drive letters you use in your batch files and so on. * * * * *
Yep. That makes sense. If you have an RW drive and RW disc, and if your ThinkPad drive will read RW discs, it might make sense to do your experimenting with the RW disc instead of wasting a lot of CD-R discs. * * * * *
Yep, bad cables are the bane of anyone who builds PCs. I keep tons of the things around, and one of the first things I do when I'm having hardware problems is swap cables. It's surprising how often what looks like a perfectly good cable turns out to be the cause. As you say, making good cables is not difficult, but it is (relatively) expensive to do right. I confess that I tend to use the cables that come with the drives by preference, but when I need to replace a set, I sometimes use the $2 no-name, plastic-bagged cables. You'd think I would have learned by now. * * * * *
Sorry for the typo. It has now been fixed.
[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] |
