Email
Robert |
Week of 4 September 2000 Friday, 05 July 2002 08:08 A (mostly) daily journal of the trials, tribulations, and random observations of Robert Bruce Thompson, a writer of computer books. |
|
wpoison
Search [tips]
|
Jump to most
recent update Monday, 4 September 2000 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Our SETI@Home group continues to crank out work units at a cumulative total of about 100 per day. Congratulations to team members Greg Lincoln and Joe Hartman, who recently passed the 100 work unit milestone, and Al Carnali, who recently passed the 50 work unit milestone. If you haven't joined our SETI effort, please do. Here's how. A month or so of short posts starts today. I have an article to write for the O'Reilly web site, due the day after tomorrow, and the new web site to get up off its knees in the next two or three weeks. So posts here will be short, and I won't have time to answer a lot of mail. There should be quite a bit of new stuff up on the new web site over the coming days and weeks, though. As I spin that site up, a lot of the technical stuff will end up being posted over there, although this site will probably continue to be updated daily. -----Original Message----- On a serious note, a. Sun do not know how to write software than runs under NT; I've experienced similar problems with SO under both NT4 and Win2K. b. On the other hand, while SO is a tad slow to start under Linux, once it's running, it's rock solid and as fast as Office 2K. c. I very much dislike SO's "desktop" metaphor as well. As I understand it, though, SO6 will modularize all the individual apps. I read this to mean the desktop window constraints will either disappear or be optional. d. So far, SO 5.2 seems to translate Word 2000 documents without error or incident. Which is a far cry from how 5.1 handled things. More next week. Yes, from reading your page, it's obvious that the Linux version of StarOffice works fine. I just wish the same could be said for the NT version. -----Original Message----- I was surprised to read about your problems with Star Office under Windows. I installed it at work last week (Win2K Pro). I was trying to see how it performed under Windows. I had no problems. I used it for a while and then went back to Office. I could have kept using it, but I was satisfied that it would work for us. We have a situation where we will need over 100 PC's for just over a month and then we are done with them. I can't really justify paying Microsoft $50000 for one month's use, so I think we will use Star Office. Hmmm. The difference between NT4 and Windows 2000, perhaps? At any rate, I just calls 'em as I sees 'em.
|
|
wpoison
Search [tips]
|
Tuesday, 5 September 2000 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] I worked all day yesterday on the article for the O'Reilly web site, with an occasional break to help Barbara do something or other with the packing and cleaning. Today is devoted to finishing the article, because they want it by tomorrow. Barbara is cleaning house today and doing more packing, so I'll probably help with that when I take a break from writing. Once I get this article polished off, I need to devote some heads-down time to getting the HardwareGuys.com web site up off its knees. I'll be starting regular updates there this coming week, so I have a lot to get done before then.
|
|
wpoison
Search [tips]
|
Wednesday, 6 September 2000 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Our SETI@Home group now has 48 members and is fast approaching 9,000 work units complete, a very respectable total. If you haven't joined our SETI effort, please do. Here's how. Is it just me, or does everyone get those begging email messages from Africa and Asia? They always tell a sad tale of family troubles, poverty, and a struggle against all the odds to get a degree in computer science or whatever, and ask me to send whatever money I can afford to help them. The stories may even be true. I got the first such message a couple of years ago, and have been getting them in increasing numbers ever since. Although I've never responded to one, I now get a couple a month and no longer even bother to read them before deleting them. I suppose that some of them may actually be legitimate pleas for help from people who have no alternative, but I suspect most are scams. And I have neither the time nor the inclination to sort them out. Barbara is off to get her hair cut and go to the grocery store. And I'm back to work on my article for the O'Reilly web site. The deadline for completion was to be today, but O'Reilly told me yesterday that the book has been delayed until early October. When I called my editor to ask what was going on, he said that was news to him. As far as he knew, the delay was from early September to 22 September, and the book was still on schedule for release on that date. But I notice that the O'Reilly Coming Attractions page now lists the book as having an October release date, so it may be true. I hope not, but we'll see.
|
|
wpoison
Search [tips]
|
Thursday, 7 September 2000 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Our SETI@Home group continues to develop nicely. We now have 50 members, and a total of nearly 9,000 work units complete. Congratulations to team member Larry See, who recently passed the 100 work unit milestone. If you haven't joined our SETI effort, please do. Here's how. Tom Syroid posts a cautionary tale about StarOffice, which apparently munged one of his documents, costing him two days' work. For all the bitching and moaning I've done about Microsoft Office, I must say that Word has never eaten one of my documents. Now, granted, it was Tom's own fault for not saving versions as he went along, which he admits. I long ago learned the wisdom of doing that, and my disk is cluttered with documents with names like PCD Processors 20000904-1.doc. But for a supposedly polished application to corrupt a data file beyond retrieval is simply unacceptable. Amazon.com is generating more furor. First they admit that they were taking payments for featuring books as recommended titles while presenting those recommendations as unbiased reviews. Then they patent stuff that should have been unpatentable. Then, recently, they send an email message to their customer base concerning the new Amazon.com privacy policy, which is basically "you don't have any privacy, and any information you provided to us in the past or provide to us in the future is ours to do with as we please." Now it turns out that Amazon has come up with a special way to treat their loyal customers. They charge them more for the same products than they charge first-time customers. It turns out that that Amazon.com cookie on your hard disk is equally convenient for you and Amazon. It lets you order with a single click. And it lets Amazon know to quote you a higher price than they would if you didn't have the cookie. Amazon, it seems, looks at loyal customers as a resource to be looted and pillaged. I haven't ordered anything from Amazon.com for a long time, and I have no plans to do so in the future. I order my books from Fatbrain, which at least seems to have some respect for its customers. The Fatbrain privacy policy isn't perfect by any means, but at least they seem not to engage in the obnoxious actions that have apparently become routine for Amazon.com. Yesterday was a typical day, or how one's been lately. Write all day. Short reading/eating break in the afternoon. More writing. Dinner. Wash up, walk dogs, come in to check email. Find message from an editor at O'Reilly saying a journalist needs a PC hardware/memory expert to quote for an article he's reading, and could I call him ASAP. Called him and talked for half an hour. The article is mainly about Rambus and will appear Friday in the Washington Post. If I'm lucky, there'll be one of my phrases quoted--probably the one I'd least like to have seen in print--and they'll spell my name right if I'm lucky. While I'm doing the interview, mail arrives from one of the agents at my literary agency, asking if I'd be interested in doing some technical articles for a web publication. The pay isn't great, $0.25 to $0.50 a word, so it'll depend on how quickly I can do them. Interestingly, short articles are often considerably harder to write than longer ones. As someone once famously observed, "I'd have made it shorter if I'd had more time..." Today, I will finish up the web article for O'Reilly. It's been growing like Topsy. I'm at around 5,000 words now (their stated minimum was 800, and I think they probably expected 1,000 to 1,500). I'll probably finish writing it today and then cut it down some. Once that's done, I'll run it past some of my colleagues for a quick-and-dirty "sanity check", as Pournelle calls it. And once I get past that, I have to get to work on the HardwareGuys.com web site. Oh, yeah, and Barbara is leaving tomorrow for a ten-day bus tour of eastern Canada with her parents. For those of you who have been sending me email, I apologize. I'm reading all of it, but I simply don't have time to respond to all of it, let alone post it, and still keep my head above water.
|
|
wpoison
Search [tips]
|
Friday, 8 September 2000 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Our SETI@Home group continues to crank out work units. Congratulations to team member JCMorales, who recently passed the 50 work unit milestone. If you haven't joined our SETI effort, please do. Here's how. We finally have a firm in-stock date for PC Hardware in a Nutshell. We finished the book in early June, and were originally hoping it'd hit the bookstores in early August. That was perhaps a bit optimistic, and we were soon told it'd be the first week of September. Then O'Reilly wasn't happy with the first iteration of the index, so they decided to have it re-done from scratch. That extended things until September 22. The first of this week, someone at O'Reilly told us that the in-stock date had slipped yet again, this time until October 10th. No one had told my editor about the latest slippage, so he looked into it for me. I'm not sure what the reason was, but he assures me that the book absolutely positively will hit the warehouse on October 10th, so that means it'll be available in bookstores not long after. We now have a link up to Fatbrain.com for anyone who wants to order the book through this site. For reasons we've made clear elsewhere, we no longer order from Amazon.com, and recommend that no one else do so either. We buy most of our books through Fatbrain.com now, and have been quite pleased with their service and prices. Full disclosure: If you order the book from Fatbrain.com via our link, they pay us a percentage of the purchase price as a commission on the transaction. That doesn't increase the price you pay, but is simply a fee that Fatbrain pays us for referring you. Any money we earn this way helps keep this site open, so if you plan to buy the book, we'd appreciate you doing so via our link.
-----Original Message----- (please use billcav@yahoo.com as my email address if you publish. thanks!) Greetings, Bob. I've enjoyed your site. The Daynotes Gang has stolen hours out of the week I would otherwise have been using for something more productive, like Diablo II. Thanks! I've gotten several very, very strange scam spams from Africa lately. They purport to be from some former government official with money that needs to be smuggled out of the country. I'm supposed to believe that somehow he received my name as someone who can help him with this. The "help" takes the form of a contribution of US dollars which will be used to facilitate the transfer of anywhere from a quarter million dollars to a hundred million, depending on the spam. I'm to receive anywhere from 10% to 25% of the total for my "help". I'm pretty tolerant. Almost all spam gets deleted without much notice, just like snail junk mail. But every once in a while I find something so offensive, evil, or insulting that I have to take action. I sent copies of the spam to the abuse mailboxes for @Home, the apparently-originating ISP, Network Solutions, and the domain that controls the originating IP. The only response I got was from @Home, which was an automated reply telling me that if the spam didn't originate from their network they didn't want to hear about it. Just another data point. Bill Cavanaugh Thanks for the kind words. I've heard about that scam, although it's not always done via the Internet. There was something in our newspaper recently about a Winston-Salem woman who fell victim to it. In the version I've heard about, it's someone from Nigeria who claims to need the details of your bank account in order to transfer a large sum of money to it, of which you will be permitted to keep a percentage. On the one hand, such scams need to be stopped. On the other, I have little sympathy for anyone who is stupid (or greedy) enough to fall for it. -----Original Message----- How come your site appears in favorites at work, with an icon that is a miniature photo of you, but not on my home machine? Win98SE at both locations. Thanks, -Ron Dunno why it acts differently at work and home. That special icon is a feature of Internet Explorer 5.0+. IE5 automatically requests the file favicon.ico. If it's present in the directory, IE uses it as an icon for the bookmark. If not, it just displays its usual icon. If you go to the Microsoft web site, choose search, and enter "favicon" as the search string, it turns up all sorts of stuff. -----Original Message----- Robert I had your new book ordered from amazon. I canceled it and ordered it from Fatbrain. I had heard about all of amazon's new policies except charging more from "loyal customers". I had been dealing with them for years. I buy books and dvd movies. I liked Fatbrain's web site. Had no problem finding your book. Mike Boyle Thanks for ordering the book. You may have gotten a deal on it. I see that Fatbrain is still listing it at $20, which they show as a 20% discount from the retail price of $25. But actually O'Reilly has set the retail price at $30, so $20 is actually a 33% discount. It'll be interesting to see if Fatbrain honors orders at the price they originally quoted. I should probably get a link up to the Fatbrain site for the book. As strange as this probably sounds to anyone who's not an author, we'll actually make more money from the referral fee from Fatbrain than we do on the royalty for each copy of the book sold. Incidentally, O'Reilly tells me that the in-stock date has slipped again, this time from 9/22 until 10/10. They do promise that this is the absolute last delay, though. -----Original Message----- . . . they send the debit and credit card transactions to Mexico to be processed. If the processor there has unscrupulous employees, your credit card numbers are sold for a premium to establish a new "YOU". How do I know this? My director now has two identities: one is her own identity and one was created by someone not her that started with an Amazon.com transaction. Apparently, the banks and credit card institutions are aware of this but, no one has gone so far as to say anything to customers unless there is a problem. At that point, the first question for credit card fraud is generally: "Have you purchased anything from Amazon.com in the past six months?" If the answer is yes, the bank and credit card institutions start to tear your accounts apart so that they can rebuild them with new information to re-establish you as the real you. So much for "privacy", eh? That's why I won't buy from Amazon.com anymore. -- That's incredible if true. I'd have assumed that the credit card transactions were processed automatically and securely within Amazon.com. -----Original Message----- Nope, they are processed somewhere in Mexico and they are not secure. It is true-- my director had to tear up all of her accounts and start over because they were able to uncover her real identity to the point of obtaining a credit card in her name! They've tried to get into her E-Trade account, her Schwab account, and her bank accounts online. The shortcoming of their system is that they still can't separate the social security number and credit card number from the shipping address (yet) so she received a couple of things in the mail that she hadn't ordered and that's what started the unraveling of the "Second Joyce". -----Original Message----- For the reader who appreciates (and misses) keyboards with F-keys on the left: http://www.cvtinc.com/kybdfeatures.htm These are the old Northgate 102 keyboards (with f-keys on the left). Best !@#$ kb in the world. Alps keyswitch can handle any amount of pounding and keeps on 'clicking'. Beware: These aren't $40 keyboards. Only the serious should apply.... I have no affiliation with this company. Thanks. We've mentioned those before, but it deserves repeating. I've never much cared which keyboard I use, although I like the Microsoft Natural Keyboard models. But for someone who likes the F-keys down the left, this keyboard is definitely a good albeit expensive choice. -----Original Message----- I admit processor envy. My main machine is a dual Pentium Pro 233 (ok, the processors are clocked a bit, but PC Power dual fan heatsinks keep the temp down) and I average 36 odd hours per unit. I have been running SETI for a while, joining the daynotes group made me realize how high my hours/unit were. When I am on the machine, seti runs about 50% load on the processors, actually about 70% on #2 and 30% on #1. It loads each up near 100% when the machine is otherwise not in use. The machine has fast drives (LVD SCSI) and plenty (256 MB) of ram and a quality Intel providence MB running NT4 sp6a. Any ideas on how to speed up? Three thoughts: First, make sure you have the SETI client set to "Data analysis always runs" in Preferences. That setting tells SETI to run all the time rather than just when the screensaver kicks in. It'll occupy 100% of your CPU whenever you're not using it (including between keystrokes when you're typing). It gets out of the way immediately whenever a higher priority process needs the CPU, and I've never been able to tell the difference performance-wise between having it running and not. The only downside is that when it's set to always run, it's always using about 15 MB of memory. That shouldn't be an issue for you, though. Second, to the best of my knowledge, SETI is not dual-processor aware. On my own dual-CPU box, SETI runs at 100% on CPU0, but at 0% on CPU1. Although I haven't tried doing it, the SETI FAQ claims that you can run a second instance of the client on the second CPU. You have to install SETI a second time in a different directory and then fire up the second copy of the program and set affinity for it to CPU1. That'd give you two copies running simultaneously. Third, set the SETI screensaver to blank the screen quickly rather than displaying the pretty SETI display, which takes a lot of CPU ticks to generate. To do that, right click on an empty area of your desktop and choose Properties. On the Screensaver page, make sure the SETI screensaver is selected and click Settings. Mark the "Go to blank screen" checkbox and set the time to something shorter than the default 30 minutes (I use 1 minute). Note that this setting does not determine how long your screensaver takes to kick in. It specifies how long the SETI display is shown before the screen goes to blank. Alternatively (if you have SETI set to run all the time), you can turn off the SETI screensaver entirely and just use blank screen.
|
|
wpoison
Search [tips]
|
Saturday, 9 September 2000 [Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Our SETI@Home group now has 51 members and will likely pass 10,000 work units complete in the next few days. (Don't pay any attention to the totals they give on the page. They're all screwed up. You have to copy and paste the table into a spreadsheet to get an accurate count). Congratulations to team member Ron Snider, who recently passed the 100 work unit milestone. If you haven't joined our SETI effort, please do. Here's how. Barbara left last night for the bus tour of Canada she's doing with her parents and sister. Actually, the tour didn't leave last night. Barbara went over to sleep at her sister's house. They all got up at oh-dark-thirty to drive over to Randleman, NC, from which the bus departs. I think Barbara said they needed to leave the house at something like 0400. Yuck. As she did last time, Barbara left me with pre-written journal entries, which I'll post each day on her diary page. So far, things are going well. The dogs settled down last night, although Malcolm continues to be a very strange puppy. He's fear-aggressive, but only with members of his pack. He's completely friendly to strangers and to people he sees infrequently, such as Hannah, the lady who comes in twice a week to help my mother. But with me and, to a lesser extent, Barbara, he's terrified. Not all the time. Sometimes he's a normal puppy. Other times, though, it seems as though he suddenly remembers that we're something to fear, and then he goes into snarly mode. This morning, as I sat on the sofa reading the newspaper, Malcolm came over to me with his normal pup grin on and started snouting my hand. As soon as I started petting him, though, his hackles went up, a growling emerged from deep in his chest, and his fangs bared. Knowing that this is just a threat display, I kept petting him, and he kept snarling and growling. So I stopped. At which point, he reached up with his paw and pulled my hand down to pet him. Which I started doing, whereupon the growly/snarly act started again. So I stopped again, and again he used his paw to pull my hand over to pet him. And again he started the growly/snarly act. Over and over again. So I stopped petting him and reached into my shirt pocket, where I keep the treats. All the dogs are well-aware of this, so Malcolm instantly dropped the growly/snarly act and sat there with his pup grin on and an expectant look on his face. I handed him the treat. He took it, switched back instantly to growly/snarly mode, and munched the treat, growling and snarling the whole time. This is one seriously confused puppy. Barbara and I have tried to figure out what has caused this behavior, although there's not always a reason, or at least not one understandable by humans. Intelligent dogs are most prone to fear aggression--they have very active imaginations--and Border Collies are so damned smart that they are more likely to become fear-aggressive than any other breed. This is our first experience with fear-aggression, though, at least with a dog that we own. We think that the time Duncan bit Malcolm's snout badly may have been the genesis of the whole thing. We're trying to condition Malcolm out of the fear-aggression, but I fear it may be permanent. He needs to learn that Barbara and I are not threats to him, but convincing him of that may be impossible. I've started picking him up and holding him in my lap, feeding him treats occasionally. So long as I don't try to pet him, he'll settle down. But as soon as I start petting, he starts snarling. And he's terrified the whole time he's snarling. I can feel him trembling. Well, enough of this. I need to get back to work. The article for the O'Reilly web site still isn't finished, because the woman at O'Reilly said that the delay in publication meant I had more time to get it knocked out. I also have lots to do on the HardwareGuys.com web site, as well as a couple of systems I need to build. Pournelle should be sending me the next draft of his column for a sanity check sometime today. And with Barbara gone, a lot of my time is taken up by mom and the dogs.
-----Original Message----- > On the other, I have little sympathy for anyone who is stupid (or greedy) enough to fall for it. I used to feel the same way until I came into contact with senior citizens who would have laughed the idea off as younger people, but who were so desperate that any chance looked good. In other words, the people with the least to lose are the most likely to lose everything. This is the most terrible thing about spam, as opposed to Real World scams -- that you can reach a much wider audience with a communication that many people don't know isn't personal, and since it's free, no assumption of a ROI is needed. There's no stationary cost, no printing cost, no postage... just a free email address and a piece of pirated mass-mail software. Good point. I've always thought that people who scam the elderly are at least as despicable as child molesters. There should be a special place in Hell for them, and the courts should expedite their trip there. -----Original Message----- Hey Robert, To continue on the keyboard theme, I have found that the old IBM or even some older Ambra keyboards are awesome. My Northgate Omnikey finally died after 9 years of service and i searched for a replacement. I found that the old IBM keyboards are usually readily available from PC recycle places for around $5-10 (Canadian no less) each. I have 4 in use right now (3 at home and i at the office) and I have about 8 in stock in case of failures. Can never have too many good old clicky keyboards around ;-). Can't wait to check out your new book. Ray -- Xprima.com Corporation rwatson@xprima.com www.xprima.com Good idea. I hadn't thought about that for years, although I had an old IBM keyboard that I used on many systems until it finally died. I've used so many keyboards that I'm pretty much a keyboard agnostic. I can switch between them with no problems, although I use Microsoft Natural Keyboards by choice. -----Original Message----- I wish I were able to use 'just any' kb. After a little more than 20 years, my hands refuse to learn new positions (PC's aren't the only machines out there, you know). However, I must say that the thing that impresses me most about these kb's is it's construction. It's real hard to kill them. Spill anything you like; drop anything you like; it keeps in ticking. And I love that feedback from an Alps keyswitch. I don't know how you can even use the MS kb; feels like marshmallows to me<g>... Cost? Yea, they do cost a bundle. But, they last forever and why cheap out on your hands and eyes. I'm also a Viewsonic fan (whore?). I run only 21" screens here (with kvm's). I currently use a P815 on my daily machine, but I have most models around here - even an old PS21 that's not green! SideNote: I was at the TechData trade show here in Florida a few weeks ago, and wow, there are some interesting lcd screens coming down the pike. Viewsonic had a 32" lcd on display that was, well, stunning. Cost was stunning too, but that's changing quickly. Who knows, in a year or two we might be able to afford 21" lcd's....cool... BTW: I google'd your site. I was looking for an answer to an odd problem I'm having. I've got this windows app I wrote eons ago using Borland C 16bit code. It's ran great from 3.1, 95, 98 and NT4. On win2k (sp1) you'll notice that if you hover the mouse over the minimize|maximize|close buttons (in the upper right corner of most windows), a tooltip or balloon type help msg pop's up (mine is in yellow) that says "Close" etc... Well, if you do this with my 16bit Borland app, it nukes WOW32.DLL. You don't know of a registry setting that will turn off that hover crap, do you? Yes, my keyboard agnosticism probably puts me in a small minority, particularly among fast touch-typists. As far as the DLL, no idea. Perhaps one of my readers will know. -----Original Message----- Robert I see #50 (no name) completed his first unit in 57 minutes. That must be some computer!!! Mike Boyle Yep, it would be if the number were accurate, but it isn't. I've noticed this happen a couple of times now when someone first joins and completes one unit. The lowest real number I've seen is 48 minutes per unit, and that's from some government agency that presumably is running the client on some kind of supercomputer. -----Original Message----- I was surfing the net for some help on fixing an issue I encounter to no avail. I have recently installed a new video card (graphic accelerator type) on my Gateway computer AND NOW MY DVD IS NOT WORKING. I was so angry that I totally removed the program (impact) from the computer. I think I went a bit too far. I know that have to disable the integrated chip set. When I called Gateway they acted like I was speaking Greek. All they can say is please call a third party at $2.99 a min. (I would say that they have some kind of scam going on! ) Can you help me or do you know of a web site that can help me? my spec: 333-pentium mmx 2 , 128 sdram,window98,intergrated board I assume you don't mean that your DVD drive has died entirely. If so, you probably disconnected a cable while you were installing the new video card, or perhaps there's some sort of resource conflict. If you mean that your DVD drive still works, but you can't play movies on it, the most likely problem is that you had DVD player software that functions only with the video adapter originally in the machine. If that's the case, you should be able to find DVD player software that'll work with your current video card. I'd start looking on www.dvdutils.com. -----Original Message----- Hi Robert, I found your site with a Google search. The web page I am looking at is a year old (dated November 99.) On your site, you had some good things to say about the Onstream Echo software for the DI-30. I am wondering if you still use the Echo software, and if you have any usage tips. I am running a simple home network of two computers. One is a Pentium MMX 200 MHz. The other is a Pentium III 600 MHz. I have always been disappointed that Onstream didn't include a simple IDE hardware driver with their product. It has become even more frustrating now that we've switched from Windows 98 to Windows 2000 pro. With the upgrade, I decided to take advantage of the multi-user features of Windows 2000. The problem that I am having is that Echo software shuts down when a user logs off. Onstream has recently provided an updated Echo (3.2) for Windows 2000 pro, but it still doesn't seem to take advantage of the multi-user or login environment of Windows 2000. On your site, you advocate Echo software as a potentially good backup solution for Windows NT. I'm curious how you would implement this. Do you know of a way that Echo backups can be run in a logged off state. Would you dedicate a single non-logon based computer (Win98 or NT) to do all backups? Just curious if your opinions have changed on this subject. Perhaps you have found ways to make the software do exactly what you want it to do. Have you ever done a complete restore using the Echo software? How did it go? I use Echo enough to know that it works, but we store all data on a network server which also happens to be my wife's main workstation. That system gets backed up with a Tecmar Travan NS20 tape drive, which backs up both network data and her workstation data. On my own main workstation, I use a Tecmar 3900 DDS3 tape drive. As far as Echo under Windows NT, the (early) version I tried had problems with it, but several subsequent releases improved the situation. I've not tried running Echo under NT4 lately. Also, I've not used Echo (or any other tape software) in a multiple logon environment. I suspect the best and perhaps only way to do that would be to run Echo while logged on as an administrator on the local machine. I have done a full restore with Echo. That's part of my testing of any tape drive and software. The full restore worked properly, or as properly as a full restore can ever work on a Windows-based machine. Unfortunately, because of the way Windows generates short directory names on the fly and then has the registry point to resources using those dynamically-generated short directory names, there is literally no way to back up a Windows-based system and then be certain that it can be restored to a functional state. That's not Echo's fault. That's a design flaw in Windows itself. -----Original Message----- Hi Robert, Looking forward to the release of your book - I've read the preview chapter from O'Reilly's website and am very impressed! The hardwareguys.com site looks good too! I'm looking at replacing my Sony 48X IDE drive with a Plextor or Toshiba SCSI drive (haven't decided which yet, but the Plextor specs do look good). I plan to do some DAE in order to create MP3s and I also want a drive that is more accurate at reading damaged disks. My friend reckons SCSI CD-ROM drives are a waste of money and that I should just use my Sony CD-RW drive for this, but I'd really rather only use the drive for the task intended - writing CDs. In your opinion, is it worth replacing the drive with a SCSI unit? The other advantages I can see are less vibration than the cheaper IDE models and lower CPU utilisation. If so, which would be better? I can get a Toshiba 40X SCSI-2 for around US$87 or a Plextor Ultra SCSI 40X for around $122 in Hong Kong, where I live. I'm willing to upgrade the SCSI controller to UltraSCSI if the Plextor is worth the extra. Keep up the great work with your website - it's an enjoyable and educational read. Thanks for the kind words. As far as the drive, I'd go with the Plextor SCSI. Not because it's a SCSI, but because it's a Plextor. CD-ROM drives differ dramatically when it comes to DAE. Some IDE units don't support DAE at all, and most that do support it at only a small fraction of their rated speed. I have a 32X Toshiba, for example, which rips at only about 6X. And that's actually pretty good for an IDE drive. Some 24X to 40X IDE drives rip at only 1X or 2X. But beyond ripping speed, there's another aspect to DAE. Quality of extracted audio. Very few IDE drives (none I can think of off the top of my head other than the Plextor IDE CD burners) do clean DAE. Every one I've tried produces some artifacts rather than doing a perfect extraction. Some people won't even notice these, particularly if the WAV file is destined to become a low- or moderate bitrate MP3, but there really is a difference. Every Plextor drive I know of does clean DAE, and does it at or near its rated speed. So might the Toshiba SCSI CD-ROM, but I haven't tried it. -----Original Message----- Well some substantial improvement: I had "data analysis always runs" as the option already. I was misinformed that SETI was dual processor aware, that info was news to me but I think you are correct. I set the processor affinity to run it on the second processor. That reduced me to about 26-27 hours per unit. Then I gave it more priority over that processor and set the screen saver to blank as you suggested. That got me to about 24 hours per unit. Might be the best I can do with a 233 mhz processor. Thanks for the help. As far as I know, you've done everything possible. My Pentium II/300's do something like 14 hours per unit, but they have 512 KB of L2 cache. If time per unit were linear, I'd expect my Pentium III/600 to do a unit in about 7 hours, but in fact it takes close to 9 hours. I suspect that's because the Pentium III/600 has only 256 KB of L2 cache. The SETI client seems sensitive to cache size more than cache speed.
|
|
wpoison
Search [tips]
|
[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] Today is Barbara's and my 17th wedding anniversary. Our SETI@Home group now has 51 members and has now passed 9,500 work units complete. Congratulations to team members Wakeolda, who recently passed the 100 work unit milestone, and Jon Barrett, who recently passed the 50 work unit milestone. If you haven't joined our SETI effort, please do. Here's how. Malcolm had another growly day yesterday. Here's a sequence of photographs taken over a minute or so yesterday evening as he approaches me for some attention. I ignored him while he was sitting on the Ottoman, so he climbed down, came over next to me, and climbed up with his front paws on my leg. I started petting him, and he started snarling (photo 3). I stopped petting him, and he reached over with his paw to pull my hand back toward him. I started petting him again, and he started snarling again (photo 4). And more of the same this morning as I tried to read the newspaper. This is one seriously confused puppy.
-----Original Message----- Robert, Thanks for your thoughts. I am about a week away from getting my MCSE certification. Backup and restoration of a Windows environment has always frustrated me - now more than ever. I have never been able to do a successful full restore with any product. Now, I guess I know why. I will be getting Norton Ghost pretty soon. I am hoping that this might provide a solution that other backup methods haven't. As far as Echo goes, right now it is on the disabled list. I will have to keep playing with it to find satisfactory operation settings. Thanks again for your advice. Well, don't hold it against Echo. Other than a cloning product like Ghost or DriveImage, I don't know of any way to restore a Windows hard disk to its original state. Cloning works fine if you need to dupe a bunch of systems, but it's certainly not ideal for backing up. -----Original Message----- Hi, Robert. >From your page today: "Thanks for the kind words. As far as the drive, I'd go with the Plextor SCSI. Not because it's a SCSI, but because it's a Plextor. CD-ROM drives differ dramatically when it comes to DAE. Some IDE units don't support DAE at all, and most that do support it at only a small fraction of their rated speed. I have a 32X Toshiba, for example, which rips at only about 6X. And that's actually pretty good for an IDE drive. Some 24X to 40X IDE drives rip at only 1X or 2X. But beyond ripping speed, there's another aspect to DAE. Quality of extracted audio. Very few IDE drives (none I can think of off the top of my head other than the Plextor IDE CD burners) do clean DAE. Every one I've tried produces some artifacts rather than doing a perfect extraction. Some people won't even notice these, particularly if the WAV file is destined to become a low- or moderate bitrate MP3, but there really is a difference. Every Plextor drive I know of does clean DAE, and does it at or near its rated speed. So might the Toshiba SCSI CD-ROM, but I haven't tried it." The best way of making sure you have an exact copy of an audio CD is to use a cardware program called Exact Audio Copy, available [here] or [here]. If you set it up properly, you are guaranteed of an exact bit by bit copy of the audio files on the CD, even if your drive won't to perfect DAE on its own. It does rip slower than some other programs (I get around 2x on an HP7200i on most disks) whereas Adaptec will get about 6x. But the copy made with EAC is audibly indistinguishable from the original CD, where as a copy made from Adaptec sounds just ever so slightly off. Well, not exactly. I read the web page, and the author doesn't guarantee an exact bit-by-bit copy unless you're using a CD-ROM drive that is capable of that. Although I'm prepared to believe that this is probably a pretty good ripper, you can't do in software what requires hardware support, as the author himself makes clear. The "Exact" in the product name is an exaggeration, in other words. As the author says in billboard text "You will get the best results when using Plextor CD-ROM drives." -----Original Message----- ROBERT, In Saturday's daynotes you comment: >I have done a full restore with Echo. That's part of my testing of any tape drive and software. The full restore >worked properly, or as properly as a full restore can ever work on a Windows-based machine. Unfortunately, >because of the way Windows generates short directory names on the fly and then has the registry point to >resources using those dynamically-generated short directory names, there is literally no way to back up a >Windows-based system and then be certain that it can be restored to a functional state. That's not Echo's >fault. That's a design flaw in Windows itself. At the 'day job', we have had good success with Norton Ghost, creating a Ghost image of the drive and then backing up the image to tape. We can then restore the image and use Ghost to re-image a drive in the problem machine. This usually give us a bootable windows recovery with everything working. Sure, you can use something like Ghost or DriveImage to do an image of a partition. That's a good method if you need to clone machines or if you want a disaster recovery solution, but that's not what most people think of as backing up. If I delete a file accidentally, for example, I want to be able to retrieve just that one file without overwriting my entire boot drive. Some image backup programs do allow you to do a selective file-by-file restore, but I've never much trusted them. But I agree that an image backup product is a good disaster recovery solution. I use DriveImage periodically on my own systems for just that reason. -----Original Message----- True, Our primary objective for that method is disaster recovery, but I think the latest Ghost will let you do selective file recovery, and with todays fast machines and fast drives, you can almost ghost a partition as fast as you can back it up to tape, unless you're doing 'serious backup' with a programmed schedule of full and incremental backups. I haven't used Ghost since before Norton bought it, but I got email from a Symantec PR person a week or so back who's supposed to be sending me the latest and greatest Norton products once they're released. Their comprehensive product includes Ghost, so I'll check it out. -----Original Message----- Bob, I think you may be right about the size of L 2 cache. I run a 500 MHz Celeron with 128KB cache. It takes me about 14-18 hours to run one unit. ( I have a telephone connection, so I must start the transfer.) John D. Vogt Yep, that sounds reasonable. My Pentium III/550 (512 KB L2 cache) does a unit in 11 hours flat, so it sounds like having an L2 cache four times larger lets the entire data set fit into cache. -----Original Message----- Apropos Bill Cavenaugh's mail... "I've gotten several very, very strange scam spams from Africa lately. They purport to be from some former government official with money that needs to be smuggled out of the country." We had a lot of that hitting company and private faxes in Sweden some years ago, usually from Nigeria. Then a few cautionary tales came back from some foolish people who had gone with the bait. The stories were all pretty consistent. The mark is led along, lured to some African country (at own expense) to meet with said official or official's go-between. Meetings occur, cash is waved, papers are signed, but ultimately, the mark has to front fees or bribes to "get things moving", "prime the pump" or provide "legitimate cover transaction". The mark just sink deeper and deeper into "expenses" until the con men figure they can't milk him for any more. Then calls are not returned, meetings are cancelled, and the mark eventually realizes he's been had. Occasionally, he is threatened, jailed, and in a couple of cases it would seem even murdered. In the Nigerian cases, it seemed clear from reports that both govt and police people were involved in the scams, because official complaints and attempts to investigate got nowhere. / Bo -- Hmm. Over here it's not quite so violent. The scammers usually make contact by telephone and claim to be representing someone in Nigeria who wants to transfer a lot of money into this country, but needs a bank account to do so. They victimized a woman here in Winston-Salem in the last month or so. It made the paper, so presumably most people will be aware of it now. Or perhaps not.
[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week] |
