Daynotes Journal Week of 13 January 2003 Latest Update : Sunday, 19 January 2003 09:18 -0500

Monday, 13 January 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

9:44 - I'm going to be doing heads-down writing until I finish the first draft of the third edition. That means posts here are likely to be short and sporadic. Today I'm working on DVD, which I hope to have knocked out by tomorrow, or Wednesday at the latest. It's a real shame that the copyright pigs have managed to tie up DVD, particularly writable DVD, to the extent they have. If it weren't for the MPAA, laws like the DMCA, and politicians like Fritz "Hollywood" Hollings, DVD would be a much more useful technology today than it is.

Instead, the copyright pigs have basically managed to make DVD illegal for Linux-based systems, or so nearly so as not to matter, while relegating DVD writers to a niche market. If it weren't for the copyright pigs, CD-RW would be a dying technology, all of us would have DVD writers on all our systems, and we'd be using $0.50 DVD blanks to make personal copies of whatever we wished.

I'm still dithering about the dueling writable DVD format issue. At first glance, DVD+R and DVD+RW seem attractive, but I think there's a good chance they'll be the losers. I say that because, although DVD+R(W) does data better than DVD-R(W) and does video better than DVD-RAM, it's inferior to DVD-R(W) for video and inferior to DVD-RAM for data. If most people had a choice and understood all the issues, they'd put DVD-RAM drives in their PCs for data and DVD-R drives for video. DVD+R(W) is based on the assumption that most people want to buy only one DVD writer, and even though DVD+R(W) isn't the best at anything it's a good compromise. The trouble is, it's an unnecessary compromise.

It'd be far better if you could buy one drive that could write DVD-RAM discs when you needed to write data, and DVD-R discs when you needed to write video. Unfortunately for DVD+R(W), you can buy such drives. So where does that leave DVD+RW? Out in the cold, as far as I can see. My impression is that the best idea for maximum flexibility is to buy a drive that can write DVD-RAM and DVD-R. Then buy some re-writable DVD-RAM discs for data, and some write-once DVD-R discs for video.

Am I wrong? Let me know over on the messageboard.

There appears to be some cause for concern about AdAware. I've been using AdAware since shortly after it was first released, but updates seem to be very slow in coming lately. I downloaded and installed Spybot Search & Destroy this morning. It's a bit different from AdAware, but it appears to do the job and scans very quickly. If you do install it, you might want to disable scanning for Usage Tracking (Settings - File Sets). I ran the program with default settings right after I installed it, and was horrified to find nearly 100 problems. Nearly all of those were Usage Tracking (e.g., WinZip keeping a list of recently used files), and so weren't problems at all.

11:56 - There's an interesting post on Bilbrey's site this morning. I find myself completely in agreement with Brian on this issue. I wouldn't ever intentionally do anything to make my site less accessible to the blind, but nor do I spend any time attempting to make it more so. It's a resource allocation issue. I have very limited time available, and choosing to spend it on one thing just means I can't spend it on something else.

I actually have thought about the accessibility issue from time to time. I concluded that my site should be reasonably accessible. It's mostly text. I use few images. Most of those, like my photograph at the top of the page and the separator bar above have descriptive filenames (photo-rbt.jpg and separator.jpg, respectively). I don't enter alternate text simply because it's one more thing to do and of very limited value to anyone, including blind visitors. Ordinarily, when I post an image I describe it in the nearby text anyway. If I have a few images that blind visitors can't access, well I'm not going to worry about it and I suspect they don't worry about it either. I am not blind, and there are a lot of images I miss out on as well. I refuse to load Shockwave and Flash, so I never see those images. I have a lot of image servers blocked, so I never see images from them, either. Missing a few images is not the end of the world, particularly on text-only pages that use text-only navigation.

As a matter of fact, one of things that I well and truly hated about Mozilla Composer was its insane insistence that I enter alternate text. Every time I embedded an image in a page, up would pop that damned dialog, insisting that I either enter alternate text or mark a checkbox to tell Composer that I didn't want to use alternate text for that image. What was missing, and I regard this as inexcusable, is an "I don't want to use alternate text for this or any other image, so never ask me again" option. I realize that the Mozilla developers were trying to do the "right thing" and force users to enter alternate text, probably both to help the blind and to ensure W3C compatibility, but I deeply resent their attempt to force me to do something I don't want to do.

That reminded me of the reason I gave up Eudora Mail and refuse to use it ever again. I wanted to send mail with return-receipts requested. Eudora Mail allowed me to request a return receipt, but I had to do so on a per-message basis. There was no option to enabled RR-requests globally, and Eudora even had a self-righteous little blurb in the on-line help that told me they wouldn't give me a global RR-request option because that wasted bandwidth. Well, screw them. I don't need people artificially limiting my choices because of their own agenda. Make the default option off for return receipts. That's fine, and accomplishes their goal. But don't refuse to allow me to do something that I want or need to do.

[Top]

Tuesday, 14 January 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

10:21 - Roland Dobbins sends this link with the subject, "Outrageous, if true". Outrageous, indeed. As the article points out, Linux folks have been expecting Microsoft to bring software patent suits to stifle Linux, but having the suit originate from SCO is a real slap in the face. It's pretty clear that Linux is unstoppable by any means other than our bloated legal system, so I fully expect Microsoft to begin serious, broad-based attacks on the software-patent front in the very near future. But to have the starting shorts fired by SCO is very disturbing.

And here's a follow-up article, and yet another follow-up article, also sent to me by Roland. From these and other articles I've read, it appears that SCO is simply asking people who have converted from SCO Unix to Linux and are still using SCO Unix libraries for compatibility with existing applications to pay a license fee for using those libraries. That's within their rights, and certainly not a threat to Linux generally if that's as far as it goes.

I sent this mailing to subscribers yesterday:

FedEx just showed up with a couple of Serial ATA Seagate Barracuda V drives. In talking with my contact at Seagate, I learned that there is a pretty big performance issue with the SIL 3112A SATA controller chip. Version 1.21 or later is recommended. Earlier versions have performance problems.

I have exactly two SATA-capable motherboards in house, an Intel D845PEBT2 and an ASUS A7N8X. I checked the SIL SATA controller chip on both, and both have rev 1.1. It's a physical chip that can't be upgraded by flashing the ROM or anything, so I'm stuck at 1.1 with these boards. Oh, well.

Fortunately, I also have a SIIG SATA controller card on the way in, so I'll have a chance to compare performance differences between 1.1 and 1.21. From what my Seagate contact told me, the 1.1 rev works, but the performance is not great. You really, really want rev 1.21 or higher on the SIL 3112A, so it's worth making a point of if you're buying a new motherboard.

I was listening to National Liberal Radio several days ago, and heard an interview with this guy. He claims, on what sounded to me very shaky evidence, that the Chinese discovered America in 1421, well before Christopher Columbus reached the New World. The interviewer didn't ask the author any hard questions, so it was difficult to judge from what was said whether his claims had any validity. Then I happened across this article on CNN. It appears that few historians are taking his claims seriously, and this excerpt from the web site may show why:

The evidence is overwhelming, and encompasses both physical entities, (such as shipwrecks of Chinese junks in America, Australia and Indonesia.) and examples such as the carved stones of Africa, the remains of Chinese peoples in South America, and artefacts scattered all over the world, inscribed with Chinese characters, in Chinese styles, and some successfully dated back to before the arrival of the Europeans.

Eh? Disregarding the fact that he can't spell artifacts, his logic is dubious at best. Say you dig up a human skeleton and find with the remains a coin dated 1421. A rational investigator notes that the person is likely to have died in 1421 or later, assuming that the coin was in fact minted in 1421. That's all one can say without further information. The person may have died in 1650, or 1913, or just yesterday. In fact, the person may have died much earlier than 1421 if there is any way that the coin could have been placed with the remains well after death, although it's a reasonable assumption in the absence of evidence to the contrary that the person was probably still walking around in 1421.

Equally, uncovering a Chinese artifact that dates to before 1492 proves nothing. Regardless of the date it was made, that artifact may have been placed where it was found any time after it was made. I have books printed more than 100 years before I was born. I've read those books, and a forensic examination would reveal my fingerprints on them. Does that prove that I read those books more than 100 years before I was born? The presence of Chinese junks also proves nothing, even if radiocarbon dating or dendrochronology establishes that the junks were made prior to 1492. What counts is not when the junk was made, or even when the wood from which it was made was cut, but when the junk sunk. That's a point that seems to have escaped this author.

This from James Chamier, which is representative of several similar messages I've received.

-------- Original Message --------
Subject: DVD debate
Date: Mon, 13 Jan 2003 18:55:11 +0000
From: James Chamier
Organization: at home with the servers...
To: Robert Bruce Thompson

Hi Bob,

I've posted on the messageboard, but I thought it may be easier to email this information instead as its getting a bit long.

I think you've actually got the emphasis wrong about the differences between the + and - DVD formats. Also DVD-RAM is an interesting media, but I can only see it being used as a random access alternative to tape by some people in the IT industry. DVD-RAM does not have enough capacity to replace backup systems in data centers, so I'll agree it's a great consumer backup product.

The big question is, what is the medium really for? I believe CD-R and RW are good enough for most home users in terms of backup. Writable DVD are sold so that people can create DVD videos that they can give to someone with a set-top-box DVD player. This is the 'big new hobby'. DVD players have sold in amazing quantities in the last 2 years, especially if you include those built in to the gaming systems PS2 and X-Box. The quality of VHS video pictures is now seen as 'bad' due to the number of people who have watched DVD movies. Modern camcorders have twice the resolution of VHS, and so the advantage of recordable DVD appears. My father has already made 3 wedding DVD's for friends, even the software now works!

The -R format is the 'official standard' from the DVD forum, -RW rewritable was invented by Pioneer and subsequently adopted by the DVD forum. While there are many combined DVD-RAM/-R drives, few other than Pioneer make -R/-RW rewritable drives.

The + format came from DVD Forum members who wanted to make a better product. It was 6 months later 'to market' which does not matter now. In the +R camp are DVD Forum members who are also some of the world's major manufacturers, including Philips, HP, Sony and Ricoh. They agreed to give Ricoh the initial hardware manufacturing contract until volumes rose.

In the UK video magazines the consensus is that there is now little to choose between -R and +R on playability, and prices of branded media have converged. It is true that unbranded -R media are cheaper, but these are reckoned to be product rejected on quality by the brand names, and apparently give more playability issues, perhaps not a medium to use for a wedding video, or data backup.

Sony have produced (and so has another brand) a dual format drive that does -R/-RW/+R/+RW. This is good news for Pioneer, but Sony are trying to avoid the marketplace problem they had with MiniDisc. They launched MD at the same time as Philips launched Digital Compact Cassette. The buying public didn't know which to buy, so neither sold! Eventually DCC died and MiniDisc is now a good secondary format used for home recordings.

Ignoring complex technical arguments about data integrity comparisons what is the user experience. -R/-RW writers showed 50% DVD players could read rewritable, over 95% could read write once, but video producers loved what they could now do with DVD, hence many commentators said -R was the 'best'. The + camp launched without write once and the rewritable +RW compatibility was the same as -RW. Within months +R was added, HP even ran a TradeIn program for their old drives.

>From the user perspective + is faster than - on both write once and rewritable for a given laser technology. Second generation -RW drive can now do 2x rewrite, but requires formatting, 1st generation +RW does 2.4x rewrite without formatting. This formatting is a significant time, taking over an hour! MPEG2 encoding is a processor intensive activity for good quality, and so DVD creation is a slow process, hence rewritable tests are essential. Hence saving the formatting before writing time is important, you have to experience this to realise how + is nicer than - to use!

An independently run website shows the differences between the technologies, the set-top-box recorder tables are very revealing.

http://www.dvdplusrw.org/

regards

James

--
James Chamier
"Thats it boys, it's a wrap!"
http://www.chamier.co.uk/james/log/

Thanks. My own feeling is that with the price of write-once DVD-R and DVD+R discs plummeting, rewritability is a non-issue for recording videos. DVD-R and DVD+R discs are fundamentally similar to CD-R discs, and there's no reason their prices shouldn't rapidly fall into the $0.50 range. Of course, there are royalty costs that I don't pretend to understand fully, but I doubt those will prevent the write-once discs from becoming commoditized just like CD-R discs. The DVD-R and DVD+R patent holders both want the huge revenue stream that will result if their products become the market standard, and with the two competing I suspect royalties of only pennies per disc will become the norm.

To my way of thinking, the ideal DVD writer is one that can write both DVD-R (for video) and DVD-RAM (for data). DVD-RW and DVD+RW are both deficient for data, not least because they don't support defect mapping, which DVD-RAM does natively. DVD+RW backers claim defect mapping support, but that's not actually a feature of the DVD+RW format. Instead, they're counting on Mt. Rainier DVD+RW drives, which aren't shipping as far as I know. Even if they were, there are no Mt. Rainier compliant operating systems.

It's going to be interesting to see what the market decides. DVD-R and DVD+R are so similar technically that there's really no need for both. The same is true of DVD-RW and DVD+RW. What's in question here isn't so much the technical aspects of the formats as the desire of the patent holders to grab the revenue streams that will result if their formats win. Rationally, any DVD-R drive should also write DVD+RW (and vice versa), and any DVD-RW drive should also write DVD+RW (and vice versa). Given that the DVD+R(W) supporters are also DVD Forum members, I suspect we'll see the DVD-R(+R) and the DVD-RW(+RW) formats merge. That leaves DVD-RAM, also backed by the DVD Forum, as the clear choice for data drives.

The only reason combo DVD-RAM/R/+R/RW/+RW haven't appeared is the additional cost of adding DVD-RAM support. That actually costs a maker $20 or so, which translates to a $75 bump in retail price. As drive volumes start increasing, that differential is likely to drop substantially. A quick look at Directron's DVD burner page (and the prices they show) is illustrative.

13:04 - More on writable DVD:

-------- Original Message --------
Subject: James Chamier's DVD Message
Date: Tue, 14 Jan 2003 11:40:04 -0600
From: Michael Hipp
To: Robert Bruce Thompson

Robert,

Mr. Chamier makes lots of good points and I thoroughly enjoyed his detailed explanation. But on one point I've got to take exception:

" DVD-RAM does not have enough capacity to replace backup systems in data centers, so I'll agree it's a great consumer backup product. "

He seems to believe that the world consists only of "consumers" and enterprises big enough to have a "data center". What he is totally ignoring is everything in-between from the SOHO to the medium-sized business that doesn't quite have anything worthy of being called a data center. These out-number the fortune 1000 by many times and all of them have gigabytes of data to backup. And tape is a good medium only for the very largest data centers that can afford the really good equipment and full-time staff.

"The rest of us" need DVD to do backups and archiving on. We've long since outgrown CD-xx as a backup medium.

Michael Hipp
Hipp & Associates

As I've said repeatedly, I don't consider any optical media a good choice for backing up (as opposed to archiving). The primary reason for that is reliability, which ranges from ranges from one to several orders of magnitude higher for tape than for hard drives, and higher still compared to optical drives. For example, a typical hard drive might generate one unrecoverable bit error per 10e14 bits. Tape drives are typically rated for one unrecoverable bit error per 10e15 to 10e19 bits. Typical optical drives have error rates around 10e12, which makes them 100 times more likely than hard drives to generate unrecoverable bit errors, and from 1,000 to 10,000,000 times more likely than tape drives.

For archiving, the higher bit error rate of optical may be acceptable. If not, the best solution is usually just to pull two copies of the data from the original source (as opposed to duping the first optical copy). For backing up current working data, the higher reliability of tape is almost always worth the cost to a business of whatever size, if they compare the cost of keeping good backups with the true potential cost of not having good backups.

For individuals and businesses operating on very tight budgets, optical may be a good choice for backup, particularly if they replicate the backups from the source. As you say, many small businesses can fit their critical data on a 4.7 GB DVD-RAM cartridge, or even on a 700 MB CD-R disc. For them, optical backup (duly replicated) may be a reasonable choice.

13:20 - Well, I said a few months ago that I thought Mandrake was in bad trouble, and a three weeks ago that I didn't expect Mandrake to be around this time next year. The Register just posted an article that says it's likely Mandrake will file for the French equivalent of Chapter 11 Bankruptcy. That's a shame, but it's not really surprising. Although Chapter 11 is for re-organization as opposed to liquidation, I don't see how Mandrake will be able to reorganize successfully. My guess is that sooner rather than later they'll end up in Chapter 7 liquidation, with what assets they have going at fire-sale prices. One of those assets is their control center software, which I hope Red Hat grabs.

[Top]

Wednesday, 15 January 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

9:43 - Yesterday, I spent some time playing with DVD-RAM under Red Hat Linux 8.0. What fun. I installed an ATAPI DVD-RAM drive as the Secondary Slave in my RH8 box. The good news is that Red Hat recognized the drive was present when I rebooted. The bad news is that it recognized it as hdd and as just another CD-ROM drive. I changed to /proc/sys/dev/cdrom and cat'd info, which tells me that Linux knows this is a DVD-RAM drive, but it's not going to access the device as a DVD drive, let alone as a DVD writer, until I do some more work on it.

So I'll be working with Linux and DVD writing today. This afternoon, Barbara and I will head for the library, pick up dinner for my mother and visit her, meet our friends for dinner at a little diner we like, and then head off to the Forsyth Astronomical Society meeting.

This from Svenson:

-------- Original Message --------
Subject: tape - CD
Date: Wed, 15 Jan 2003 09:39:29 +0100
From: Swijsen J
To: thompson@ttgnet.com

<quote> ...Tape drives are typically rated for one unrecoverable bit error per 10e15 to 10e19 bits.... </quote>

Are the error rates you mention errors occuring while writing the media or while they lay 'in storage' (occuring after writing) ?

--
Kind regards,
Jan Swijsen.

Media reliability is certainly an issue with tape, as it is for all magnetic media. That's why I recommend tape for backing up, but not for archiving. Glass-house mainframe operations do archive to tape, but they refresh those tapes every year or so. In standard PC/server environments, tape simply isn't reliable enough for archiving. Tape drive and tape manufacturers tell me that DDS tapes should be reliably readable for at least two years after they are written, but I've never seen any reason to put that to the test. I use tape for rotating backups and CD-R/DVD-RAM (replicated) for archiving.

[Top]

Thursday, 16 January 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

-------- Original Message --------
Subject: MandrakeSoft files for bankruptcy.
Date: 15 Jan 2003 13:07:50 -0800
From: Roland Dobbins
To: thompson@ttgnet.com

http://www.mandrakesoft.com/company/press/briefs?n=/mandrakesoft/news/2405\

-------- Original Message --------
Subject: Re: SuSE Linux
Date: Wed, 15 Jan 2003 21:40:09 +0000
From: Holden Aust
To: Robert Bruce Thompson <thompson@ttgnet.com>

Bob,

You make a good point. It certainly makes sense for you to get up to speed in Red Hat before trying other distributions.

My only concern is that it appears that many of your readers, of both your books and your websites, take your word as "gospel" and, if they get the impression - whether correctly or not - that you think that the only Linux to use is Red Hat, I think that would be unfortunate for several reasons.

One of the real advantages of Linux is that there is considerable competition between the various distributions of Linux and that is one of the reasons that the pace of development is so rapid in Linux.

I remember when Caldera introduced the first GUI installer, when all the other distros were using character-based installers. Several of the character-based installers worked very well, but the Caldera GUI installer caused a sensation at the LinuxWorld Expo where it was introduced and it looked much more familiar and friendlier to a user used to Windows. Within six months to a year, all the major commercial distributions had developed their own GUI installers and several of them eventually surpassed the Caldera installer that had inspired them.

Seeing that rapid recognition of the clear advantage of the GUI installer and seeing how rapidly all the other major commercial distros adopted and improved upon the concept made me realize that competition in the operating system was a major advantage that Linux would benefit from.

Red Hat 4.1 was the first "packaged" version of Linux that I tried and I've tried a number of different Red Hat versions since then. The first version of SuSE that I tried was 5.0 and I've also tried Mandrake, Caldera, Corel Linux, and several other distributions but, so far, I've always ended up preferring SuSE. I haven't tried Red Hat 8.0, so I can't make direct comparisons of Red Hat 8.0 and SuSE 8.1, but I can say that every time, in the past, that I've compared the current version of Red Hat to the current version of SuSE, I've preferred SuSE.

The LSB (Linux Standard Base - SuSE, Red Hat, and all the major commercial distributions now conform to the LSB specs, I believe) and United Linux projects are important steps toward the goal of making it eventually possible to have "shrink-wrapped" Linux application software on the shelves of a computer store. And as long as these standards make it easier to switch between different distributions, we will have the advantage of the competition between the different distributions and yet still have the ability to use standard application packages with any distribution.

So, for these reasons, I encourage you to try SuSE when you get a chance and I look forward to hearing your impressions of it.

--- Holden

[Top]

Friday, 17 January 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

-------- Original Message --------
Subject: Computer throws a hissy fit
Date: Fri, 17 Jan 2003 10:33:28 -0800 (PST)
From: David L. Griffin
To: Jerry Pournelle, Robert Bruce Thompson, Tom Syroid, Dave Farquhar

Here's an unusual failure mode.

http://www.wiu.edu/users/jrt111/hissy.html

Dave

[Top]

Saturday, 18 January 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

[Top]

Sunday, 19 January 2003

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums] [TechnoMayhem.com Forums]

[Top]

Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004 by Robert Bruce Thompson. All Rights Reserved.