Can somone give the useless and ad-ridden articles at rojakpot their own section, so I can filter them all out automatically? If I wanted a graphics card review that actually gave useful information, I'd visit a site with real content in that area, like Tom's hardware.
You're thinking of the "Take a ride on the Reading" Chance card. I instead find myself humming "Going off the rails..." line from Ozzy Osbourne's "Crazy Train" when reading about this subject.
I've got about 1500 CDs to convert, some of which go back to when I started buying in 1984. In the first 50 I've done on a test run, selected as a random mix from various time periods, two of them had subtle error problems that required me to either clean the CD or get another one altogether to get a perfect read. One was flagged by the drive's C2 error detection, the other was only caught only by a comparison using the AccurateRip database (luckily a high percentage of the really old and well played CDs are in their database already). Don't know if that will remain consistant or not, but at that rate I expect 40 glitches per thousand albums converted.
As the parent's collection is even larger than mine, I'll bet his has been around the block with scars to prove it as well, and therefore I think his vigilance is fully justified. The main reason I'm just getting started with this project right now is that I've only recently become satisfied that enough databases are available for me to be confident my rips are accurate.
Hey, don't knock lesbianteenelectronics.com They may sometimes sell products as new that are in fact slightly used, but I've found that pictures already on the memory card more than make up for it.
Yeah yeah, we all were suspicious of #3 and #9. But read the quote again: "only 4% of Internet users can flag 100% of phishing e-mails as fraudulent". The only way you fail to be part of the elite 4% is if you misidentify a phishing attempt as a legitimate one, which you did not do. You, like myself and everybody else commenting here, correctly identified all the phishing attemps as such, which is the statistic they're quoting. The fact that we're so paranoid we sometimes distrust legitimate mails as well doesn't figure into that number.
Re:Tag CDs which are copy controlled
on
Amazon Goes Wiki
·
· Score: 2, Interesting
I'm sure you can add warnings all day to an title listed as the #55,186 best seller in music. I'm curious to see someone do that sort of thing with a popular title. My experience with Amazon's review features has been that I can rant about unpopular titles all day and those comments hang around, which reinforces the idea that they allow positive and negative reviews. Obviously biased comments, and ones that are negative about the topic of the book, seem to persist as well, also reinforcing that they're allowed. But try to put detailed warnings in about problems with a popular title, well those can silently go *poof*.
Details: I've written many reviews about stock trading books. I've now tried to submit a review talking about exactly what's wrong with Michael Covel's "Trend Trading", ranked #1,415 in books, three times, each time changing the text around a bit. Every time it's dissapeared quickly afterward. Someone is editing out "+5 insightful" negative comments about this book, while leaving those unlikely to dissuade a buyer alone. It's all very curious.
Oh, I get it. That book on lesbians was a gift. Wink. I went through a similarly painful period where my recommendations were embarassing after ordering several country music CDs for my relatives during the holidays.
There are two easy ways to fix this problem. If you hit the "Your Account" tab, scroll down to "Improve Your Recommendations", and there's the option to exclude purchases. You can fine-tine your entire list quickly there.
The other way is to note that every time you're given a recommendation, there's a little note saying some variant on "recommended for your because...". Just click on that for the ones that stick out as obviously out of place, and you'll get a list of titles it used; turn off the "used to make recommendations" checkbox and the book is gone.
This is a data capacity issue right now; Ubuntu is bigger than Mepis and can't squeeze everything on a single CD-R. If you download the Ubuntu DVD, it includes an expanded version of the Live CD with more apps, as well as the installer. You need a DVD drive to use it, of course, but as long as you have that hardware and can burn a DVD-R somewhere it works great.
It's is a shame that someone from the 140-160 IQ range, a group who aren't too bright, is commenting on prodigies as if they have some understanding of their problems. It must have been frustrating for the poster in the public schools, too smart to be easily taught material yet not bright enough to create their own instead. Tragic, really.
There is an excellent product available from the Bacardi corporation that is specificially aimed at reducing the social anxiety people with IQs around 151 experience. While I was technically above its target demographic, I found that with a sufficiently large dose of Bacardi 151 even my IQ was dulled enough that I could get along just fine with the common people.
Whereas the last ECS motherboard I bought was so bad, I only wish it had caught on fire; would have saved me lots of time troubleshooting problems with its BIOS.
The album has been released on their own label. Their 2000 album came out on Sire records, which is a division of Warner Bros. So this is not your standard indie artist story; this is a band who had a big contract at one point but has now given up on the majors and decided to do it themselves.
According to the RIAA, their most successful album, "Where Have All the Merrymakers Gone?", was certified a Gold record in early 1999, which means it sold at least 500,000 copies. Since it wasn't upgraded to Platinum, that means it didn't sell as many as 1M. As the single for "Flagpole Sitta" from there hit #3 on Billboard's "Modern Rock Tracks" chart, they may have sold a good number of those as well. I suspect they're planted firmly in one-hit wonder territory myself, but it's quite possible they could have a lucky hit or something and sell 750,000 copies of the new album.
Whoa, you've even gone all the way up to their 160GB model? Sheesh, with your data set, you might as well be extrapolating based on who made the best stone tablets.
Maxtor had a good run from around 500MB-100GB in storage space; many of the drives produced during that period were excellent, and much of the goodwill people have for the brand is based on that period. Somewhere around 200GB, their designs changed enough that the Maxtor drives stopped being even slightly reliable. This was also the same time period their drives were consistantly the cheapest/GB on the market and I think that's not a coincidence.
As several people have suggested, the issues Maxtor drives have seem to be a direct result of how hot the drives run; their heat dispersion techniques are just not effective for larger capacity models. Even looking at the obviously pro-Maxtor review that's the subject of this discussion, you can see how hot it and the Hitachi drive (and we all know how reliable Deathstar designs are, right?) run compared with the similar Seagate models. This particular review makes this hard to conclude, because of the size differences, but similar reviews where the capacities were matched also show the latest Seagate designs are much cooler than Maxtor or Hitachi (I'm not sure about Western Digital, because we're talking about drives that some people think are reliable here).
Let's see, now I have to throw out my anecdotal experience, right? There was a year long period not too long ago where I bought a 200GB and then two 250GB drives from Maxtor, spaced evenly throughout the year (mid-2002 to mid-2003 I think), so certainly not all from one batch. When the first one failed, I was happy because the replacement was a 250GB model so I had a matched set. But after the day where both the original 250GB drives died, even though one of them wasn't even plugged in and running at the time, it stressed even my meticulous backup methods to recover from that. Total result: my failure rate on the Maxtor drives is 133%. How did I get over 100%? Well, one of the 250GB drives that was RMA'd and replaced with a refurbished unit, which just last week has now died itself.
During that same era and since, I've bought around 8 Seagate drives, most of them 160GB models. They were a little smaller than the Maxtor units from the same era, but I haven't seen Seagate's RMA page yet.
And if we're going to swap university tech support stories, in later 2003 I worked at a school where all the freshmen had 250GB Maxtor drives similar to mine in their machines. The guy running the hardware part of their support deparment tells me he's never seen a failure rate like this in the ten years he's been doing that work; if you said "Maxtor" near him he'd jump like he'd been poked.
Anyway, regardless of which war stories you believe, the fact is that the hotter a drive runs, the more likely it is to fail quickly. If you care about reliability, you should always buy the drives that run the coolest. And that's sure not Maxtor or Hitachi right now, but instead Seagate. To even hesitate on that decision, there would need to be a compelling, recent body of evidence condemning Seagate's reliability record combined with an outstanding and also recent period where Maxtor did better, and I'm not seeing either of those right now, but rather the exact opposite on both sides.
From what I've seen, there's something driving what reviews get filtered out of Amazon that goes beyond any simple rules on their part.
I've been getting paid to write reviews of various sorts for a long time now, and I'm certainly not writing anything that would get filtered because of errors in its content, poor writing, foul language, or the like. Recently I got into reading about the stock market, and I wrote reviews for 10 different books that I had something to say about. These are a mix of the very best and the very worst of the books I read.
All but one of these reviews were untouched, including several 1 and 2 star ones where I panned the title as being basically useless or a waste of money (all books that, upon checking, I discovered don't sell well). But one of the reviews just dissapeared one day. It was a title called "Trend Following", by Michael Covel; Amazon ranks it #805 in books as I write this, which means it sells quite well indeed, and the average review gives it 4.5 stars. I said exactly what was wrong with the book. Shortly afterwards, review goes poof.
Now I'm curious. I wrote another review, with the same basic tone and an equally "+5 Insightful" commentary on why the book wasn't any good, but this time being especially careful to stick to the facts and not let my tone get negative beyond that. That one was also deleted.
The interesting part here is that there are many negative reviews listed for the book--all of which are basically useless themselves, because they either have an obvious ax to grind or they don't give enough information on the flaws of the book to really dissuade buyers. Although negative, they weren't convincing, so those are still there. After all, we wouldn't want all the reviews to be positive ones, right? That would look fake. (I'm laughing here because most of the reviews of "Trend Following" are incredibly obvious shills)
So someone here is noticing that my review is not just negative, it's genuinely likely to turn away a customer, so off it goes. Some number of reviews that are negative but not informative enough to convince are kept around just so it all looks fair. Who's driving this process? Amazon? The author? Someone who works for the publisher? I resubmitted my second version of the review again, and it was also whacked in short order, so whoever is doing it is at least consistant. But this much I am sure of: if I'm considering writing a negative review, I check the sales rank first and don't bother if it's high; just wasting my time otherwise.
Obviously Hot Russian Teens need to be at the top of the suspect list.
Re:Why are we even paying attention to this?
on
Hot Coffee Cooling Off
·
· Score: 2, Interesting
Your reaction matched my initial one. I've since come to realize I was wrong. Parents who are hoping the ESRB rating process can help them select titles for their children don't care one bit about the realities of the development process. The clarification developers are getting here is that if a title is shipped with a certain rating, all of the content on the media should meet that rating, whether it shows up in the official game or not. As you point out, this will increase the cost of doing business for companies shiping products with ESRB labels on them, but everyone on the rating side of things considers that the developer's problem. We should actually be glad that this is the title involved in setting this precedent, because the outrage here is muted by the fact that it's hard to find a parent who approved of the official GTA:SA who is then going to care about the additional dirt of Hot Coffee.
I see someone else has already posted a great example of the kind of easter egg that could really cause trouble; don't miss the story of unlocking Freak Show mode in "Happy Bunny and the Carnival Mystery" at http://games.slashdot.org/comments.pl?sid=156674&c id=13135433
I personally have used the Verizon EVDO service in Baltimore and New York. In the areas of the East Coast I've been where I don't get full EVDO speeds, you can at least connect at slow speeds over Verizon's nationwide network and do simple work. Big difference between that and a tiny little area plan like Richochet. I even get a strong EVDO signal here in the suburbs of Baltimore ~15 miles out of the center of the network where I live, beyond where DSL reaches even.
Your legal and licensing comments are of course spot on and I believe that's the intractable part of all this.
But on the subject of needing to have dedicated documentation writers, it's possible that releasing whatever crumbs you do have could actually help the company internally. Considering that people doing reverse-engineering work can end up with good documentation on how a device works even though they started with nothing, I'd suggest that if you gave them whatever shreds of data you did have, they could actually help figuring out what was missing in your documentation and fill it in. The end result would be documents audited for correctness and completeness at very low cost to the manufacturer. Why not put the open source programmer who really wants that chip to be supported to work for you to everyone's benefit?
As for the other replies in this thread suggesting that documentation should be developed first as a sort of spec for the code, that entire line of thinking is so old school now. It's now recognized that developing code to build something teaches you things that alter the design as you go. This is why methodologies like Extreme Programming and Agile Development are so popular. For developing small to medium sized projects on an aggressive schedule (which usually describes driver development), particularly if parts of the design aren't understood well enough by the people doing the development, these approaches can work very well. To quote from Brooks himself: "The discard and redesign may be done in one lump, or it may be done piece-by-piece. But all large-system experience shows that it will be done."
Starting with full documentation so complete that it serves as a spec and then implementing is absurd when you recognize the implications of this reality. I'm not saying you shouldn't design and plan, but by the time you have a product that works well, you may not have actually built it as originally envisioned, so why go so crazy writing that plan down before starting?
This is particularly true in driver development, where you can easily find yourself painted into a corner by the hard requirements of the device you're controlling, requirements that might not even be obvious when starting. Simple example for demonstration sake: let's say you're writing something that reads data from a serial port. Based on the project goals, you might do some computations and start with a polled setup at the beginning, thinking that at the data rate you're using the buffers in the device will suffice to keep up with the flow of data; using polling makes the whole design simpler. But after actually building such a thing, you find out that some low-level mechanisms actually take longer than you had thought to run, and now you're forced to use an interrupt based approach. Next thing you know, you're tossing out large pieces of code and replacing them with new one as you split your original monolithic polling handler into sections, because now you have to make the interrupt service routine tight and move the rest of the processing into another section. While in this particular example, it might seem obvious that the developer should have used interrupts from the beginning, but that's only true if you've developed something like this before. Since device manufacturers are so often treading onto new ground, there isn't so much of the benefit of experience available to guide you in cases like these.
Slashdot Mirror
Can somone give the useless and ad-ridden articles at rojakpot their own section, so I can filter them all out automatically? If I wanted a graphics card review that actually gave useful information, I'd visit a site with real content in that area, like Tom's hardware.
You're thinking of the "Take a ride on the Reading" Chance card. I instead find myself humming "Going off the rails..." line from Ozzy Osbourne's "Crazy Train" when reading about this subject.
I've got about 1500 CDs to convert, some of which go back to when I started buying in 1984. In the first 50 I've done on a test run, selected as a random mix from various time periods, two of them had subtle error problems that required me to either clean the CD or get another one altogether to get a perfect read. One was flagged by the drive's C2 error detection, the other was only caught only by a comparison using the AccurateRip database (luckily a high percentage of the really old and well played CDs are in their database already). Don't know if that will remain consistant or not, but at that rate I expect 40 glitches per thousand albums converted.
As the parent's collection is even larger than mine, I'll bet his has been around the block with scars to prove it as well, and therefore I think his vigilance is fully justified. The main reason I'm just getting started with this project right now is that I've only recently become satisfied that enough databases are available for me to be confident my rips are accurate.
Hey, don't knock lesbianteenelectronics.com They may sometimes sell products as new that are in fact slightly used, but I've found that pictures already on the memory card more than make up for it.
> Make sure the porn you include by default is under a kosher license
But what about people who aren't that fond of Jewish girls?
I'm sure all three of the comic book geeks who read Forbes are going to be very displeased with this oversight.
I love that they sell counseling for Internet Addiction via chat room. That makes about as much sense as having your AA meeting at a bar.
I wonder how much I could charge to connect people's chat clients to an Eliza bot?
Yeah yeah, we all were suspicious of #3 and #9. But read the quote again: "only 4% of Internet users can flag 100% of phishing e-mails as fraudulent". The only way you fail to be part of the elite 4% is if you misidentify a phishing attempt as a legitimate one, which you did not do. You, like myself and everybody else commenting here, correctly identified all the phishing attemps as such, which is the statistic they're quoting. The fact that we're so paranoid we sometimes distrust legitimate mails as well doesn't figure into that number.
I'm sure you can add warnings all day to an title listed as the #55,186 best seller in music. I'm curious to see someone do that sort of thing with a popular title. My experience with Amazon's review features has been that I can rant about unpopular titles all day and those comments hang around, which reinforces the idea that they allow positive and negative reviews. Obviously biased comments, and ones that are negative about the topic of the book, seem to persist as well, also reinforcing that they're allowed. But try to put detailed warnings in about problems with a popular title, well those can silently go *poof*.
Details: I've written many reviews about stock trading books. I've now tried to submit a review talking about exactly what's wrong with Michael Covel's "Trend Trading", ranked #1,415 in books, three times, each time changing the text around a bit. Every time it's dissapeared quickly afterward. Someone is editing out "+5 insightful" negative comments about this book, while leaving those unlikely to dissuade a buyer alone. It's all very curious.
Oh, I get it. That book on lesbians was a gift. Wink. I went through a similarly painful period where my recommendations were embarassing after ordering several country music CDs for my relatives during the holidays.
There are two easy ways to fix this problem. If you hit the "Your Account" tab, scroll down to "Improve Your Recommendations", and there's the option to exclude purchases. You can fine-tine your entire list quickly there.
The other way is to note that every time you're given a recommendation, there's a little note saying some variant on "recommended for your because...". Just click on that for the ones that stick out as obviously out of place, and you'll get a list of titles it used; turn off the "used to make recommendations" checkbox and the book is gone.
This is a data capacity issue right now; Ubuntu is bigger than Mepis and can't squeeze everything on a single CD-R. If you download the Ubuntu DVD, it includes an expanded version of the Live CD with more apps, as well as the installer. You need a DVD drive to use it, of course, but as long as you have that hardware and can burn a DVD-R somewhere it works great.
/ is where the DVD torrent is at.
http://us.releases.ubuntu.com/cdimage/dvd/current
It's is a shame that someone from the 140-160 IQ range, a group who aren't too bright, is commenting on prodigies as if they have some understanding of their problems. It must have been frustrating for the poster in the public schools, too smart to be easily taught material yet not bright enough to create their own instead. Tragic, really.
There is an excellent product available from the Bacardi corporation that is specificially aimed at reducing the social anxiety people with IQs around 151 experience. While I was technically above its target demographic, I found that with a sufficiently large dose of Bacardi 151 even my IQ was dulled enough that I could get along just fine with the common people.
It's obvious from the related News of the World article that Andrew is in fact a butt pirate.
Whereas the last ECS motherboard I bought was so bad, I only wish it had caught on fire; would have saved me lots of time troubleshooting problems with its BIOS.
The album has been released on their own label. Their 2000 album came out on Sire records, which is a division of Warner Bros. So this is not your standard indie artist story; this is a band who had a big contract at one point but has now given up on the majors and decided to do it themselves.
According to the RIAA, their most successful album, "Where Have All the Merrymakers Gone?", was certified a Gold record in early 1999, which means it sold at least 500,000 copies. Since it wasn't upgraded to Platinum, that means it didn't sell as many as 1M. As the single for "Flagpole Sitta" from there hit #3 on Billboard's "Modern Rock Tracks" chart, they may have sold a good number of those as well. I suspect they're planted firmly in one-hit wonder territory myself, but it's quite possible they could have a lucky hit or something and sell 750,000 copies of the new album.
Whoa, you've even gone all the way up to their 160GB model? Sheesh, with your data set, you might as well be extrapolating based on who made the best stone tablets.
Maxtor had a good run from around 500MB-100GB in storage space; many of the drives produced during that period were excellent, and much of the goodwill people have for the brand is based on that period. Somewhere around 200GB, their designs changed enough that the Maxtor drives stopped being even slightly reliable. This was also the same time period their drives were consistantly the cheapest/GB on the market and I think that's not a coincidence.
As several people have suggested, the issues Maxtor drives have seem to be a direct result of how hot the drives run; their heat dispersion techniques are just not effective for larger capacity models. Even looking at the obviously pro-Maxtor review that's the subject of this discussion, you can see how hot it and the Hitachi drive (and we all know how reliable Deathstar designs are, right?) run compared with the similar Seagate models. This particular review makes this hard to conclude, because of the size differences, but similar reviews where the capacities were matched also show the latest Seagate designs are much cooler than Maxtor or Hitachi (I'm not sure about Western Digital, because we're talking about drives that some people think are reliable here).
Let's see, now I have to throw out my anecdotal experience, right? There was a year long period not too long ago where I bought a 200GB and then two 250GB drives from Maxtor, spaced evenly throughout the year (mid-2002 to mid-2003 I think), so certainly not all from one batch. When the first one failed, I was happy because the replacement was a 250GB model so I had a matched set. But after the day where both the original 250GB drives died, even though one of them wasn't even plugged in and running at the time, it stressed even my meticulous backup methods to recover from that. Total result: my failure rate on the Maxtor drives is 133%. How did I get over 100%? Well, one of the 250GB drives that was RMA'd and replaced with a refurbished unit, which just last week has now died itself.
During that same era and since, I've bought around 8 Seagate drives, most of them 160GB models. They were a little smaller than the Maxtor units from the same era, but I haven't seen Seagate's RMA page yet.
And if we're going to swap university tech support stories, in later 2003 I worked at a school where all the freshmen had 250GB Maxtor drives similar to mine in their machines. The guy running the hardware part of their support deparment tells me he's never seen a failure rate like this in the ten years he's been doing that work; if you said "Maxtor" near him he'd jump like he'd been poked.
Anyway, regardless of which war stories you believe, the fact is that the hotter a drive runs, the more likely it is to fail quickly. If you care about reliability, you should always buy the drives that run the coolest. And that's sure not Maxtor or Hitachi right now, but instead Seagate. To even hesitate on that decision, there would need to be a compelling, recent body of evidence condemning Seagate's reliability record combined with an outstanding and also recent period where Maxtor did better, and I'm not seeing either of those right now, but rather the exact opposite on both sides.
From what I've seen, there's something driving what reviews get filtered out of Amazon that goes beyond any simple rules on their part.
I've been getting paid to write reviews of various sorts for a long time now, and I'm certainly not writing anything that would get filtered because of errors in its content, poor writing, foul language, or the like. Recently I got into reading about the stock market, and I wrote reviews for 10 different books that I had something to say about. These are a mix of the very best and the very worst of the books I read.
All but one of these reviews were untouched, including several 1 and 2 star ones where I panned the title as being basically useless or a waste of money (all books that, upon checking, I discovered don't sell well). But one of the reviews just dissapeared one day. It was a title called "Trend Following", by Michael Covel; Amazon ranks it #805 in books as I write this, which means it sells quite well indeed, and the average review gives it 4.5 stars. I said exactly what was wrong with the book. Shortly afterwards, review goes poof.
Now I'm curious. I wrote another review, with the same basic tone and an equally "+5 Insightful" commentary on why the book wasn't any good, but this time being especially careful to stick to the facts and not let my tone get negative beyond that. That one was also deleted.
The interesting part here is that there are many negative reviews listed for the book--all of which are basically useless themselves, because they either have an obvious ax to grind or they don't give enough information on the flaws of the book to really dissuade buyers. Although negative, they weren't convincing, so those are still there. After all, we wouldn't want all the reviews to be positive ones, right? That would look fake. (I'm laughing here because most of the reviews of "Trend Following" are incredibly obvious shills)
So someone here is noticing that my review is not just negative, it's genuinely likely to turn away a customer, so off it goes. Some number of reviews that are negative but not informative enough to convince are kept around just so it all looks fair. Who's driving this process? Amazon? The author? Someone who works for the publisher? I resubmitted my second version of the review again, and it was also whacked in short order, so whoever is doing it is at least consistant. But this much I am sure of: if I'm considering writing a negative review, I check the sales rank first and don't bother if it's high; just wasting my time otherwise.
While this is true, I think that we all agree that if you're an end user who has to eat a baby, Apple makes the tastiest one.
Obviously Hot Russian Teens need to be at the top of the suspect list.
Your reaction matched my initial one. I've since come to realize I was wrong. Parents who are hoping the ESRB rating process can help them select titles for their children don't care one bit about the realities of the development process. The clarification developers are getting here is that if a title is shipped with a certain rating, all of the content on the media should meet that rating, whether it shows up in the official game or not. As you point out, this will increase the cost of doing business for companies shiping products with ESRB labels on them, but everyone on the rating side of things considers that the developer's problem. We should actually be glad that this is the title involved in setting this precedent, because the outrage here is muted by the fact that it's hard to find a parent who approved of the official GTA:SA who is then going to care about the additional dirt of Hot Coffee.
c id=13135433
I see someone else has already posted a great example of the kind of easter egg that could really cause trouble; don't miss the story of unlocking Freak Show mode in "Happy Bunny and the Carnival Mystery" at
http://games.slashdot.org/comments.pl?sid=156674&
I am shocked that this story has been posted so long without someone adding a "stock charts YOU!" comment to it.
> Kids now days have the memory retention of a goldfish
This comment is insulting to goldfish. Unless you're talking about the snacks; then it's a fair comparison.
I personally have used the Verizon EVDO service in Baltimore and New York. In the areas of the East Coast I've been where I don't get full EVDO speeds, you can at least connect at slow speeds over Verizon's nationwide network and do simple work. Big difference between that and a tiny little area plan like Richochet. I even get a strong EVDO signal here in the suburbs of Baltimore ~15 miles out of the center of the network where I live, beyond where DSL reaches even.
It's obvious you have never met someone who was an Amiga fan if you think having those four will be enough to keep you safe from OS trash talking.
Your legal and licensing comments are of course spot on and I believe that's the intractable part of all this.
But on the subject of needing to have dedicated documentation writers, it's possible that releasing whatever crumbs you do have could actually help the company internally. Considering that people doing reverse-engineering work can end up with good documentation on how a device works even though they started with nothing, I'd suggest that if you gave them whatever shreds of data you did have, they could actually help figuring out what was missing in your documentation and fill it in. The end result would be documents audited for correctness and completeness at very low cost to the manufacturer. Why not put the open source programmer who really wants that chip to be supported to work for you to everyone's benefit?
As for the other replies in this thread suggesting that documentation should be developed first as a sort of spec for the code, that entire line of thinking is so old school now. It's now recognized that developing code to build something teaches you things that alter the design as you go. This is why methodologies like Extreme Programming and Agile Development are so popular. For developing small to medium sized projects on an aggressive schedule (which usually describes driver development), particularly if parts of the design aren't understood well enough by the people doing the development, these approaches can work very well. To quote from Brooks himself: "The discard and redesign may be done in one lump, or it may be done piece-by-piece. But all large-system experience shows that it will be done."
Starting with full documentation so complete that it serves as a spec and then implementing is absurd when you recognize the implications of this reality. I'm not saying you shouldn't design and plan, but by the time you have a product that works well, you may not have actually built it as originally envisioned, so why go so crazy writing that plan down before starting?
This is particularly true in driver development, where you can easily find yourself painted into a corner by the hard requirements of the device you're controlling, requirements that might not even be obvious when starting. Simple example for demonstration sake: let's say you're writing something that reads data from a serial port. Based on the project goals, you might do some computations and start with a polled setup at the beginning, thinking that at the data rate you're using the buffers in the device will suffice to keep up with the flow of data; using polling makes the whole design simpler. But after actually building such a thing, you find out that some low-level mechanisms actually take longer than you had thought to run, and now you're forced to use an interrupt based approach. Next thing you know, you're tossing out large pieces of code and replacing them with new one as you split your original monolithic polling handler into sections, because now you have to make the interrupt service routine tight and move the rest of the processing into another section. While in this particular example, it might seem obvious that the developer should have used interrupts from the beginning, but that's only true if you've developed something like this before. Since device manufacturers are so often treading onto new ground, there isn't so much of the benefit of experience available to guide you in cases like these.