If they like this system so much, I'm sure they will have no problem paying out to all the patent-holders they infringe upon, according to the same idiotic legal principles they believe should protect their own works.
Of course, if any more than a handful of crooks started following these rules, that would make the software industry impossible. Not even Microsoft could ever know what they infringe. Even if the baby jesus came down from heaven today and told them the four hundred thousand patents they infringe, they would be lost again tomorrow, when 10,000 more patents were filed.
The only way this absurd system of legalized corporate mugging is truly going to end is when Microsoft and the other lobbyists behind it themselves lose Real Money (i.e. billions of US monopoly dollars) to other patent holders.
I am wishing Implicit all the best in their bullshit lawsuit.
It's gratifying to read some comments by someone familiar with the issues.
I have a few questions:
I read the article to suggest that the plants, big machines with parts that must age over time at various rates, not always known, are decaying without an adequate regime in place to continuously inspect and replace all those components, ala the FAA and airplanes. Is there a regime in place for nuke plants sufficiently comprehensive?
You've said the article is misleading on "radiation-induced embrittlement and stress-corrosion cracking."...and, can you go into more detail on what you thought the article implied and what the realities are - do pipes carrying i.e. irradiated coolant decay at the same rates as ordinary pipes?
No comment on the "agency capture" allegations at the NRC? Is he correct about the NRC reusing text provided to it? About the ethics scandal - that really happened?
I am inured to metaphors in the news. I'm afraid Fox News, etc. has these guys beat by about 100x, so "zombie" doesn't impress me particularly. But as a story about a decaying regulatory agency, and the complexities of determining a safety regime for a set of fiercely complex 40 year old machines (which incidentally could cause a spectacular incident that would very likely end your industry, and thus your civilian career, if allowed to fail), it comes across to the layman as entirely plausible - and I do not exactly see your counterarguments yet - assuming you actually disagree with the premise, which as far as I read, was only that we need more stringent oversight of the industry (and not even that the extensions and capacity upgrades should be stopped or rolled back).
...the most salient criticism raised by the "Greenes" was that we were not, as a people, disposed to live up to the "zero tolerance" policy for failure that large scale industrial use of nuclear materials really demands. We always make mistakes eventually. Even if it takes 50 or 100 years, then it means we only have 50 or 100 years until a major nuclear disaster and i.e. epic human suffering, unprecedented economic calamity, the depopulation of a major urban area, the success of a fanatical act of terrorism, etc.
This article rather underscores the point. We have become complacent that we are smart enough and organized enough to use nuclear power safely. As we become complacent, this leads to a false sense of security, laziness and corruption on the part of operators and regulators, apathy on the part of the public, and the decline of safety culture. Now I am sure you will have no problem moving your family in down the street from one of these plants, right?
Using terms such as 'zombie', "decrepit" and 'unprecidented' without a shred of evidence
Hoping to fool people who didn't read the article?
It presents copious evidence by citing numerous specific incidents at various facilities, and clearly detailing how these incidents are related to age and lax safety culture.
Hence "decrepit."
It also discusses specific regulators at NRC, their backgrounds, and their resumes (which involve jumping between the regulatory agency and cushy jobs at the companies they regulate). It cites a specific ethics violation.
"Zombie" is perfectly valid analogy considering that these plants are unquestionably operating beyond their original design "lifetime." Quite a bit less vivid than many other terms and analogies I've been subjected to by the news media lately: i.e. who "hates America," who'se part of a "Nazi regime," who'se "socialist," who "sides with terrorists," and so forth.
The events described in the article, both in terms of safety incidents and regulatory activity, are prima facie unprecedented.
Weeks -> months. But yes, AMD are not the only offenders in the "launch" game.
Still, it takes some audacity to compare this debacle to the Wii launch, even before they stacked a november failed launch with the same silicon they couldn't ship in their september cards.
Bottom line, if you're a shareholder, all those excuses are just that... they blew it. So why say otherwisde?
Yup, that seems to be the case. But it _is_ AMD's fault that they claimed they had a "launch" in September - in fact they go right on with the fantasy and "launch" even more products they can't actually sell in November. Why not instead just be honest with senior management, and with customers and retailers, and call this what it is: "limited availability."
Of course they shipped a few thousand of them. I should have added a disclaimer at the bottom that I didn't need to hear from each one of the lucky few who managed to get one.:)
What was I expecting? To be able to buy their "launched" product, within 2-3 months after the launch, at or near MSRP.
Hey, all I'm saying is, to call this a "launch" is a joke. I think AMD would be smart to say "limited availability" and jack the price so that they are capturing ebay dollars instead of letting lucky people like you make a profit on MSRP vs. real value of the goods, and screw over themselves and their retailers.
Lots of products are launched for the holidays. If you can't do something ("launch", "MSRP") don't advertise that you can. But they can't do that, because then some pissant vendors and/or middle management would have to fess up to the fact that they blew their deadline, and they cannot capture all the dollars they could have, if for instance they actually could do what they claimed (sell product X at price Y, right now).
Oh, I have no doubt a few thousand of them have shipped overall. But these cards launched in September. They are still so far off from meeting demand that it is a joke.
Have a look around for your card today. You will find every retailer out of stock. ETAs are now running into December when they are given at all. And you will find it on ebay, for a ~25-50% premium. Most customers who want this (at anywhere near the MSRP) are still waiting 2 months after the launch, and that could turn to 3 months or more.
This is a comment on AMD's business, marketing, and PR, rather than their technical team. AMD has unquestionably won the latest round against NVidia, who will have to wait until next year (and miss the holidays) before they have a shot at retaking the top performer and price-performance crowns back.
But let's be real. The 5850 and 5870 have already "launched" too. But unfortunately AMD's idea of a "launch" is "you can buy it 4-16 weeks from now."
I see a lot of companies "making their deadlines" this way. i.e. by not actually making them. Surprised at how often the press gives them a pass on it.
Well, we've convinced ourselves that laissez faire capitalism is actually brilliant and the only reason the 19th century and the Great Depression were so non-fun was that we "just weren't doing it right." Time for round two.
You've got economists with the bullshit theory that markets are efficient and that consumers make rational, perfectly informed choices about every economic decision in their lives. This hilarious concept is actually used as cover by policy makers and regulators, to justify their actions, since it is easier to say this than to admit, "hey, I took a bribe!" The result is a "free" market where participants can be fucked over with impugnity via scams that are generally illegal in other countries, and then blamed for being victimized.
By now we have reached a level of intellectual depravity where we actually believe we can have an "efficient market" of i.e. cell phone service, land line service, or electricity - concepts which are funny, ludicrous and pathetic, in that order.
For cell phones, sure, you can make a new competitor, if you can raise several billion dollars - no problem, right? This market is so "efficient" that the 5 US participants rig prices in broad daylight. Unless you actually think the cost of sending an SMS is not only measurable but has increased 400% since ~2000. As usual, the rest of the first world (Europe, Asia) regulates these business in a more sane way and have had vastly better and cheaper cell phone service for many years. Luckily Americans have no idea what goes on in the rest of the world, so our sense of superiority need not be threatened.
Land lines are even more fun. No point in recapping our amusing attempts to apply antitrust law to The Telephone Company, only to let all the little parts practically merge back together again. Verizon's 200 billion monopoly dollars in annual revenue is put to entirely efficient uses, you can be assured. Fortunately if you don't like them, you can switch to an entirely different brand name for your Verizon land-line. At least we have the self-respect to let an unaccountable plutocrat own and profit from the infrastructure, rather than a democratic government, in Jesus's Name, I Pray.
Electricity is even more fun, since as a system it was basically designed by ex-Soviet robber barons. As Republican-patrons Enron demonstrated in federal court, the market is so "efficient" you can even create a shortage just by never building more power plants, and even turning off the ones you have. This is why privatized electricity has been so good at reducing costs in every instance it was tried. By negative dollar amounts.
You know what? I can see why privatization is so attractive. What's the point of letting government try to fix anything. Just let some lucky individuals own these natural monopolies. Not only can we not vote with our wallets, we cannot figure out how to vote, period. If the voters are too stupid to defend themselves from these scams, they don't deserve to keep their money, right? Just a bunch of children in need of some Catholic school vouchers, right?
We're hardly starting at zero developing test cases for filesystems. A reputable OS shop will have a battery of them. As the state of the art in that field progresses, new ones are required continuously, since we keep finding new and interesting ways for a FS to fail. I personally have kept seeing many novel and amusing ones over the years.:)
I neither have test cases nor bug fixes for ext4. I only say that you should have a much, much harder time finding bugs, patches, and blog posts recounting stories of woe and data loss, before you decide a filesystem is ready for Linux's largest group of low-skilled users.
They do not care about the advantages of delayed allocation or nanosecond timestamps. But when their thesis or novel or day's video shoots go poof and our answer is some blame-the-victim nonsense about backups, or reading all the fine print... who wants this on their conscience, really? Especially when it was so easily avoidable, either by not taking silly chances, or warning more loudly not to trust us.
Yes, and Reiser was nuking people's data for years. He used to claim his FS was stable before his consistency checker was even finished. XFS was/is better by comparison, and neither are anyone's default filesystem. Obviously it's the default that matters, not what other options you offer (where presumably someone making an affirmative choice knows what they're doing, or will learn the hard way fast enough).
When talking about an mp3 player, and only one person out of 100 says "it crashed" then no one gets heated. When talking about a filesystem, someone saying "But I used filesystem X for two years and never lost data," and expecting that fact to matter, is really just saying "I do not understand how to discuss filesystem reliability." A 1/100 or 1/000 shot of data loss in filesystem-land is not a finished product.
It does seem as if Fedora switched to ext4 as a default in April. Good old Redhat. I remember back in the day when they switched to PAM while it was still in the low 0.# versions, and was barely functional or documented. Many a server needed a rescue CD after one of several rpm upgrades "disturbed" the pam configs and all authentication would then fail... But the sloppiness and race to the bleeding edge was just typical for them, and among Linux users at the time, it was hardly a scandal, just yet another headache, to be expected. IMO Fedora never made a serious effort to be usable by non-experts. I used to expect better from Canonical, but apparently not anymore.
How did you figure out how long the average newbie would go before hitting the bug? And how did you estimate the time to the fix? By saying "Remember, every file system has weird bugs", I'm sure you don't imply there aren't certain widely accepted standards for stability in a filesystem? Not zero bugs (an untestable condition), but on the other hand, not "we down to a dozen patches a release, and there are only a few open data loss bugs. Well, wait, there's another."...:)
Your ability to estimate dates and quality levels is no doubt fearsome. But there is an easier way: simply shipping with an ext3 default for another 6 months (or however long it takes until the code stabilizes to the appropriate degree for a filesystem). Most users will not even notice. This is the way you behave if you have any conscience - or if moral arguments don't appeal, if you care about your reputation, or want any repeat users of your products or services.
In case you are not keeping up, Linus Torvalds himself reported losing a filesystem to ext4 in a kernel bugreport last week.
This is just not a filesystem that is ready for primetime. And that is no slight on anyone - these things take time. For a filesystem, in particular, many years. Where I do have a problem is when some mean-spirited person is out there tricking people into using it before it's stabilized, without properly communicating the risks.
Risks which, if you can make the average Ubuntu user understand them, they would almost never choose to take.
Oh, so no one can just be a first time user, use the thing for a while, put a lot of stuff on there, and get bitten when they finally cross some wacky hidden threshold that Canonical buried on their website? Right I forgot, assuming everyone has backups is just how we do things...
Your way of thinking about filesystem reliability, while interesting, isn't that useful to anyone.
Work with any file larger than 512MB and you get intermittent corruption, and you seem to imply it's not that bad?
Read the bug, for god's sake, people's apt cache gets corrupted before they even start heaven forfend installing real apps, doing development, working with media, etc...
All they had to do was just not use ext4 until it was ready. I.e. no open data loss bugs for a little while. Probably a conservative approach is warranted, it being a filesystem and all... That's _all_ they had to do.
Intel video driver issues were huge. Also, for Kubuntu users, the transition to KDE4 (which started at 8.10) was botched and caused a mass exodus from KDE and Kubuntu - some of which was down to KDE and a lot of which was down to Canonical. Linus Torvalds stopped using KDE during this period IIRC, though I don't have the link handy for his (characteristically scathing) comments. By comparison, KDE3 in 8.04 was a shockingly polished, stable system.
To this day, I see 9.04 kernel-crashing occasionally on the best supported laptops (thinkpads of various vintages) where before, crashes on that same hardware were unheard of with ubuntu/kubuntu.
Amazingly, if you click on the release notes link all the way down at the bottom of the "cool new features" page, and read about 2/3 of the way down that page, oh yeah by the way:
Possible corruption of large files with ext4 filesystem
There have been some reports of data corruption with fresh (not upgraded) ext4 file systems using the Ubuntu 9.10 kernel when writing to large files (over 512MB). The issue is under investigation, and if confirmed will be resolved in a post-release update. Users who routinely manipulate large files may want to consider using ext3 file systems until this issue is resolved. (453579)
What... the... fuck... are these morons thinking. They make ext4 their default filesystem, and release to the world with a bug like this open. I don't care whether they charge for this, give it away for free, or pay people to use it. This is not a release, it's an ugly prank. A giant fuck you to the entire world. "Hey, here's an awesome free OS! Just kidding it eats your files, lulz!!!!"
How hard is it to put a giant warning label ("MAY EAT YOUR FILES, TBD") on the front page? Why not just say, whoops the release will be late, sorry? What is it about their ego that makes it more important than peoples' data? You don't have to do work for free, just don't fucking trick people!
Between the increasing mess of 8.10 and 9.04, and this debacle, I am just losing all respect for Canonical.
Also, btw, the whole story is wrong and 9.10 is not available for download yet. Thank god.
Don't get me wrong - I think Indilinux looks to have a good future, and I am glad that they supported their users. I just can't really praise them for not finishing the product before they started selling it.
The real problem here was about "time to market" - companies were hustling to get to a lower price point before their competitors, and build their reputation in a new market. In that race, it happens again and again that corners get cut. No surprise really, and I don't think it will ever change.
As for the best bang for the buck, I find it harder to make as definitive a statement as you have. In any case, I don't think I could add anything to the (fairly nuanced) words of tomshardware or anandtech.
Not anymore, I think, but there was a period of time a little while back when the firmware updates were coming thick and fast.
Even if every firmware update works, that's not an "OK" situation for many people/companies, and not every update is guaranteed to work. Compare this to Intel drives, which have had a very low single digit number of updates over their model life on X25-M G1/G2, for instance. The firmware was "finished" before the product shipped, it had comparatively few issues (like halting or weird slowdowns, though there was one early on IIRC) and its performance has been consistent and generally better than other drives for the specs that matter most on the desktop (random reads and writes).
Thank you. The brands/models were the critical piece of information.
You're probably aware that SSD's have been in the server space, at a very different price point, for a few years now, without any extraordinary reliability debacles. To some extent, this is a case of getting what you pay for. I did a moderate amount of research on SSD drives, relying especially on the independent review sites, and quickly eliminated all of the brands you described.
As is frequent in fairly new markets, there are a few smaller and less well-run companies trying to dive in, and their first customers get to beta test their v0.* and v1.* offerings.
The prevailing wisdom seemed to me (and to people like i.e. Torvalds) that Intel was far and away the top of the heap in terms of performance and reliability, and some drives based on a newer Samsung controller (i.e. OCZ Summit) were a perhaps credible alternative. Other brands were clearly struggling to even be in the game, with frequent firmware updates and outright debacles (i.e. Indilinux, Micron) and we're in the process of shaking out who will make it and who will not.
I have only fielded a few consumer-grade SSDs over about the same amount of time as you, but going with Intel's G1 and G2 MLC products has so far yielded zero failures.
If you are already in the market for an SSD, and you are ready to spend premium money for premium performance, you should go the whole distance and go with Intel, the current market leader. See also the latest news on these models.
I agree. That said, I actually sat through the entire video, and sadly, it was a disappointment. I ignore everything that he did that came before. He has great intelligence, a wonderful idea, some vague but intriguing suggestions on how to accomplish it, and very little working code.
At least he is up front about this - if by up front, you count a litany of near-impossible problems you do not even have an idea about how to solve in the last 8 minutes of your 35 minute video.:) I do hope for the death of the fad of using videos where a perfectly good blog post will do.
In truth I think he is on to something, or at least, that problems of parallelism will be successfully attacked in ways similar to what he suggests. I can't really cast blame on anyone for being excited by an idea or building a tiny prototype that does something cool. But I would personally be way too embarrassed to say "hey guys look at this" with so little meaningful work done.
Ah, some actual points one can respond to. That uncommented link to the top of a giant chatlog... sorry tl, dr.
I suspect being able to run Windows in Bootcamp and Fusion was a major factor in Apple's switching to Intel, yes? Now only Apple computers can legally run both operating systems. And you've found some people chatting about that in IRC... OK...?
And someone griping in IRC that Miguel's article got posted while groklaw's extensive treatment on the same issue was passed over... Actually that was rather informative. Hardly makes them a loon, no?
Well, keep the links coming. Right now you two look like the loons, but I am still prepared to be enlightened. I don't see anything loony at all in what I actually cited, and apparently neither do you, or I'm sure you would have pointed it out. But with your ad hominem attacks, the beauty is that you don't have to - you just have to find something, anything vaguely embarrassing said by anyone associated with the site, and you've "scored" your "point." Carry on, gentlemen (and/or ladies).
Ah, mods... this is a factual and informative post about Debian policy.
The controversy over Mono in Debian is hardly offtopic and the viewpoint expressed here is quite rational, given that we are still in the midst of a massive Microsoft-backed anti-free-software lawsuit and FUD campaign (SCO, tomtom, etc).
Slashdot Mirror
If they like this system so much, I'm sure they will have no problem paying out to all the patent-holders they infringe upon, according to the same idiotic legal principles they believe should protect their own works.
Of course, if any more than a handful of crooks started following these rules, that would make the software industry impossible. Not even Microsoft could ever know what they infringe. Even if the baby jesus came down from heaven today and told them the four hundred thousand patents they infringe, they would be lost again tomorrow, when 10,000 more patents were filed.
The only way this absurd system of legalized corporate mugging is truly going to end is when Microsoft and the other lobbyists behind it themselves lose Real Money (i.e. billions of US monopoly dollars) to other patent holders.
I am wishing Implicit all the best in their bullshit lawsuit.
It's gratifying to read some comments by someone familiar with the issues.
I have a few questions:
I am inured to metaphors in the news. I'm afraid Fox News, etc. has these guys beat by about 100x, so "zombie" doesn't impress me particularly. But as a story about a decaying regulatory agency, and the complexities of determining a safety regime for a set of fiercely complex 40 year old machines (which incidentally could cause a spectacular incident that would very likely end your industry, and thus your civilian career, if allowed to fail), it comes across to the layman as entirely plausible - and I do not exactly see your counterarguments yet - assuming you actually disagree with the premise, which as far as I read, was only that we need more stringent oversight of the industry (and not even that the extensions and capacity upgrades should be stopped or rolled back).
...the most salient criticism raised by the "Greenes" was that we were not, as a people, disposed to live up to the "zero tolerance" policy for failure that large scale industrial use of nuclear materials really demands. We always make mistakes eventually. Even if it takes 50 or 100 years, then it means we only have 50 or 100 years until a major nuclear disaster and i.e. epic human suffering, unprecedented economic calamity, the depopulation of a major urban area, the success of a fanatical act of terrorism, etc.
This article rather underscores the point. We have become complacent that we are smart enough and organized enough to use nuclear power safely. As we become complacent, this leads to a false sense of security, laziness and corruption on the part of operators and regulators, apathy on the part of the public, and the decline of safety culture. Now I am sure you will have no problem moving your family in down the street from one of these plants, right?
Right?
Using terms such as 'zombie', "decrepit" and 'unprecidented' without a shred of evidence
Hoping to fool people who didn't read the article?
It presents copious evidence by citing numerous specific incidents at various facilities, and clearly detailing how these incidents are related to age and lax safety culture.
Hence "decrepit."
It also discusses specific regulators at NRC, their backgrounds, and their resumes (which involve jumping between the regulatory agency and cushy jobs at the companies they regulate). It cites a specific ethics violation.
"Zombie" is perfectly valid analogy considering that these plants are unquestionably operating beyond their original design "lifetime." Quite a bit less vivid than many other terms and analogies I've been subjected to by the news media lately: i.e. who "hates America," who'se part of a "Nazi regime," who'se "socialist," who "sides with terrorists," and so forth.
The events described in the article, both in terms of safety incidents and regulatory activity, are prima facie unprecedented.
You fail. Good day, sir.
Weeks -> months. But yes, AMD are not the only offenders in the "launch" game.
Still, it takes some audacity to compare this debacle to the Wii launch, even before they stacked a november failed launch with the same silicon they couldn't ship in their september cards.
Bottom line, if you're a shareholder, all those excuses are just that... they blew it. So why say otherwisde?
Since you can sell it on ebay right now for a $100 premium, I would guess they are not really around all that much. :)
Yup, that seems to be the case. But it _is_ AMD's fault that they claimed they had a "launch" in September - in fact they go right on with the fantasy and "launch" even more products they can't actually sell in November. Why not instead just be honest with senior management, and with customers and retailers, and call this what it is: "limited availability."
Of course they shipped a few thousand of them. I should have added a disclaimer at the bottom that I didn't need to hear from each one of the lucky few who managed to get one. :)
What was I expecting? To be able to buy their "launched" product, within 2-3 months after the launch, at or near MSRP.
Hey, all I'm saying is, to call this a "launch" is a joke. I think AMD would be smart to say "limited availability" and jack the price so that they are capturing ebay dollars instead of letting lucky people like you make a profit on MSRP vs. real value of the goods, and screw over themselves and their retailers.
Lots of products are launched for the holidays. If you can't do something ("launch", "MSRP") don't advertise that you can. But they can't do that, because then some pissant vendors and/or middle management would have to fess up to the fact that they blew their deadline, and they cannot capture all the dollars they could have, if for instance they actually could do what they claimed (sell product X at price Y, right now).
Oh, I have no doubt a few thousand of them have shipped overall. But these cards launched in September. They are still so far off from meeting demand that it is a joke.
Have a look around for your card today. You will find every retailer out of stock. ETAs are now running into December when they are given at all. And you will find it on ebay, for a ~25-50% premium. Most customers who want this (at anywhere near the MSRP) are still waiting 2 months after the launch, and that could turn to 3 months or more.
Imagine if every company "launched" this way.
This is a comment on AMD's business, marketing, and PR, rather than their technical team. AMD has unquestionably won the latest round against NVidia, who will have to wait until next year (and miss the holidays) before they have a shot at retaking the top performer and price-performance crowns back.
But let's be real. The 5850 and 5870 have already "launched" too. But unfortunately AMD's idea of a "launch" is "you can buy it 4-16 weeks from now."
I see a lot of companies "making their deadlines" this way. i.e. by not actually making them. Surprised at how often the press gives them a pass on it.
Well, we've convinced ourselves that laissez faire capitalism is actually brilliant and the only reason the 19th century and the Great Depression were so non-fun was that we "just weren't doing it right." Time for round two.
You've got economists with the bullshit theory that markets are efficient and that consumers make rational, perfectly informed choices about every economic decision in their lives. This hilarious concept is actually used as cover by policy makers and regulators, to justify their actions, since it is easier to say this than to admit, "hey, I took a bribe!" The result is a "free" market where participants can be fucked over with impugnity via scams that are generally illegal in other countries, and then blamed for being victimized.
By now we have reached a level of intellectual depravity where we actually believe we can have an "efficient market" of i.e. cell phone service, land line service, or electricity - concepts which are funny, ludicrous and pathetic, in that order.
For cell phones, sure, you can make a new competitor, if you can raise several billion dollars - no problem, right? This market is so "efficient" that the 5 US participants rig prices in broad daylight. Unless you actually think the cost of sending an SMS is not only measurable but has increased 400% since ~2000. As usual, the rest of the first world (Europe, Asia) regulates these business in a more sane way and have had vastly better and cheaper cell phone service for many years. Luckily Americans have no idea what goes on in the rest of the world, so our sense of superiority need not be threatened.
Land lines are even more fun. No point in recapping our amusing attempts to apply antitrust law to The Telephone Company, only to let all the little parts practically merge back together again. Verizon's 200 billion monopoly dollars in annual revenue is put to entirely efficient uses, you can be assured. Fortunately if you don't like them, you can switch to an entirely different brand name for your Verizon land-line. At least we have the self-respect to let an unaccountable plutocrat own and profit from the infrastructure, rather than a democratic government, in Jesus's Name, I Pray.
Electricity is even more fun, since as a system it was basically designed by ex-Soviet robber barons. As Republican-patrons Enron demonstrated in federal court, the market is so "efficient" you can even create a shortage just by never building more power plants, and even turning off the ones you have. This is why privatized electricity has been so good at reducing costs in every instance it was tried. By negative dollar amounts.
You know what? I can see why privatization is so attractive. What's the point of letting government try to fix anything. Just let some lucky individuals own these natural monopolies. Not only can we not vote with our wallets, we cannot figure out how to vote, period. If the voters are too stupid to defend themselves from these scams, they don't deserve to keep their money, right? Just a bunch of children in need of some Catholic school vouchers, right?
We're hardly starting at zero developing test cases for filesystems. A reputable OS shop will have a battery of them. As the state of the art in that field progresses, new ones are required continuously, since we keep finding new and interesting ways for a FS to fail. I personally have kept seeing many novel and amusing ones over the years. :)
I neither have test cases nor bug fixes for ext4. I only say that you should have a much, much harder time finding bugs, patches, and blog posts recounting stories of woe and data loss, before you decide a filesystem is ready for Linux's largest group of low-skilled users.
They do not care about the advantages of delayed allocation or nanosecond timestamps. But when their thesis or novel or day's video shoots go poof and our answer is some blame-the-victim nonsense about backups, or reading all the fine print... who wants this on their conscience, really? Especially when it was so easily avoidable, either by not taking silly chances, or warning more loudly not to trust us.
Yes, and Reiser was nuking people's data for years. He used to claim his FS was stable before his consistency checker was even finished. XFS was/is better by comparison, and neither are anyone's default filesystem. Obviously it's the default that matters, not what other options you offer (where presumably someone making an affirmative choice knows what they're doing, or will learn the hard way fast enough).
When talking about an mp3 player, and only one person out of 100 says "it crashed" then no one gets heated. When talking about a filesystem, someone saying "But I used filesystem X for two years and never lost data," and expecting that fact to matter, is really just saying "I do not understand how to discuss filesystem reliability." A 1/100 or 1/000 shot of data loss in filesystem-land is not a finished product.
It does seem as if Fedora switched to ext4 as a default in April. Good old Redhat. I remember back in the day when they switched to PAM while it was still in the low 0.# versions, and was barely functional or documented. Many a server needed a rescue CD after one of several rpm upgrades "disturbed" the pam configs and all authentication would then fail... But the sloppiness and race to the bleeding edge was just typical for them, and among Linux users at the time, it was hardly a scandal, just yet another headache, to be expected. IMO Fedora never made a serious effort to be usable by non-experts. I used to expect better from Canonical, but apparently not anymore.
You amaze me, man.
How did you figure out how long the average newbie would go before hitting the bug? And how did you estimate the time to the fix? By saying "Remember, every file system has weird bugs", I'm sure you don't imply there aren't certain widely accepted standards for stability in a filesystem? Not zero bugs (an untestable condition), but on the other hand, not "we down to a dozen patches a release, and there are only a few open data loss bugs. Well, wait, there's another."... :)
Your ability to estimate dates and quality levels is no doubt fearsome. But there is an easier way: simply shipping with an ext3 default for another 6 months (or however long it takes until the code stabilizes to the appropriate degree for a filesystem). Most users will not even notice. This is the way you behave if you have any conscience - or if moral arguments don't appeal, if you care about your reputation, or want any repeat users of your products or services.
In case you are not keeping up, Linus Torvalds himself reported losing a filesystem to ext4 in a kernel bugreport last week.
This is just not a filesystem that is ready for primetime. And that is no slight on anyone - these things take time. For a filesystem, in particular, many years. Where I do have a problem is when some mean-spirited person is out there tricking people into using it before it's stabilized, without properly communicating the risks.
Risks which, if you can make the average Ubuntu user understand them, they would almost never choose to take.
Oh, so no one can just be a first time user, use the thing for a while, put a lot of stuff on there, and get bitten when they finally cross some wacky hidden threshold that Canonical buried on their website? Right I forgot, assuming everyone has backups is just how we do things...
Your way of thinking about filesystem reliability, while interesting, isn't that useful to anyone.
You've got to be kidding me.
Work with any file larger than 512MB and you get intermittent corruption, and you seem to imply it's not that bad?
Read the bug, for god's sake, people's apt cache gets corrupted before they even start heaven forfend installing real apps, doing development, working with media, etc...
All they had to do was just not use ext4 until it was ready. I.e. no open data loss bugs for a little while. Probably a conservative approach is warranted, it being a filesystem and all... That's _all_ they had to do.
Intel video driver issues were huge. Also, for Kubuntu users, the transition to KDE4 (which started at 8.10) was botched and caused a mass exodus from KDE and Kubuntu - some of which was down to KDE and a lot of which was down to Canonical. Linus Torvalds stopped using KDE during this period IIRC, though I don't have the link handy for his (characteristically scathing) comments. By comparison, KDE3 in 8.04 was a shockingly polished, stable system.
To this day, I see 9.04 kernel-crashing occasionally on the best supported laptops (thinkpads of various vintages) where before, crashes on that same hardware were unheard of with ubuntu/kubuntu.
Amazingly, if you click on the release notes link all the way down at the bottom of the "cool new features" page, and read about 2/3 of the way down that page, oh yeah by the way:
Possible corruption of large files with ext4 filesystem
There have been some reports of data corruption with fresh (not upgraded) ext4 file systems using the Ubuntu 9.10 kernel when writing to large files (over 512MB). The issue is under investigation, and if confirmed will be resolved in a post-release update. Users who routinely manipulate large files may want to consider using ext3 file systems until this issue is resolved. (453579)
What... the... fuck... are these morons thinking. They make ext4 their default filesystem, and release to the world with a bug like this open. I don't care whether they charge for this, give it away for free, or pay people to use it. This is not a release, it's an ugly prank. A giant fuck you to the entire world. "Hey, here's an awesome free OS! Just kidding it eats your files, lulz!!!!"
How hard is it to put a giant warning label ("MAY EAT YOUR FILES, TBD") on the front page? Why not just say, whoops the release will be late, sorry? What is it about their ego that makes it more important than peoples' data? You don't have to do work for free, just don't fucking trick people!
Between the increasing mess of 8.10 and 9.04, and this debacle, I am just losing all respect for Canonical.
Also, btw, the whole story is wrong and 9.10 is not available for download yet. Thank god.
Don't get me wrong - I think Indilinux looks to have a good future, and I am glad that they supported their users. I just can't really praise them for not finishing the product before they started selling it.
The real problem here was about "time to market" - companies were hustling to get to a lower price point before their competitors, and build their reputation in a new market. In that race, it happens again and again that corners get cut. No surprise really, and I don't think it will ever change.
As for the best bang for the buck, I find it harder to make as definitive a statement as you have. In any case, I don't think I could add anything to the (fairly nuanced) words of tomshardware or anandtech.
Not anymore, I think, but there was a period of time a little while back when the firmware updates were coming thick and fast.
Even if every firmware update works, that's not an "OK" situation for many people/companies, and not every update is guaranteed to work. Compare this to Intel drives, which have had a very low single digit number of updates over their model life on X25-M G1/G2, for instance. The firmware was "finished" before the product shipped, it had comparatively few issues (like halting or weird slowdowns, though there was one early on IIRC) and its performance has been consistent and generally better than other drives for the specs that matter most on the desktop (random reads and writes).
Thank you. The brands/models were the critical piece of information.
You're probably aware that SSD's have been in the server space, at a very different price point, for a few years now, without any extraordinary reliability debacles. To some extent, this is a case of getting what you pay for. I did a moderate amount of research on SSD drives, relying especially on the independent review sites, and quickly eliminated all of the brands you described.
As is frequent in fairly new markets, there are a few smaller and less well-run companies trying to dive in, and their first customers get to beta test their v0.* and v1.* offerings.
The prevailing wisdom seemed to me (and to people like i.e. Torvalds) that Intel was far and away the top of the heap in terms of performance and reliability, and some drives based on a newer Samsung controller (i.e. OCZ Summit) were a perhaps credible alternative. Other brands were clearly struggling to even be in the game, with frequent firmware updates and outright debacles (i.e. Indilinux, Micron) and we're in the process of shaking out who will make it and who will not.
I have only fielded a few consumer-grade SSDs over about the same amount of time as you, but going with Intel's G1 and G2 MLC products has so far yielded zero failures.
If you are already in the market for an SSD, and you are ready to spend premium money for premium performance, you should go the whole distance and go with Intel, the current market leader. See also the latest news on these models.
I agree. That said, I actually sat through the entire video, and sadly, it was a disappointment. I ignore everything that he did that came before. He has great intelligence, a wonderful idea, some vague but intriguing suggestions on how to accomplish it, and very little working code.
At least he is up front about this - if by up front, you count a litany of near-impossible problems you do not even have an idea about how to solve in the last 8 minutes of your 35 minute video. :) I do hope for the death of the fad of using videos where a perfectly good blog post will do.
In truth I think he is on to something, or at least, that problems of parallelism will be successfully attacked in ways similar to what he suggests. I can't really cast blame on anyone for being excited by an idea or building a tiny prototype that does something cool. But I would personally be way too embarrassed to say "hey guys look at this" with so little meaningful work done.
OK. Wow, so loony. Anything else?
Ah, some actual points one can respond to. That uncommented link to the top of a giant chatlog... sorry tl, dr.
I suspect being able to run Windows in Bootcamp and Fusion was a major factor in Apple's switching to Intel, yes? Now only Apple computers can legally run both operating systems. And you've found some people chatting about that in IRC... OK...?
And someone griping in IRC that Miguel's article got posted while groklaw's extensive treatment on the same issue was passed over... Actually that was rather informative. Hardly makes them a loon, no?
Well, keep the links coming. Right now you two look like the loons, but I am still prepared to be enlightened. I don't see anything loony at all in what I actually cited, and apparently neither do you, or I'm sure you would have pointed it out. But with your ad hominem attacks, the beauty is that you don't have to - you just have to find something, anything vaguely embarrassing said by anyone associated with the site, and you've "scored" your "point." Carry on, gentlemen (and/or ladies).
Ah, mods... this is a factual and informative post about Debian policy.
The controversy over Mono in Debian is hardly offtopic and the viewpoint expressed here is quite rational, given that we are still in the midst of a massive Microsoft-backed anti-free-software lawsuit and FUD campaign (SCO, tomtom, etc).
References: