It has to do with the analog nature of the storage. If you had 0, 1, 1, 0, and you overwrote that with zeros, you'd then have 0,.1,.1, 0. Chances are that the drive itself (without at least modified firmware) can't tell the difference, but a data recovery lab can. You can actually still read data after it's been written between 5 and 20 times - each time, subtract the obvious and multiply the residue.
For starters, DNA is digital. Yes, transcription errors will be made in cloning, just as they are in normal reproduction. But this isn't scary, it's just mutation, and it's a driving force of evolution. Not all mutation turns you into Radioactive-Man, in fact, the vast majority has no impact whatsoever.
I think there is a parallel between this and nuclear physics - another field where a little bit of information is a very dangerous thing. Towards the beginning, things were very secretive, and the goal of many governments was to make nucler weapons. Had it remained entirely secretive, we'd have even more nuclear weapons and not much else. Instead, the information opened up, and we now have nuclear power, a much better understanding of the atom, particle accelerators, different competing theories of everything, and so fourth. The technology has both good and evil uses, but lowering the "barrier of entry" has actually favored the good uses. It created a much larger field of researchers than small secretive teams could have sustained.
The same thing can happen in biotech. If the information exists, a determined person or government can get it, regardless of how high the artifical barriers to entry are. So the obvious evil uses will be fulfilled either way. The good guys, however, aren't typically as determined. (they won't be spying, using fake ID's, etc. to get at "deadly information.") So you'll have a much smaller feild of research going on, and technology will be slowed, without a significant impact on the "evil" uses.
Well, it's an excercise in both space efficiency and ergonomics. 1cm3 cubes aren't easy to manipulate, and you can't write much on them. If it was me, I'd make three designs. One would be the size of a fat CD-case and work like a Zip disk. This would hold about 50cm3 = >50 GB. You'd carry one of them with you, and it'd be your life, until it runs out. For digital cameras and portable Vorbis players (this is the future, remember?), you'd have a smaller thing, like a CF card only with about a GB. A small network might buy 1dm3 cubes from time to time for data backup, and huge corperate data-farms might buy 1m3 washing-machines.
Methinks authors use word processors, so the book itself isn't its source.
It is very harmful for you to pretend to be someone who would know about these issues, and then spread misinformation about them, when in fact you are not who you claim to be.
It depends on whether the distribution is commercial or not. If it's a noncommercial distribution, you can tell people to get the source from wherever you got it from (kernel.org) under 3(c). But you can only choose 3(c) if it's noncommercial, so an embedded hardware company would need to follow 3(a) or (b), either giving the source with the binary or giving a written offer, valid for 3 years, to get the source under the GPL for a nominal charge.
Exactly - if you violate the GPL, you get sued for copyright infringement because you didn't comply with the only license that lets you make copies.
I agree that you have to give out your.config, which is a "script used to control compilation." But obfuscating the code isn't a loophole - the GPL says:
The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable.
So unless a company thinks it can convince a court that obfuscated or de-compiled code is the "preferred form of the work for making modifications to it," it better not do that.
Furthermore, even if they do end up GPL-ing the infringing code, we're not talking about their "entire codebase." They use a modified Linux kernel, but the interesting things happen in userspace. Linus's note at the top of linux/COPYING says non-GPL userspace programs are OK. To comply, they'd only have to open drivers they've written for the special hardware.
All this guy is asking for are the kernel mods that let Linux run on that chip, so he can make his own PC-cards work.
A little clarification is in order. Every computer owner has the choice to use Windows or not. But the vast majority didn't make an informed decision based soley on the relative merits of the competing operating systems.
I would guess 75% of desktop users don't even know there are options (True quote: "You don't use Windows? You don't like Bill Gates? But you're a computer guy! And Bill Gates wrote Windows!"). These are the people who bought a computer but didn't consider the OS. Probably 15% wish they didn't have to use Windows, but are "forced" to by outside pressures: jobs, compatibility, killer apps, etc. Finally, 10% for the rest of the OSs, mostly Linux and Mac.
So, 75% didn't know they had a real choice, 15% made a choice under duress, and 10% made a free choice.
While on the subject of Mozilla's marketing, check out Firebird's webpage. IMHO, they do a great job of showcasing the features in non-technical language. They may be a little heavy on the next-generation solution of the future best-of-breed easy-to-use buzzwords of success, but for an open source project, it's a very polished webpage.
Seriously, we need more marketing people, artists, and UI designers helping out with open source. They did a good job with Firebird.
Regarding 2.6.0, it's a little late to speculate on the pre-releases. 2.6.0-test11 is out now, and it will be the last test release. In two or three weeks, after the bug reports subside to a dull roar, 2.6.0 will be out. It will, however, be interesting to see how Andrew Morton takes care of 2.6.x (x > 0) releases.
Using the VFS layer, which is likely dead anyway, isn't a good idea, but there are still safe ways. There are some patches that use their own minimal system to write to swap partitions on IDE disks. Safer still, there could be an option (not default) to dump to a floppy using a self-contained system. Probably safest would be to write the dump somewhere in RAM that won't be erased during a warm boot, and save that memory on the next boot until userspace can save it somewhere.
When the kernel crashes, it doesn't just stop, it makes some effort to tell you how it crashed - I just think if it were easier to get dumps without manual copying or special hardware, more dumps would be returned. At least the "save in RAM" technique could go in mainstream or distro kernels - what's the possible harm in that?
Congradulations, you run servers. But the majority of desktop users (who are more likely to run testing kernels anyway) do run X and don't have serial consoles. I do think an oops dumping system that didn't require no X or a serial console would get a lot more dumps back to the kernel developers.
I'll tell you what I see for the future of instant messaging. There will be a bunch of companies trying to make one IM to rule them all, each with their own incompatible protocols and clients. IM is so low-bandwidth that it's practical to have one centralized server, which gives companies the ability to advertise and the ability to sneak software onto the computer via the client. Chances are Microsoft will win this battle in the long run (by bundling with Windows as they already do), though AIM won't be far behind. Secondarily, there will be a few free or adware clients trying to communicate with all protocols. This is somewhat good for users, but whoever has the greater market share will try to ban that client, because having a universal client makes it harder to lock in customers.
Meanwhile, I plan to wash my hands of the whole mess and use Jabber. Remember back when we had standards, and the internet was decentralized? It actually worked - there wasn't a single point of failure. When was the last time the entire email system went down? Jabber can offer the same reliability, and you don't aren't locked into a single server or client.
Besides being decentralized, Jabber tries to offer gateways, and many Jabber clients (such as GAIM) also play the "keep up with the proprietary protocol" game. So have the best of both worlds - get a Jabber account somewhere, and whenever your friends's servers lock out their clients of choice, convince them to get a Jabber account also.
Unless one of the people working on the ACPI subsystem have your laptop, they have no way of knowing about the problem. So submit a bug report - if the fix is a one-liner, it might still go in.
I really wish some of the "save oops to disk (or high memory, or floppy)" patches went into the mainline kernel. Lets face it - everyone runs X, and nobody has a serial console, so most people won't see the oopses. And even if they do, few people want to copy them down by hand. This really limits the amount of useful bug reports the kernel developers are getting.
After these patches become mainstream, somebody could make an automated system to ask the user to describe the problem, then send a bug report with the oops,.config, dmesg, etc.
You wouldn't think so, but there is a good reason for it. To make a sweeping generalization, Slashdotters want to see Linux succeed, and we have the technical knowledge to compile kernels and make informative bug reports if needs be. Publishing this story on Slashdot will entice more people to test the new kernel, thus ensuring that the 2.6.0 release will be Bug-free(tm).
mod_gzip does save some overhead, but it's not perfect. This is http://www.alistapart.com/d/slashdot/index.html, in it's original form and with all spaces squeezed to a single space, uncompressed and gzipped:
-rw-r--r-- 1 tom users 30667 Nov 25 22:24 nospace.html -rw-r--r-- 1 tom users 33923 Nov 20 14:38 index.html
-rw-r--r-- 1 tom users 9613 Nov 25 22:24 nospace.html.gz -rw-r--r-- 1 tom users 10015 Nov 20 14:38 index.html.gz
So, without gzip you save 9.6%, and with gzip you save 4.0% by eliminating spaces. Actually, you could save even a little more - my pipeline just turned all whitespace into one space, whereas spaces between tags could be dropped for even greater savings.
Obviously, for a demonstration meant to be perused by Slashdotters, indentation is a good thing. But I think for the real Slashdot servers, the 4% bandwidth savings might be a good idea.
While I don't have a benchmark, I do think a wider space is a readability advantage. I figure, people generally start and stop reading at a sentence boundary, if not in a paragraph boundary. So for the same reason a double space or indent before a paragraph is useful, two spaces between sentences is good - it's easy for the eye to search for that. Yes, because the period is small, you get that to some degree automatically, but you get the same for commas. The initial capital helps also, but that also happens before proper nouns. Basically, a period, two spaces, and a capital is the easiest pattern to search for.
This is one of the most insightful articles about DRM I've read in a long time, because it doesn't listen to the RIAA/MPAA's cover stories about DRM. The weaknesses of DRM schemes are obvious - any DRM will be eventually cracked. Even if Palladium is implemented flawlessly, there will still be the analog hole - something that can't be fixed without an encrypted digital channel to a cochlear implant. Finally, people are being told that DRM isn't about piracy. Although the article doesn't explicitly state it, the real target of DRM is fair use - a sense of "owning" the content you buy, an ability to use it how you see fit, so long as you don't run afoul of copyright laws.
In the 80s with VCRs and tape recorders, people showed that they wanted time- and space-shifting fair use rights, and the law followed. Now the law is swinging back, as the DMCA can make those things technically illegal - consider that if the DMCA and the broadcast bit existed then, VCRs would be illegal now. But the content owners were unable to stop Xerox machines, VCRs, tape dubbers, digital audio extraction, CD-RWs, and portable MP3 players, because people really do want to "own" content.
When you make a sale, both sides get something they want. The RIAA wants money, theoretically so they can pay artists to make music. People want music. Specifically, they want to "own" music, as in, "to have the ability to play it, whenever, wherever." This is where the balance lies - if people could redistribute, artists wouldn't get paid, but if people couldn't "own" (in the sense of sovereignty, not copyright), they wouldn't buy it, and again, the artist starves. DRM tries to do just that - take away "ownership," in return for, nothing but inconvenience. I don't think this would happen in a competitive market. I can only hope it won't happen in the present market.
Yeah, primary/secondary is valid too, and that's what I usually say (until now). But anyone who's offended by master/slave deserves to be offended, because they're stupid. It's OK to be ignorant of computer terminology, but if you are, don't be writing laws to govern it.
Well, I hope you're not taking what I said the wrong way. I like some popular music (which is why I care), and besides, its quality is irrelevant. And I'm certainly in favor of copyright and against mass copyright infringement (most P2P). So I think I agree with you. As a practical matter, however, DRM makes it difficult for me to do what I want with my music - organize and play it in any OS with any software, burn it, back it up, transfer it to any portable player, and generally do whatever I can do with a CD. In a word, I want convenience.
I just think that in a competitive market, DRM wouldn't happen, because it offers nothing but potential limitations and incompatibility. I'm angry that it's happening anyway, I'm angry that it gets legal protection (DMCA), and I'm angry that so many people, even Slashdotters who should know better (not pointing at you), accept the line that it's about piracy. That's all. Sorry if I sound melodramatic.
Slashdot Mirror
It has to do with the analog nature of the storage. If you had 0, 1, 1, 0, and you overwrote that with zeros, you'd then have 0, .1, .1, 0. Chances are that the drive itself (without at least modified firmware) can't tell the difference, but a data recovery lab can. You can actually still read data after it's been written between 5 and 20 times - each time, subtract the obvious and multiply the residue.
For starters, DNA is digital. Yes, transcription errors will be made in cloning, just as they are in normal reproduction. But this isn't scary, it's just mutation, and it's a driving force of evolution. Not all mutation turns you into Radioactive-Man, in fact, the vast majority has no impact whatsoever.
The same thing can happen in biotech. If the information exists, a determined person or government can get it, regardless of how high the artifical barriers to entry are. So the obvious evil uses will be fulfilled either way. The good guys, however, aren't typically as determined. (they won't be spying, using fake ID's, etc. to get at "deadly information.") So you'll have a much smaller feild of research going on, and technology will be slowed, without a significant impact on the "evil" uses.
Perhaps he sniffed a password to log in as a non-root user on the box and used this local root exploit to get root.
Well, it's an excercise in both space efficiency and ergonomics. 1cm3 cubes aren't easy to manipulate, and you can't write much on them. If it was me, I'd make three designs. One would be the size of a fat CD-case and work like a Zip disk. This would hold about 50cm3 = >50 GB. You'd carry one of them with you, and it'd be your life, until it runs out. For digital cameras and portable Vorbis players (this is the future, remember?), you'd have a smaller thing, like a CF card only with about a GB. A small network might buy 1dm3 cubes from time to time for data backup, and huge corperate data-farms might buy 1m3 washing-machines.
It is very harmful for you to pretend to be someone who would know about these issues, and then spread misinformation about them, when in fact you are not who you claim to be.
It depends on whether the distribution is commercial or not. If it's a noncommercial distribution, you can tell people to get the source from wherever you got it from (kernel.org) under 3(c). But you can only choose 3(c) if it's noncommercial, so an embedded hardware company would need to follow 3(a) or (b), either giving the source with the binary or giving a written offer, valid for 3 years, to get the source under the GPL for a nominal charge.
I agree that you have to give out your .config, which is a "script used to control compilation." But obfuscating the code isn't a loophole - the GPL says:
So unless a company thinks it can convince a court that obfuscated or de-compiled code is the "preferred form of the work for making modifications to it," it better not do that.All this guy is asking for are the kernel mods that let Linux run on that chip, so he can make his own PC-cards work.
I would guess 75% of desktop users don't even know there are options (True quote: "You don't use Windows? You don't like Bill Gates? But you're a computer guy! And Bill Gates wrote Windows!"). These are the people who bought a computer but didn't consider the OS. Probably 15% wish they didn't have to use Windows, but are "forced" to by outside pressures: jobs, compatibility, killer apps, etc. Finally, 10% for the rest of the OSs, mostly Linux and Mac.
So, 75% didn't know they had a real choice, 15% made a choice under duress, and 10% made a free choice.
Seriously, we need more marketing people, artists, and UI designers helping out with open source. They did a good job with Firebird.
And why would changes to Mozilla or Konqueror affect various other software? It's not like they're part of the OS.
Regarding 2.6.0, it's a little late to speculate on the pre-releases. 2.6.0-test11 is out now, and it will be the last test release. In two or three weeks, after the bug reports subside to a dull roar, 2.6.0 will be out. It will, however, be interesting to see how Andrew Morton takes care of 2.6.x (x > 0) releases.
When the kernel crashes, it doesn't just stop, it makes some effort to tell you how it crashed - I just think if it were easier to get dumps without manual copying or special hardware, more dumps would be returned. At least the "save in RAM" technique could go in mainstream or distro kernels - what's the possible harm in that?
Congradulations, you run servers. But the majority of desktop users (who are more likely to run testing kernels anyway) do run X and don't have serial consoles. I do think an oops dumping system that didn't require no X or a serial console would get a lot more dumps back to the kernel developers.
Meanwhile, I plan to wash my hands of the whole mess and use Jabber. Remember back when we had standards, and the internet was decentralized? It actually worked - there wasn't a single point of failure. When was the last time the entire email system went down? Jabber can offer the same reliability, and you don't aren't locked into a single server or client.
Besides being decentralized, Jabber tries to offer gateways, and many Jabber clients (such as GAIM) also play the "keep up with the proprietary protocol" game. So have the best of both worlds - get a Jabber account somewhere, and whenever your friends's servers lock out their clients of choice, convince them to get a Jabber account also.
Unless one of the people working on the ACPI subsystem have your laptop, they have no way of knowing about the problem. So submit a bug report - if the fix is a one-liner, it might still go in.
After these patches become mainstream, somebody could make an automated system to ask the user to describe the problem, then send a bug report with the oops, .config, dmesg, etc.
You wouldn't think so, but there is a good reason for it. To make a sweeping generalization, Slashdotters want to see Linux succeed, and we have the technical knowledge to compile kernels and make informative bug reports if needs be. Publishing this story on Slashdot will entice more people to test the new kernel, thus ensuring that the 2.6.0 release will be Bug-free(tm).
Who says the evil bit is un-implemented?
Obviously, for a demonstration meant to be perused by Slashdotters, indentation is a good thing. But I think for the real Slashdot servers, the 4% bandwidth savings might be a good idea.
While I don't have a benchmark, I do think a wider space is a readability advantage. I figure, people generally start and stop reading at a sentence boundary, if not in a paragraph boundary. So for the same reason a double space or indent before a paragraph is useful, two spaces between sentences is good - it's easy for the eye to search for that. Yes, because the period is small, you get that to some degree automatically, but you get the same for commas. The initial capital helps also, but that also happens before proper nouns. Basically, a period, two spaces, and a capital is the easiest pattern to search for.
In the 80s with VCRs and tape recorders, people showed that they wanted time- and space-shifting fair use rights, and the law followed. Now the law is swinging back, as the DMCA can make those things technically illegal - consider that if the DMCA and the broadcast bit existed then, VCRs would be illegal now. But the content owners were unable to stop Xerox machines, VCRs, tape dubbers, digital audio extraction, CD-RWs, and portable MP3 players, because people really do want to "own" content.
When you make a sale, both sides get something they want. The RIAA wants money, theoretically so they can pay artists to make music. People want music. Specifically, they want to "own" music, as in, "to have the ability to play it, whenever, wherever." This is where the balance lies - if people could redistribute, artists wouldn't get paid, but if people couldn't "own" (in the sense of sovereignty, not copyright), they wouldn't buy it, and again, the artist starves. DRM tries to do just that - take away "ownership," in return for, nothing but inconvenience. I don't think this would happen in a competitive market. I can only hope it won't happen in the present market.
Yeah, primary/secondary is valid too, and that's what I usually say (until now). But anyone who's offended by master/slave deserves to be offended, because they're stupid. It's OK to be ignorant of computer terminology, but if you are, don't be writing laws to govern it.
I just think that in a competitive market, DRM wouldn't happen, because it offers nothing but potential limitations and incompatibility. I'm angry that it's happening anyway, I'm angry that it gets legal protection (DMCA), and I'm angry that so many people, even Slashdotters who should know better (not pointing at you), accept the line that it's about piracy. That's all. Sorry if I sound melodramatic.