It's just a bunch of gobbledygook. "Storage lockers"? Um, yeah. Blogs having some special property that renders virus scanners inoperative? Not last time I checked. Really, I don't see any sense in the whole thing besides "hey guys, there's some adware and stuff on blogs now." "hey, thanks for the heads-up, I guess that had to happen eventually."
Haha... I like the "lookaside buffer" bit. I'm sure that it came from TLB "translation lookaside buffer", but on its own it seems like the exotic cousin of lookahead and lookbehind. Kind of like elevators going sideways.
Some of the latest versions of the stuff on marillat are indeed incompatible with Ubuntu, as they depend on a version of glibc which hasn't made it in yet. It's basically just a case of "Ubuntu isn't Debian". Wait for the problem to go away, for someone to create a repository that works on Ubuntu, compile your own, or use something like Totem.
Nothing even close to modern runs on Debian stable, and the compatiblity nightmare they call "testing" is just as bad. I've been defending (and using) Debian for a lot of years now, but they need to get their act together, stop complaining, build a decent project, and release it, or just shut up and die. Metaphorically, that is. Ubuntu is a net positive for the users. Whether it's a net positive for Debian is mostly up to Debian.
Is a more lightweight, disposable version of those plastic keyboard covers. It would be a membrane thin and flexible enough that it wouldn't interfere with your typing, but which could be thrown out at the end of the day. It would also, conveniently, protect your keyboard from wayward food particles and corrosive finger oils.
I can only assume that if making such a thing were easy it would have been done by now.
That's not actually what it does, though; it does Fancy UDP Tricks (tm) that require a third party for the initial setup phase, and then allows direct communication between the two original parties with no extra help. It's not even a secret how it works, so I wonder that we don't see it around more often.
No, some people (who, by the way, work for Skype) told you that Skype conversations are encrypted end-to-end. But because the source isn't available, and the Skype developers believe that obscurity is the best security, you have no way of knowing that. For all anyone knows, it could be that there's no real encryption at all, but that the data is just whitened by a PRNG so it looks encrypted. It could be that there is real crypto going on, but the key-exchange is boobytrapped so that Skype and/or the appropriate TLAs have snooping power. It could be that any one of a number of flaws makes what was intended to be an effective algorithm vulnerable. We don't know any of these things, but the fact that the people who wrote the encryption software in the first place don't trust it to remain unbroken in the face of public scrutiny means that you shouldn't trust it either.
Apparently the operating costs of a space elevator will be approximately 0.45359 -- no units -- no matter where it's going or how much stuff you're lifting. That's potentially a good thing, although we still have to figre out how to come up with 0.45359. Has anyone ever seen a number? (No, I don't mean a numeral.)
Er... what about dmix? That's certainly at a lower level than the sound servers you mention. It's actually done in the ALSA libraries, not the kernel (which is probably a good thing), but it's synchronized, and it works, and as long as everything on your system uses ALSA, there are no problems.
You mean "By default, most distributions don't configure ALSA to do the software mixing that 90% of AC'97 cards need because they're too cheap to do anything right", yes?
I was including everything in namespaces (default), category, wikipedia, and template. Maybe a little more than you, but probably not much. In any case, I'm working on some dictionary compression of my own; it looks pretty effective so far.
Interesting. I was looking into dictionary compression, but I hadn't gotten there yet; in the meantime, I was using an algorithm based on bzip2, arranged in such a way as to keep the compression benefits of a larger block size, but to still allow (mostly) random access; the full 'en' archive weighs in at 440MB with a 128k blocksize, or about 400MB if the block size is bumped up some more.
This is school we're talking about. Being talented and creative are dangerous and not recommended. Trying hard only means that you wasted more of your time than the next guy. If you want to write something and actually have it read, try the intarweb. Or write a book.
You could if you studied it for a minute. The rules are actually relatively simples. For example, let's look at everyone's favorite word: "Aktiengesellschaft", meaning a public corporation, and usually abbreviated "AG".
Semantically, it can be broken into Aktien / Gesell / -schaft; taken at this decomposed level it means something a little bit like "stock fellowship".
Orthographically, well, you're in luck; as usual (always?), it breaks down along the same lines, according to the rules. What are the rules? I couldn't tell you exactly, but they're simple, and they're similar to Latin's. Anyway, it breaks down to Ak/tien/ge/sell/schaft. Breaks occur between consonants that don't form clusters, between vowels that don't form diphthongs, and otherwise before consonants.
Still, I don't understand this. They must be doing something in the most simplistic way possible, or including all of the media (not just Commons). I've been working on a project recently, to create wikipedia CDs, and I've been able to get the full text of the English wikipedia, in a browsable format, into under 450MB. Searching isn't implemented yet, but I think it should be possible to fit a search index, plus all of the images from Commons, onto a single CD.
No, you can't; the grandparent post actually makes no sense. A random developer can't say "you don't have the right to use my code anymore" and take away any rights that the GPL already granted you. That developer could release all of his further changes under the "GPL except CherryOS can't use this code" license, but said license is clearly GPL-incompatible:)
What does affect CherryOS is section 4 of the GPL itself, which essentially states that any attempt to violate the GPL terminates any rights that the GPL might have granted you. Combined with section 5, that should mean that CherryOS has no right to distribute PearPC code.
Now, for some unrelated speculation. Maybe they're planning on releasing some "bleached" source, and then say "look, guys, we opened the source just to make you happy, and prove to you that we never used PearPC code" ?
If you deny MS the right to use your code, then it's not GPL. Of course, it's hard to see a situation where it would really be necessary to do such a thing; when you're Microsoft, the GPL is about the biggest disincentive there is.
Slashdot Mirror
It's just a bunch of gobbledygook. "Storage lockers"? Um, yeah. Blogs having some special property that renders virus scanners inoperative? Not last time I checked. Really, I don't see any sense in the whole thing besides "hey guys, there's some adware and stuff on blogs now." "hey, thanks for the heads-up, I guess that had to happen eventually."
OP didn't say that a standard-bearer was a flagpole, merely that they were related. And obviously they are.
Maybe you should read what you're replying to. 'practical' and 'prgamatic' are, um, practically synonyms :)
Haha... I like the "lookaside buffer" bit. I'm sure that it came from TLB "translation lookaside buffer", but on its own it seems like the exotic cousin of lookahead and lookbehind. Kind of like elevators going sideways.
Anyway, the answer is: If the design is competent, probably 30 years. Otherwise, 30 minutes.
Some of the latest versions of the stuff on marillat are indeed incompatible with Ubuntu, as they depend on a version of glibc which hasn't made it in yet. It's basically just a case of "Ubuntu isn't Debian". Wait for the problem to go away, for someone to create a repository that works on Ubuntu, compile your own, or use something like Totem.
Nothing even close to modern runs on Debian stable, and the compatiblity nightmare they call "testing" is just as bad. I've been defending (and using) Debian for a lot of years now, but they need to get their act together, stop complaining, build a decent project, and release it, or just shut up and die. Metaphorically, that is. Ubuntu is a net positive for the users. Whether it's a net positive for Debian is mostly up to Debian.
Is a more lightweight, disposable version of those plastic keyboard covers. It would be a membrane thin and flexible enough that it wouldn't interfere with your typing, but which could be thrown out at the end of the day. It would also, conveniently, protect your keyboard from wayward food particles and corrosive finger oils.
I can only assume that if making such a thing were easy it would have been done by now.
That's not actually what it does, though; it does Fancy UDP Tricks (tm) that require a third party for the initial setup phase, and then allows direct communication between the two original parties with no extra help. It's not even a secret how it works, so I wonder that we don't see it around more often.
No, some people (who, by the way, work for Skype) told you that Skype conversations are encrypted end-to-end. But because the source isn't available, and the Skype developers believe that obscurity is the best security, you have no way of knowing that. For all anyone knows, it could be that there's no real encryption at all, but that the data is just whitened by a PRNG so it looks encrypted. It could be that there is real crypto going on, but the key-exchange is boobytrapped so that Skype and/or the appropriate TLAs have snooping power. It could be that any one of a number of flaws makes what was intended to be an effective algorithm vulnerable. We don't know any of these things, but the fact that the people who wrote the encryption software in the first place don't trust it to remain unbroken in the face of public scrutiny means that you shouldn't trust it either.
Apparently the operating costs of a space elevator will be approximately 0.45359 -- no units -- no matter where it's going or how much stuff you're lifting. That's potentially a good thing, although we still have to figre out how to come up with 0.45359. Has anyone ever seen a number? (No, I don't mean a numeral.)
Er... what about dmix? That's certainly at a lower level than the sound servers you mention. It's actually done in the ALSA libraries, not the kernel (which is probably a good thing), but it's synchronized, and it works, and as long as everything on your system uses ALSA, there are no problems.
You mean "By default, most distributions don't configure ALSA to do the software mixing that 90% of AC'97 cards need because they're too cheap to do anything right", yes?
I know it's one of those nasty, confusing ones, but the word you were looking for there is "hindrance".
I was including everything in namespaces (default), category, wikipedia, and template. Maybe a little more than you, but probably not much. In any case, I'm working on some dictionary compression of my own; it looks pretty effective so far.
Interesting. I was looking into dictionary compression, but I hadn't gotten there yet; in the meantime, I was using an algorithm based on bzip2, arranged in such a way as to keep the compression benefits of a larger block size, but to still allow (mostly) random access; the full 'en' archive weighs in at 440MB with a 128k blocksize, or about 400MB if the block size is bumped up some more.
This is school we're talking about. Being talented and creative are dangerous and not recommended. Trying hard only means that you wasted more of your time than the next guy. If you want to write something and actually have it read, try the intarweb. Or write a book.
What you're saying is more-or-less in line with what the "Wikipedia 1.0" project has been talking about; maybe you could chime in there.
You could if you studied it for a minute. The rules are actually relatively simples. For example, let's look at everyone's favorite word: "Aktiengesellschaft", meaning a public corporation, and usually abbreviated "AG".
Semantically, it can be broken into Aktien / Gesell / -schaft; taken at this decomposed level it means something a little bit like "stock fellowship".
Orthographically, well, you're in luck; as usual (always?), it breaks down along the same lines, according to the rules. What are the rules? I couldn't tell you exactly, but they're simple, and they're similar to Latin's. Anyway, it breaks down to Ak/tien/ge/sell/schaft. Breaks occur between consonants that don't form clusters, between vowels that don't form diphthongs, and otherwise before consonants.
Indeed; this is what I mentioned up above. It can be done, and it can be done reasonably well in a cross-platform app.
Still, I don't understand this. They must be doing something in the most simplistic way possible, or including all of the media (not just Commons). I've been working on a project recently, to create wikipedia CDs, and I've been able to get the full text of the English wikipedia, in a browsable format, into under 450MB. Searching isn't implemented yet, but I think it should be possible to fit a search index, plus all of the images from Commons, onto a single CD.
No, you can't; the grandparent post actually makes no sense. A random developer can't say "you don't have the right to use my code anymore" and take away any rights that the GPL already granted you. That developer could release all of his further changes under the "GPL except CherryOS can't use this code" license, but said license is clearly GPL-incompatible :)
What does affect CherryOS is section 4 of the GPL itself, which essentially states that any attempt to violate the GPL terminates any rights that the GPL might have granted you. Combined with section 5, that should mean that CherryOS has no right to distribute PearPC code.
Now, for some unrelated speculation. Maybe they're planning on releasing some "bleached" source, and then say "look, guys, we opened the source just to make you happy, and prove to you that we never used PearPC code" ?
I'm sure it did; the problem is that I can't read it and have it rhyme.
MAPS can't do any harm on their own. The real problem is people who use MAPS' braindead advice as part of their policy.
If you deny MS the right to use your code, then it's not GPL. Of course, it's hard to see a situation where it would really be necessary to do such a thing; when you're Microsoft, the GPL is about the biggest disincentive there is.