Symantec actually make a separate Norton Removal Tool available from their website, which allegedly (this is second-hand info, I don't know this from personal experience) actually works to uninstall Norton. (It's advertised as a tool to help you recover from problems during the installation, incidentally; I wonder if that's an attempt to justify the existence of a tool that should be redundant to the uninstaller?)
The reason people are interested here is that the main very-short-term energy store in the human body - the chemical used to store energy generated by respiration until it's needed - is in pretty much all known life adenosine triphosphate (ATP). (Because the body can't store ATP, it uses chemicals like glucose for typical short-term storage; ATP is used pretty much immediately after being generated.) The way it works is that the chemical's actually stable with just one, or more commonly, two phosphate groups; the extra phosphate groups take energy to attach, and consequently, if a reaction disattaches the groups as a side-effect, it can be driven in the "forwards" direction even if it normally wouldn't. As a result of this, a whole bunch of enzymes in a typical organism are dedicated to trying to split ATP molecules as a side-effect of apparently unrelated reactions. It's quite surprising that an organism has managed to adapt all its ATP-using enzymes to be based on something else (according to the speculation here on Slashdot, presumably an ATP variant that used arsenic rather than phosphorous), given how widespread it is.
The other really interesting use of phosphorous in typical Earth organisms is that DNA and RNA, two of the most important data-carrying chemicals in human bodies, incorporate phosphate in their construction. (They use a base-4 code with about 4 different variants for each nucleotide in the chain; there are 4 main ones, and a bunch of mostly irrelevant variants which pop up now and then; in most animals and plants, DNA is used for long-term storage, and RNA is used to make temporary copies of DNA, with DNA rarely being used directly except to make copies from. Which nucleic acid is used for what is far from universal in "typical" life in general, though.) ATP (with the two extra phosphate groups removed) actually happens to be one of the four possible major nucleotides that makes up RNA, and is used as a feedstock to synthesize RNA and DNA from. A life-form that didn't use phosphorous would be unable to have the same DNA/RNA as the rest of life on Earth, although it might use something chemically similar for the same purpose. I'm personally more interested in how the life-form in question deals with its energy supplies, something that's likely to have ramifications on the entire organism, rather than just on its data storage, even though DNA is rather more famous than ATP.
So in short: you'd expect it to be carbon-based, and many of its chemicals involved would be much the same (using glucose for respiration, etc.; although the organism in question is in such a weird environment that it may well use a different energy source entirely); but it would have to find a different (if only slightly different) way to store energy for immediate use (meaning that most of the enzymes in the body, that actually do something, would need to be different), and it would need to use something other than DNA or RNA to store its genetic code (likely a variant using arsenic rather than phosphorous).
It's actually one of the options Youtube gives to owners of copyrighted content that other people are posting to Youtube without their permission. When they catch something like that, they give the copyright owners the option to simply track the usage of their content, block the video altogether, or to have run adverts alongside any use of it and gain the resulting money. So sometimes you end up with adverts on random videos because they'd picked copyrighted background music, or something like that, and the owner of the music chose to allow the video to stay up in return for a bit of money.
Well, if a link's flagged as "we think there's a decent chance this link is irrelevant or spam and will contaminate your search results", what sensible search engine would pay attention to it? This is one of the cases where nofollow helps both the website that uses it, and the search engine that looks at it.
No, nobody pushed hard on the "they were shipping Linux" angle because the relevant court case would be SCO vs. IBM, which has been stayed pending the outcome of SCO vs. Novell (which still hasn't happened; SCO are appealing yet again). I'm sure IBM's lawyers would use it if that case ever went forwards.
I'd be very surprised if nobody had profiled the Linux kernel to look for all such speedups already. This is likely, instead, one of those speedups obtained via changing to a better algorithm, which can be even better than those made by optimising inner loops. (Speedups from optimising inner loops are generally around 30 to 50 percent unless you did something really stupid, in my experience; I've known speedups from improving algorithms to be well over 10000 percent (or 99 percent, depending on what you measure it relative to).)
The GPL lets you do anything you could ordinarily do under copyright law without a license (it doesn't try to add restrictions on top of copyright law, but rather add some specific permissions whereby you have to do something specific to benefit from the permission). FedEx, etc., aren't copying the code, but rather just delivering media; what they're doing is perfectly fine under copyright law even if, say, they were shipping a copy of Microsoft Windows instead. The ISP is technically copying the code, but I'm pretty sure that there's a (US) law meaning that ISPs aren't liable for copyright infringement if they don't inspect the contents of the packet they're transmitting. (On the other hand, the story relates to Australia, which may well have different relevant laws.)
Relevant quote from GPLv3:
To “propagate” a work means to do anything with it that, without permission, would make you directly or secondarily liable for infringement under applicable copyright law, except executing it on a computer or modifying a private copy. Propagation includes copying, distribution (with or without modification), making available to the public, and in some countries other activities as well.
The restrictions on transmitting-binary-must-also-transmit-source only apply to "conveying" the work, which the GPL defines in terms of propagation. Thus, people like FedEx or an ISP, which wouldn't be breaking copyright law via transferring anything, don't break the GPL either (not that it could do anything about it anyway, even if it were a restrictive EULA that said you couldn't do that). Telstra's case is a little more complex, and, not being a lawyer, I'm not sure if what they were doing would in general require a license or not. (If it does require a license and they aren't transmitting source, then they don't have a license, because the only possible license would be the GPL, and it doesn't license that activity. If it doesn't, they don't need to worry about the GPL because they can do it without a license.)
(As for the case of linking to someone else's copy of the source, the GPLv3 specifically allows that in the case where you're hosting the binaries on a server, and the source is also on someone's server accessible via much the same means, and you also ensure that the source remains available for as long as the binaries are, even if it's not you hosting it. The GPLv2 does not have such a clause, meaning that for GPLv2-only code, if you need a license to be able to distribute the binaries, you're going to have to host the sources in the same place if you want to comply with the letter of the license.)
SQL injections, which were apparently used, have nothing to do with the operating system the system is running on; rather, they exploit errors in (usually) custom-built applications mixing up data and code before sending it to the database (which cannot really distinguish an SQL injection from an actual command). Thus, posting this is really a bit misleading; a huge number of things are Microsoft's fault, but this is probably not one of them.
I know that last time I claimed for a ticket refund, I sent the documentation and tickets like they requested. They replied with a form letter asking me to supply the tickets. But they had them now, not me, so I couldn't reply...
At least one site (I can't remember which; it was a video-sharing site whose name starts with "v", but unfortunately there are at least two popular websites fitting those criteria) has started using video adverts with quick-time-events in (situations where you have to do such and such an action within such and such a time limit), to prevent you simply looking away until the advert has finished. My reaction has been to just ignore the site and refuse to look at its ads that way.
I visit several websites which are paid for via ads, with ad-blockers on, and get used to them. Sometimes I visit them on other people's computers, or public computers, too (in fact, I'm visiting Slashdot via a shared computer at the moment), and in those cases, I see the ads.
If adblocking didn't exist, I likely wouldn't be so interested in the websites in the first place, and so it's possible I see more ads with my adblocker, than I would without, due to different site visiting patterns. So it's not just word-of-mouth between users that's important in this situation.
I use a Nintendo DS as an alarm clock (because it's one of the few things I have to hand that I remember to keep the battery charged on...) and it woke me an hour early today (I'm in Europe). I wonder why the iPhone bug went the other way?
3. Don't optimise without profiling to see where the problem actually is.
I've needed to go past 1 and 2 before now; but I wouldn't add a micro-optimisation to code without being able to say both "this code is too slow without it because X", and "if I do this change, the code speeds up 30%" (or whatever amount). (And, of course, you have to know if 30% is actually a useful optimisation in the circumstances; if you don't know just how fast the code has to be, you have to read rules 1 and 2 again.)
Macro-optimisations, where you pick an algorithm that's better by computational orders, are less dangerous to try; changing an O(n^2) algorithm to an O(n log n) algorithm can sometimes save 10 orders of magnitude or more, depending on the problem (more commonly, it's more like 6 or 7 orders). Even then, you normally have to go through the first two rules and use the third; an O(n^2) algorithm is perfectly acceptable if n happens to be small in your situation, so you ideally want to know where the problem is before you spend effort (both yours and that of future maintainers) on a more efficient algorithm.
It wouldn't really make much sense for Microsoft to sue.NET implementors until after the implementors had no viable alternative languages left. Embrace/Extend/Extinguish (or in this case, Embrace/Sue/Extinguish) doesn't really work until you get the "Embrace" step to happen first.
Pretty much this happens in student union elections in the UK (most universities use more or less the same system, with a single-transferable-vote system and a "none of the above" option, personified as RON, the "ReOpen Nominations" candidate). If RON wins, the entire election is repeated; although the same candidates can stand again if they wish, they rarely do (as the electorate has already indicated that they'd prefer someone else to win).
It's pretty rare for RON to win an election (or indeed, come anywhere but last), but I've known it to happen on occasion. (Each time, it's caused quite a few extra candidates to run, and in my opinion at least the resulting elected candidate genuinely was better than the originals.)
I had the opposite experience with Outlook Web Access; eventually, I ended up using it exclusively with Firefox to get the degraded version, as the full version on IE was so buggy and annoying. The degraded version might have needed a few more clicks, but at least it worked and was internally consistent.
The thing that confuses me about that command is that the "repo" command isn't installed on my (stock Ubuntu 10.04) system, nor is it in the repositories. Is it a typo for something? Or some nonstandard command I don't have? (I doubt it would exist on Windows either; if it's Mac-only, it would be somewhat ironic given the topic of conversation...)
Interestingly, this is one of only two laws I was taught in school (the other being the Race Relations Act). I think it's on the official school curricula here in the UK.
If their circuitry is vulnerable to RF interference (not completely implausible for an analogue synthesiser, but still a bit unlikely), why on earth wasn't it shielded from radio signals? It's a simple case of putting the thing into a metal box...
It is indeed "orchestrae" it seems. The word does indeed come via a latin root orchestra, from the Ancient Greek word which transliterates into the English alphabet as orcheestra (/me glares at Slashdot's Unicode filter) according to several etymological dictionaries I checked, so using a Latin plural is at least vaguely appropriate (unlike in the case of words that come directly from Greek). The genitive of orchestra is orchestrae, making it a first declension feminine word; for such words, the plural is equal to the genitive, also orchestrae.
Of course, pluralising loanwords in their original language (usually Latin) seems to be something that, although common on online fora (and yes, I just looked that plural up as well just to make sure...), is hardly used in general conversation. Arguably, English long ago moved on from using the original plural, and so the true correct plural is to follow English rules, to end up with "orchestras".
Exactly. If I recall previous statements from Microsoft properly, Office and Office for Mac are based on two completely separate source trees.
This seems pretty plausible, especially due to the wide difference in bugs between them. The fact that this is the case is rather worrying, though; well-written code is nearly always easier to port by changing platform-specific parts, than by rewriting it, unless it mostly consists of optimised assembler or something inherently nonportable like that, and I don't see why Word would be. Even in the case of Mac OS X vs. Windows, where there are very different interface guidelines, you'd still expect to have a shared engine and separate interfaces. (Perhaps it's for marketing reasons? It's entirely possible that the Windows version of Word has been deliberately tied very strongly to Windows, and I recall something about undocumented APIs used by various internal teams at Microsoft to gain an advantage over other internal teams; although that wouldn't inherently mean there wasn't a reason why the Mac version couldn't be ported to Windows, Microsoft might not want to so they could maintain the tying.)
The only even remotely common one where it isn't is Mac OS Classic (i.e. pre-OSX), nowadays. (Although Windows will convert \n into \r\n on output to a textmode file, this will happen whether it's written as \015\012 or \r\n.) So you don't really gain anything by doing this. (A better method is to set the "binary mode" flag on the filehandle, e.g. by using "binmode" in Perl, in order to turn off platform-specific newline translation; this will avoid the \n to \r\n translation on Windows and not hurt on other common platforms. It wouldn't surprise me if this was the default for network sockets anyway, though.)
It's installed by default, but doesn't phone home without permission (it's opt-in, and just sits there unless you decide to let it start sending). I turned it on in my own Ubuntu installation (Ubuntu borrows a lot of packages from Debian, after all): System | Administration | Software Sources | Statistics. IIRC the collected statistics are available on Debian's website somewhere, although I don't know if that counts Ubuntu installations too.
Due to the existence of popcon, therefore, canonical-census seems like it can only be designed to be turned on by default; otherwise, there would be no reason for it to exist.
Slashdot Mirror
Symantec actually make a separate Norton Removal Tool available from their website, which allegedly (this is second-hand info, I don't know this from personal experience) actually works to uninstall Norton. (It's advertised as a tool to help you recover from problems during the installation, incidentally; I wonder if that's an attempt to justify the existence of a tool that should be redundant to the uninstaller?)
It'd still be carbon-based, by and large.
The reason people are interested here is that the main very-short-term energy store in the human body - the chemical used to store energy generated by respiration until it's needed - is in pretty much all known life adenosine triphosphate (ATP). (Because the body can't store ATP, it uses chemicals like glucose for typical short-term storage; ATP is used pretty much immediately after being generated.) The way it works is that the chemical's actually stable with just one, or more commonly, two phosphate groups; the extra phosphate groups take energy to attach, and consequently, if a reaction disattaches the groups as a side-effect, it can be driven in the "forwards" direction even if it normally wouldn't. As a result of this, a whole bunch of enzymes in a typical organism are dedicated to trying to split ATP molecules as a side-effect of apparently unrelated reactions. It's quite surprising that an organism has managed to adapt all its ATP-using enzymes to be based on something else (according to the speculation here on Slashdot, presumably an ATP variant that used arsenic rather than phosphorous), given how widespread it is.
The other really interesting use of phosphorous in typical Earth organisms is that DNA and RNA, two of the most important data-carrying chemicals in human bodies, incorporate phosphate in their construction. (They use a base-4 code with about 4 different variants for each nucleotide in the chain; there are 4 main ones, and a bunch of mostly irrelevant variants which pop up now and then; in most animals and plants, DNA is used for long-term storage, and RNA is used to make temporary copies of DNA, with DNA rarely being used directly except to make copies from. Which nucleic acid is used for what is far from universal in "typical" life in general, though.) ATP (with the two extra phosphate groups removed) actually happens to be one of the four possible major nucleotides that makes up RNA, and is used as a feedstock to synthesize RNA and DNA from. A life-form that didn't use phosphorous would be unable to have the same DNA/RNA as the rest of life on Earth, although it might use something chemically similar for the same purpose. I'm personally more interested in how the life-form in question deals with its energy supplies, something that's likely to have ramifications on the entire organism, rather than just on its data storage, even though DNA is rather more famous than ATP.
So in short: you'd expect it to be carbon-based, and many of its chemicals involved would be much the same (using glucose for respiration, etc.; although the organism in question is in such a weird environment that it may well use a different energy source entirely); but it would have to find a different (if only slightly different) way to store energy for immediate use (meaning that most of the enzymes in the body, that actually do something, would need to be different), and it would need to use something other than DNA or RNA to store its genetic code (likely a variant using arsenic rather than phosphorous).
It's actually one of the options Youtube gives to owners of copyrighted content that other people are posting to Youtube without their permission. When they catch something like that, they give the copyright owners the option to simply track the usage of their content, block the video altogether, or to have run adverts alongside any use of it and gain the resulting money. So sometimes you end up with adverts on random videos because they'd picked copyrighted background music, or something like that, and the owner of the music chose to allow the video to stay up in return for a bit of money.
Well, if a link's flagged as "we think there's a decent chance this link is irrelevant or spam and will contaminate your search results", what sensible search engine would pay attention to it? This is one of the cases where nofollow helps both the website that uses it, and the search engine that looks at it.
No, nobody pushed hard on the "they were shipping Linux" angle because the relevant court case would be SCO vs. IBM, which has been stayed pending the outcome of SCO vs. Novell (which still hasn't happened; SCO are appealing yet again). I'm sure IBM's lawyers would use it if that case ever went forwards.
I'd be very surprised if nobody had profiled the Linux kernel to look for all such speedups already. This is likely, instead, one of those speedups obtained via changing to a better algorithm, which can be even better than those made by optimising inner loops. (Speedups from optimising inner loops are generally around 30 to 50 percent unless you did something really stupid, in my experience; I've known speedups from improving algorithms to be well over 10000 percent (or 99 percent, depending on what you measure it relative to).)
The GPL lets you do anything you could ordinarily do under copyright law without a license (it doesn't try to add restrictions on top of copyright law, but rather add some specific permissions whereby you have to do something specific to benefit from the permission). FedEx, etc., aren't copying the code, but rather just delivering media; what they're doing is perfectly fine under copyright law even if, say, they were shipping a copy of Microsoft Windows instead. The ISP is technically copying the code, but I'm pretty sure that there's a (US) law meaning that ISPs aren't liable for copyright infringement if they don't inspect the contents of the packet they're transmitting. (On the other hand, the story relates to Australia, which may well have different relevant laws.)
Relevant quote from GPLv3:
The restrictions on transmitting-binary-must-also-transmit-source only apply to "conveying" the work, which the GPL defines in terms of propagation. Thus, people like FedEx or an ISP, which wouldn't be breaking copyright law via transferring anything, don't break the GPL either (not that it could do anything about it anyway, even if it were a restrictive EULA that said you couldn't do that). Telstra's case is a little more complex, and, not being a lawyer, I'm not sure if what they were doing would in general require a license or not. (If it does require a license and they aren't transmitting source, then they don't have a license, because the only possible license would be the GPL, and it doesn't license that activity. If it doesn't, they don't need to worry about the GPL because they can do it without a license.)
(As for the case of linking to someone else's copy of the source, the GPLv3 specifically allows that in the case where you're hosting the binaries on a server, and the source is also on someone's server accessible via much the same means, and you also ensure that the source remains available for as long as the binaries are, even if it's not you hosting it. The GPLv2 does not have such a clause, meaning that for GPLv2-only code, if you need a license to be able to distribute the binaries, you're going to have to host the sources in the same place if you want to comply with the letter of the license.)
SQL injections, which were apparently used, have nothing to do with the operating system the system is running on; rather, they exploit errors in (usually) custom-built applications mixing up data and code before sending it to the database (which cannot really distinguish an SQL injection from an actual command). Thus, posting this is really a bit misleading; a huge number of things are Microsoft's fault, but this is probably not one of them.
I know that last time I claimed for a ticket refund, I sent the documentation and tickets like they requested. They replied with a form letter asking me to supply the tickets. But they had them now, not me, so I couldn't reply...
At least one site (I can't remember which; it was a video-sharing site whose name starts with "v", but unfortunately there are at least two popular websites fitting those criteria) has started using video adverts with quick-time-events in (situations where you have to do such and such an action within such and such a time limit), to prevent you simply looking away until the advert has finished. My reaction has been to just ignore the site and refuse to look at its ads that way.
I visit several websites which are paid for via ads, with ad-blockers on, and get used to them. Sometimes I visit them on other people's computers, or public computers, too (in fact, I'm visiting Slashdot via a shared computer at the moment), and in those cases, I see the ads. If adblocking didn't exist, I likely wouldn't be so interested in the websites in the first place, and so it's possible I see more ads with my adblocker, than I would without, due to different site visiting patterns. So it's not just word-of-mouth between users that's important in this situation.
I use a Nintendo DS as an alarm clock (because it's one of the few things I have to hand that I remember to keep the battery charged on...) and it woke me an hour early today (I'm in Europe). I wonder why the iPhone bug went the other way?
There's a rule 3 as well, but it's not as snappy:
3. Don't optimise without profiling to see where the problem actually is.
I've needed to go past 1 and 2 before now; but I wouldn't add a micro-optimisation to code without being able to say both "this code is too slow without it because X", and "if I do this change, the code speeds up 30%" (or whatever amount). (And, of course, you have to know if 30% is actually a useful optimisation in the circumstances; if you don't know just how fast the code has to be, you have to read rules 1 and 2 again.)
Macro-optimisations, where you pick an algorithm that's better by computational orders, are less dangerous to try; changing an O(n^2) algorithm to an O(n log n) algorithm can sometimes save 10 orders of magnitude or more, depending on the problem (more commonly, it's more like 6 or 7 orders). Even then, you normally have to go through the first two rules and use the third; an O(n^2) algorithm is perfectly acceptable if n happens to be small in your situation, so you ideally want to know where the problem is before you spend effort (both yours and that of future maintainers) on a more efficient algorithm.
It wouldn't really make much sense for Microsoft to sue .NET implementors until after the implementors had no viable alternative languages left. Embrace/Extend/Extinguish (or in this case, Embrace/Sue/Extinguish) doesn't really work until you get the "Embrace" step to happen first.
Pretty much this happens in student union elections in the UK (most universities use more or less the same system, with a single-transferable-vote system and a "none of the above" option, personified as RON, the "ReOpen Nominations" candidate). If RON wins, the entire election is repeated; although the same candidates can stand again if they wish, they rarely do (as the electorate has already indicated that they'd prefer someone else to win).
It's pretty rare for RON to win an election (or indeed, come anywhere but last), but I've known it to happen on occasion. (Each time, it's caused quite a few extra candidates to run, and in my opinion at least the resulting elected candidate genuinely was better than the originals.)
I had the opposite experience with Outlook Web Access; eventually, I ended up using it exclusively with Firefox to get the degraded version, as the full version on IE was so buggy and annoying. The degraded version might have needed a few more clicks, but at least it worked and was internally consistent.
The thing that confuses me about that command is that the "repo" command isn't installed on my (stock Ubuntu 10.04) system, nor is it in the repositories. Is it a typo for something? Or some nonstandard command I don't have? (I doubt it would exist on Windows either; if it's Mac-only, it would be somewhat ironic given the topic of conversation...)
Interestingly, this is one of only two laws I was taught in school (the other being the Race Relations Act). I think it's on the official school curricula here in the UK.
If their circuitry is vulnerable to RF interference (not completely implausible for an analogue synthesiser, but still a bit unlikely), why on earth wasn't it shielded from radio signals? It's a simple case of putting the thing into a metal box...
It is indeed "orchestrae" it seems. The word does indeed come via a latin root orchestra, from the Ancient Greek word which transliterates into the English alphabet as orcheestra (/me glares at Slashdot's Unicode filter) according to several etymological dictionaries I checked, so using a Latin plural is at least vaguely appropriate (unlike in the case of words that come directly from Greek). The genitive of orchestra is orchestrae , making it a first declension feminine word; for such words, the plural is equal to the genitive, also orchestrae. Of course, pluralising loanwords in their original language (usually Latin) seems to be something that, although common on online fora (and yes, I just looked that plural up as well just to make sure...), is hardly used in general conversation. Arguably, English long ago moved on from using the original plural, and so the true correct plural is to follow English rules, to end up with "orchestras".
Annoyingly, Slashdot "fixed" that bug, in that the adjective doesn't show up at all if it contradicts the score. I preferred it the old way.
This seems pretty plausible, especially due to the wide difference in bugs between them. The fact that this is the case is rather worrying, though; well-written code is nearly always easier to port by changing platform-specific parts, than by rewriting it, unless it mostly consists of optimised assembler or something inherently nonportable like that, and I don't see why Word would be. Even in the case of Mac OS X vs. Windows, where there are very different interface guidelines, you'd still expect to have a shared engine and separate interfaces. (Perhaps it's for marketing reasons? It's entirely possible that the Windows version of Word has been deliberately tied very strongly to Windows, and I recall something about undocumented APIs used by various internal teams at Microsoft to gain an advantage over other internal teams; although that wouldn't inherently mean there wasn't a reason why the Mac version couldn't be ported to Windows, Microsoft might not want to so they could maintain the tying.)
The only even remotely common one where it isn't is Mac OS Classic (i.e. pre-OSX), nowadays. (Although Windows will convert \n into \r\n on output to a textmode file, this will happen whether it's written as \015\012 or \r\n.) So you don't really gain anything by doing this. (A better method is to set the "binary mode" flag on the filehandle, e.g. by using "binmode" in Perl, in order to turn off platform-specific newline translation; this will avoid the \n to \r\n translation on Windows and not hurt on other common platforms. It wouldn't surprise me if this was the default for network sockets anyway, though.)
As if there weren't enough reasons already, this argument probably explains ACTA.
It's installed by default, but doesn't phone home without permission (it's opt-in, and just sits there unless you decide to let it start sending). I turned it on in my own Ubuntu installation (Ubuntu borrows a lot of packages from Debian, after all): System | Administration | Software Sources | Statistics. IIRC the collected statistics are available on Debian's website somewhere, although I don't know if that counts Ubuntu installations too.
Due to the existence of popcon, therefore, canonical-census seems like it can only be designed to be turned on by default; otherwise, there would be no reason for it to exist.