One thing you left out (or, at least, under-emphasized): if you push your changes back upstream, then you're no longer the sole maintainer of your fork, and others likely (almost certainly) can and will help maintain and enhance and fix the bugs. That's always a big win.
The flip side of this (something I've actually encountered) is when your add-ons are so specific to your particular company that upstream isn't interested in accepting your patches.
I've gotten paid, here-and-there, to send patches upstream to various free/libre/open-source projects since the mid eighties. No company I've worked for has ever regretted such actions, but one, as I mentioned above, once regretted their inability to perform such an action.
closing the tab in firefox 3.0.1 on Ubuntu 8.04 works for me.
On Ubuntu (and presumably other X11-based systems), the "middle-click clipboard" (which is what I use 90% of the time) is completely unaffected. Only the annoying ^X/^C clipboard, which I almost never use, is affected. I cut-and-pasted the above quote even though ^V currently gives me: http://www.evil.com./
I could probably be hit with the real version of this thing and not notice for upwards of a week.:)
(I'll bet money that there's no rebooting required to clean this out either. At the very worst, restarting the X server should be more than adequate.)
Right now we can ONLY reliably spot "planets" in systems with larger than Jupiter gas giants in relatively close orbits to their parent star. Given that, we have no way of identifying probability of inward migration of gas giants,
No, but we can say something about what percentage of star systems have gas giants that have migrated inwards (number of star systems examined divided by number of star systems with easy-to-detect inner system gas giants). If there are far more close-in giants than our current theories allow, percentage-wise, then we know that our current theories are incorrect.
Ease-of-detection does not bias us here. All it does is make it easier to confirm or deny our hypotheses.
We still don't know what percentage of gas giants migrate inwards, but we know that more (in absolute numbers) migrate inwards than we thought would be the case. And we know that this means that the likelihood of Earth-like planets in the habitable zone is lower than we thought.
If 60% of the stars we examine (pulling a number out of my ass) have close-in gas giants, than we can already say that at least 60% of the stars won't have habitable planets, even though we can't reliably detect habitable planets yet.
(Of course, this all assumes that close-in giants are all migrants, but that seems like fairly solid reasoning itself.)
I was thinking about the very problem of trusting your compiler, and the only thing I could come up with is building one from an open assembler.
I built gcc (1.4) with a C interpreter. It was slow as hell (and we did it mainly to stress-test the interpreter), but when I fed the source of gcc to the result, it did what I expected--built a system the same as the one that regular gcc built.
But the simple fact of the matter is that a little common sense should reveal that the whole notion is impossible in the real world. At the time Thompson wrote, there was, basically, one C compiler and one version of login, and neither one changed very much, so it was at least theoretically possible for a fairly simple program to recognize them. The sources to gcc have changed too much over the years to be recognizable to anything less than a hard-AI system, i.e., something that doesn't exist (and if it did, you'd notice, since it would take hours to compile even the simplest app). Toss in drastically different compilers from vendors like Sun, IBM, Intel and HP, and the whole thing becomes even more ridiculous. But if you really want to check, write your compiler in another language (one that doesn't compile to assembler, like Java or Python).
My WinXP never crashes--and neither does my flying car!:)
On the other hand, your claim that "NT never had stability issues" is pure nonsense. I stress-tested dozens of NT systems to destruction, and it wasn't because of hardware problems. Get the load high enough, and they simply died. If all you do is play the occasional video game and write the occasional memo, it probably works well enough, but put it in a situation where it has to do some real work, and it fell apart like a girl scout in a minefield.
XP might be more solid than NT. I don't know and don't care. Systems with auto-raise and click-to-focus are unusable IMO.:)
I don't know why people feel the need to throw in an extra "i", but "viri", at least, is typical of geek humor. Linguistically, it's arguably better than "octopi", which has made it into the dictionary even though it's an etymological nightmare (should be "octopodes"). My mom, who is a language hacker (i.e. writer), not a computer geek, independently coined the term "viri"—along with a lot of other misapplied-but-humorous pluralizations—many years before it appeared in mainstream hacker culture. It may be sad to see people who don't know better, but that doesn't mean it's not funny, and that's why I think Tom Christiansen is an ass.:)
I think the point is that anything that needs judges is not a sport
And if it's not a sport, it's not interesting? What an odd thing to suggest on Slashdot of all places.:)
Anyway, the second-most idiotic thing about Taco's statement (after the insane sexism that makes me want to kick his scrawny little ass around the block--hey, maybe we can make that an Olympic sport!) is the suggestion that I might only watch a sporting event because I care who wins. Apparently, I'm not allowed to appreciate the sheer athletic ability and prowess and maybe even artistry of the competitors and enjoy the event even when (not if) I disagree with the judges.
At least he stopped short of suggesting that the competitors I'll want to watch are those from my own country. That kind of makes-me-want-to-rip-their-throats-out jingoism is best left to the TV broadcasters.:)
You're missing the point. This isn't about merely discovering random facts. Yes, it will be nice to know the facts, no matter what, but science is more than a random collection of unanalyzed facts. Some results will do more than merely give us another random truth to add to our collection; some results will allow us to falsify certain theories and not waste time on them any more, which is better than a result that leaves us just as confused as we are now.
And in response to Nutria, who also commented: you have it exactly backwards. A result which eliminates more theories is a better result from a scientific POV. If this were about scientists clinging to their pet theories, then a result which left more theories open would be better (since it would allow more scientists to cling to their favorites), but that's pretty much the opposite of what JohnFluxx was suggesting.
Apologies if I'm overlooking an Academia inside joke there, but merely covering someone's song (especially if that someone is a good friend) is not usually considered "plagiarism". Ferexample, George Harrison may have been convicted of plagiarism for "My Sweet Lord", but John Lennon got off scott free for this.
For some perspective on the other side, the site Hanzi Smatter (run by a friend of the owner of engrish.com) has a great collection of equally high-quality use of Chinese (and Japanese and Korean) by westerners. The best part is that westerners really seem to like to use Hanzi/Kanji in tattoos; the result is a bit harder to fix than a gaffe in a manual or a sign.:)
I'm pretty sure "Midori" is an internal code name here, like "Longhorn", so that shouldn't be a problem, as long as it's not an official marketing brand. Not the smartest code name they could have picked, perhaps, since they may have to be a little more careful in how they use the term than they would have otherwise, but not really a problem.
Is Transmeta even around any more? I kind of lost track after Linus left.
btw, Firefox keeps telling me that "movie" and "movies" does not exist in the English UK dictionary, why is that? (I am not an native English speaker... fwiw)
I believe (though I'm willing to be corrected on this) that the UK standard is "film", and that "movie" is considered a barbaric Americanism. This is why (for example) Wikipedia uses "film" for its categories--it's an acceptable term pretty-much everywhere.
Seems a little odd to me since "film" is merely the medium and is used for static images as well--"movie" seems like a far more appropriate term for these moving images--but as an American, I'm hardly in a position to question the oddities of other people's usages.:)
(I'm just glad that my country is on the winning side in the balance of trade for "u"s.):)
A possible solution is to start seeding masses of innocent data so that they simply cannot identify the music sharers from anyone else.
You're assuming that the "music sharers" are copyright violators. In fact, over at Etree, they're already sharing terabytes of perfectly legal music, much of it from fairly big names that happen to allow taping/trading (the rest is from small obscure bands that also allow taping/trading). Etree is probably the largest single fully-legal torrent site in the world. Jon Hart, who is famously suing Comcast for their interference with his torrents is the main west-coast taper for one of the taper-friendly bands, and Etree is where he ran them.
Your idea of seeding random data is interesting, but I think it's going to be hard to get more than a tiny number of people to participate. Etree, on the other hand, already has huge amounts of data, and, while it's dominated by hippie jam-band junk that most people won't be interested in, it has enough other styles of music there that a whole lot of people can probably find something of interest (and you can filter out the grateful dead and phish--I have the already-filtered page bookmarked). So if you want to help drive up the amount of legal torrent traffic in the world, looking for something that might interest you on Etree is probably a good place to start.
(Note that many of the bands which allow taping and trading do so only under non-commercial terms, so Etree is, of necessity, completely ad-free.)
It's fine for government to change the rules when they need improving. Retroactive change, however, is not fine, and that appears to be what the PTO is attempting to do.
No, that's what the Supreme Court already did! The PTO is simply trying to find the best way of coping with recent Supreme Court decisions like KSR International Co. v. Teleflex Inc. and Microsoft v. AT&T (perhaps suprisingly to a lot of slashdotters, Microsoft were the Good Guys in that latter), and Quanta v. LG Electronics.
If the blame lies anywhere for the various messes that will inevitably ensue from all this, it is with the Federal Circuit which greatly overstepped its bounds by misinterpreting patent law in favor of those with patents.
It becomes really tough when you have to guess how government will change the rules for the past.
According to the Supremes, the rules aren't changing for the past, they're simply being interpreted and enforced correctly now, which they weren't before. It's true that some will be hurt by all of this, but the alternatives are equally ugly, e.g. selective enforcement of existing laws, and a get-out-of-jail-free card for those who have been trying to take advantage of apparent loopholes in the law which never actually existed in the first place.
And the people who end up footing the bill? The citizen - the one who always ends up paying.
No more so than the way we've been paying for patent trolls and other patent abusers. We ended up in a situation with no easy way out, so we might as well take the right way out, which should cause less pain in the long run.
What ever the differing beliefs of the churches of vi and emacs, most of them (except a few old curmudgeons) use bash, and in bash, two ^Ws is sufficient to delete "web 2.0".
Most people hate annoying ads. Few people object to informative, well-targeted ads that tell them about products they actually want. I hate annoying ads, but I bought something from a random ad on a random website just last week. It was an interesting product and the ad was well-targeted (science fiction themed t-shirts on a fantasy webcomic) and clever and unobtrusive. It was a very small vendor (one man shop) with a interesting product (I had to restrain myself from buying his entire line) which I probably would have never known about if not for Internet advertising.
The comic artist was happy because I clicked on his ad, the vendor/artist was happy because I bought his product, and I was happy because I got a kick-ass product. Seems like a win-win-win to me!:)
Frankly, I and almost everyone I know frequently goes looking specifically for ads, e.g., when I want take-out, I check the ads in the Yellow Pages. I do use flashblock and I usually turn off image animation, but I absolutely do not use adblock nor do I have any interest in doing so. As a result, I have much better t-shirts than you do!:p;)
(And no, I'm not going to tell you the site. I have to put up with a lot of annoying ads to spot the few gems, and I see no reason why you should benefit from my suffering. If you're going to cop a pompous, arrogant, know-it-all attitude about how you're too good to be subjected to any ads ever, you're just going to have to live with the consequences.)
I disagree! I learned $crusty_old_language back when computers had six bits and were lucky to get 'em, and I see no reason why anyone else should benefit from the research and development that's taken place in the last forty years. A pox on all these $trendy_new_language advocates who want to replace my ancient $crusty_old_language-based systems with something that's actually readable and maintainable! I say, force the kids to start with $crusty_old_language like I did, and hopefully most of them will be discouraged from entering the field and competing with my out-of-date skills, and the few that actually manage to persevere will be so traumatized by the pain of learning $crusty_old_language that they'll be completely insane and will want to help me maintain my stale, fragile, unreadable pride-and-joy. Now, where'd I put that deck of punch cards?...
You must be kidding! Surely Slashdot's high-quality editorial and fact-checking department wouldn't have let a mistake of this magnitude through to hit the front page, would they?
If the designer goes through the trouble of using a liquid cooler, it is probably safe to say that conduction only heat removal is not sufficient.
Not by the time it's at full operating temperature, but it presumably takes a non-finite amount of time to warm up, during which time, the metal in the sink will be melting. Furthermore, the metal will be carrying heat away even before it melts, so it's not like there won't be any cooling until all the metal is liquid. As the temperature of the CPU rises, so will the cooling ability of the heat sink. As I said previously, it's an engineering challenge, but I doubt it's a serious obstacle--even if it's not the route they're taking.
I'm actually more worried about the magnetic pump. I'm not sure what the effect of having a small-but-powerful electromagnet right next to your CPU would be, but I can't imagine that it's a good thing, and that's going to be an issue no matter what the melting temperature of the metal coolant.
The comic won a Hugo award, and I believe it may be the only comic ever to have done so. At the least, it's one of a very elite few. In other words, definitely worth reading.
Slashdot Mirror
One thing you left out (or, at least, under-emphasized): if you push your changes back upstream, then you're no longer the sole maintainer of your fork, and others likely (almost certainly) can and will help maintain and enhance and fix the bugs. That's always a big win.
The flip side of this (something I've actually encountered) is when your add-ons are so specific to your particular company that upstream isn't interested in accepting your patches.
I've gotten paid, here-and-there, to send patches upstream to various free/libre/open-source projects since the mid eighties. No company I've worked for has ever regretted such actions, but one, as I mentioned above, once regretted their inability to perform such an action.
That's easy to address if you just assume the creator was a turtle: then it's turtles all the way up! :)
I'll just link to the last time I addressed this issue.
Executive summary: yes, I checked, but I already knew I didn't need to: Thompson's hack won't work any more.
I have people I'd trust with every bit of credentials I have, yet I don't give it up to them either.
I'd say that's almost tautological. The people I'm likely to trust with my passwords are, almost by definition, those who don't need them. :)
closing the tab in firefox 3.0.1 on Ubuntu 8.04 works for me.
On Ubuntu (and presumably other X11-based systems), the "middle-click clipboard" (which is what I use 90% of the time) is completely unaffected. Only the annoying ^X/^C clipboard, which I almost never use, is affected. I cut-and-pasted the above quote even though ^V currently gives me: http://www.evil.com./
I could probably be hit with the real version of this thing and not notice for upwards of a week. :)
(I'll bet money that there's no rebooting required to clean this out either. At the very worst, restarting the X server should be more than adequate.)
All of the windows mobile phones i've used had terrible interfaces that were more suited to a PDA than a phone...
All the Windows Mobile PDAs I've used had terrible interfaces that were more suited to a desktop than a PDA.
All the Windows desktop systems I've used had terrible interfaces. Period. :)
Right now we can ONLY reliably spot "planets" in systems with larger than Jupiter gas giants in relatively close orbits to their parent star. Given that, we have no way of identifying probability of inward migration of gas giants,
No, but we can say something about what percentage of star systems have gas giants that have migrated inwards (number of star systems examined divided by number of star systems with easy-to-detect inner system gas giants). If there are far more close-in giants than our current theories allow, percentage-wise, then we know that our current theories are incorrect.
Ease-of-detection does not bias us here. All it does is make it easier to confirm or deny our hypotheses.
We still don't know what percentage of gas giants migrate inwards, but we know that more (in absolute numbers) migrate inwards than we thought would be the case. And we know that this means that the likelihood of Earth-like planets in the habitable zone is lower than we thought.
If 60% of the stars we examine (pulling a number out of my ass) have close-in gas giants, than we can already say that at least 60% of the stars won't have habitable planets, even though we can't reliably detect habitable planets yet.
(Of course, this all assumes that close-in giants are all migrants, but that seems like fairly solid reasoning itself.)
I was thinking about the very problem of trusting your compiler, and the only thing I could come up with is building one from an open assembler.
I built gcc (1.4) with a C interpreter. It was slow as hell (and we did it mainly to stress-test the interpreter), but when I fed the source of gcc to the result, it did what I expected--built a system the same as the one that regular gcc built.
But the simple fact of the matter is that a little common sense should reveal that the whole notion is impossible in the real world. At the time Thompson wrote, there was, basically, one C compiler and one version of login, and neither one changed very much, so it was at least theoretically possible for a fairly simple program to recognize them. The sources to gcc have changed too much over the years to be recognizable to anything less than a hard-AI system, i.e., something that doesn't exist (and if it did, you'd notice, since it would take hours to compile even the simplest app). Toss in drastically different compilers from vendors like Sun, IBM, Intel and HP, and the whole thing becomes even more ridiculous. But if you really want to check, write your compiler in another language (one that doesn't compile to assembler, like Java or Python).
My WinXP never crashes--and neither does my flying car! :)
On the other hand, your claim that "NT never had stability issues" is pure nonsense. I stress-tested dozens of NT systems to destruction, and it wasn't because of hardware problems. Get the load high enough, and they simply died. If all you do is play the occasional video game and write the occasional memo, it probably works well enough, but put it in a situation where it has to do some real work, and it fell apart like a girl scout in a minefield.
XP might be more solid than NT. I don't know and don't care. Systems with auto-raise and click-to-focus are unusable IMO. :)
Sure there is: it's the plural of virius. :)
I don't know why people feel the need to throw in an extra "i", but "viri", at least, is typical of geek humor. Linguistically, it's arguably better than "octopi", which has made it into the dictionary even though it's an etymological nightmare (should be "octopodes"). My mom, who is a language hacker (i.e. writer), not a computer geek, independently coined the term "viri"—along with a lot of other misapplied-but-humorous pluralizations—many years before it appeared in mainstream hacker culture. It may be sad to see people who don't know better, but that doesn't mean it's not funny, and that's why I think Tom Christiansen is an ass. :)
I think the point is that anything that needs judges is not a sport
And if it's not a sport, it's not interesting? What an odd thing to suggest on Slashdot of all places. :)
Anyway, the second-most idiotic thing about Taco's statement (after the insane sexism that makes me want to kick his scrawny little ass around the block--hey, maybe we can make that an Olympic sport!) is the suggestion that I might only watch a sporting event because I care who wins. Apparently, I'm not allowed to appreciate the sheer athletic ability and prowess and maybe even artistry of the competitors and enjoy the event even when (not if) I disagree with the judges.
At least he stopped short of suggesting that the competitors I'll want to watch are those from my own country. That kind of makes-me-want-to-rip-their-throats-out jingoism is best left to the TV broadcasters. :)
You're missing the point. This isn't about merely discovering random facts. Yes, it will be nice to know the facts, no matter what, but science is more than a random collection of unanalyzed facts. Some results will do more than merely give us another random truth to add to our collection; some results will allow us to falsify certain theories and not waste time on them any more, which is better than a result that leaves us just as confused as we are now.
And in response to Nutria, who also commented: you have it exactly backwards. A result which eliminates more theories is a better result from a scientific POV. If this were about scientists clinging to their pet theories, then a result which left more theories open would be better (since it would allow more scientists to cling to their favorites), but that's pretty much the opposite of what JohnFluxx was suggesting.
Apologies if I'm overlooking an Academia inside joke there, but merely covering someone's song (especially if that someone is a good friend) is not usually considered "plagiarism". Ferexample, George Harrison may have been convicted of plagiarism for "My Sweet Lord", but John Lennon got off scott free for this.
For some perspective on the other side, the site Hanzi Smatter (run by a friend of the owner of engrish.com) has a great collection of equally high-quality use of Chinese (and Japanese and Korean) by westerners. The best part is that westerners really seem to like to use Hanzi/Kanji in tattoos; the result is a bit harder to fix than a gaffe in a manual or a sign. :)
I'm pretty sure "Midori" is an internal code name here, like "Longhorn", so that shouldn't be a problem, as long as it's not an official marketing brand. Not the smartest code name they could have picked, perhaps, since they may have to be a little more careful in how they use the term than they would have otherwise, but not really a problem.
Is Transmeta even around any more? I kind of lost track after Linus left.
btw, Firefox keeps telling me that "movie" and "movies" does not exist in the English UK dictionary, why is that? (I am not an native English speaker... fwiw)
I believe (though I'm willing to be corrected on this) that the UK standard is "film", and that "movie" is considered a barbaric Americanism. This is why (for example) Wikipedia uses "film" for its categories--it's an acceptable term pretty-much everywhere.
Seems a little odd to me since "film" is merely the medium and is used for static images as well--"movie" seems like a far more appropriate term for these moving images--but as an American, I'm hardly in a position to question the oddities of other people's usages. :)
(I'm just glad that my country is on the winning side in the balance of trade for "u"s.) :)
If only we could persuade some of the botnets to launch a huge DDoS attack against him! :)
A possible solution is to start seeding masses of innocent data so that they simply cannot identify the music sharers from anyone else.
You're assuming that the "music sharers" are copyright violators. In fact, over at Etree, they're already sharing terabytes of perfectly legal music, much of it from fairly big names that happen to allow taping/trading (the rest is from small obscure bands that also allow taping/trading). Etree is probably the largest single fully-legal torrent site in the world. Jon Hart, who is famously suing Comcast for their interference with his torrents is the main west-coast taper for one of the taper-friendly bands, and Etree is where he ran them.
Your idea of seeding random data is interesting, but I think it's going to be hard to get more than a tiny number of people to participate. Etree, on the other hand, already has huge amounts of data, and, while it's dominated by hippie jam-band junk that most people won't be interested in, it has enough other styles of music there that a whole lot of people can probably find something of interest (and you can filter out the grateful dead and phish--I have the already-filtered page bookmarked). So if you want to help drive up the amount of legal torrent traffic in the world, looking for something that might interest you on Etree is probably a good place to start.
(Note that many of the bands which allow taping and trading do so only under non-commercial terms, so Etree is, of necessity, completely ad-free.)
It's fine for government to change the rules when they need improving. Retroactive change, however, is not fine, and that appears to be what the PTO is attempting to do.
No, that's what the Supreme Court already did! The PTO is simply trying to find the best way of coping with recent Supreme Court decisions like KSR International Co. v. Teleflex Inc. and Microsoft v. AT&T (perhaps suprisingly to a lot of slashdotters, Microsoft were the Good Guys in that latter), and Quanta v. LG Electronics.
If the blame lies anywhere for the various messes that will inevitably ensue from all this, it is with the Federal Circuit which greatly overstepped its bounds by misinterpreting patent law in favor of those with patents.
It becomes really tough when you have to guess how government will change the rules for the past.
According to the Supremes, the rules aren't changing for the past, they're simply being interpreted and enforced correctly now, which they weren't before. It's true that some will be hurt by all of this, but the alternatives are equally ugly, e.g. selective enforcement of existing laws, and a get-out-of-jail-free card for those who have been trying to take advantage of apparent loopholes in the law which never actually existed in the first place.
And the people who end up footing the bill? The citizen - the one who always ends up paying.
No more so than the way we've been paying for patent trolls and other patent abusers. We ended up in a situation with no easy way out, so we might as well take the right way out, which should cause less pain in the long run.
What ever the differing beliefs of the churches of vi and emacs, most of them (except a few old curmudgeons) use bash, and in bash, two ^Ws is sufficient to delete "web 2.0".
Uh yes, most people do hate ads
Most people hate annoying ads. Few people object to informative, well-targeted ads that tell them about products they actually want. I hate annoying ads, but I bought something from a random ad on a random website just last week. It was an interesting product and the ad was well-targeted (science fiction themed t-shirts on a fantasy webcomic) and clever and unobtrusive. It was a very small vendor (one man shop) with a interesting product (I had to restrain myself from buying his entire line) which I probably would have never known about if not for Internet advertising.
The comic artist was happy because I clicked on his ad, the vendor/artist was happy because I bought his product, and I was happy because I got a kick-ass product. Seems like a win-win-win to me! :)
Frankly, I and almost everyone I know frequently goes looking specifically for ads, e.g., when I want take-out, I check the ads in the Yellow Pages. I do use flashblock and I usually turn off image animation, but I absolutely do not use adblock nor do I have any interest in doing so. As a result, I have much better t-shirts than you do! :p ;)
(And no, I'm not going to tell you the site. I have to put up with a lot of annoying ads to spot the few gems, and I see no reason why you should benefit from my suffering. If you're going to cop a pompous, arrogant, know-it-all attitude about how you're too good to be subjected to any ads ever, you're just going to have to live with the consequences.)
I disagree! I learned $crusty_old_language back when computers had six bits and were lucky to get 'em, and I see no reason why anyone else should benefit from the research and development that's taken place in the last forty years. A pox on all these $trendy_new_language advocates who want to replace my ancient $crusty_old_language-based systems with something that's actually readable and maintainable! I say, force the kids to start with $crusty_old_language like I did, and hopefully most of them will be discouraged from entering the field and competing with my out-of-date skills, and the few that actually manage to persevere will be so traumatized by the pain of learning $crusty_old_language that they'll be completely insane and will want to help me maintain my stale, fragile, unreadable pride-and-joy. Now, where'd I put that deck of punch cards?...
You must be kidding! Surely Slashdot's high-quality editorial and fact-checking department wouldn't have let a mistake of this magnitude through to hit the front page, would they?
If the designer goes through the trouble of using a liquid cooler, it is probably safe to say that conduction only heat removal is not sufficient.
Not by the time it's at full operating temperature, but it presumably takes a non-finite amount of time to warm up, during which time, the metal in the sink will be melting. Furthermore, the metal will be carrying heat away even before it melts, so it's not like there won't be any cooling until all the metal is liquid. As the temperature of the CPU rises, so will the cooling ability of the heat sink. As I said previously, it's an engineering challenge, but I doubt it's a serious obstacle--even if it's not the route they're taking.
I'm actually more worried about the magnetic pump. I'm not sure what the effect of having a small-but-powerful electromagnet right next to your CPU would be, but I can't imagine that it's a good thing, and that's going to be an issue no matter what the melting temperature of the metal coolant.
The comic won a Hugo award, and I believe it may be the only comic ever to have done so. At the least, it's one of a very elite few. In other words, definitely worth reading.