When you have several gigabytes of data to process in a non-CPU-intensive manner, you can't affor to save everything to a central server. The time taken up by NFS (AFS,SMB,whatever) is too high. You want to get done today (or tommorrow, but definately not next week!). Now, you could copy the data from the server, process it, and copy the results back, but this takes extra time (and requires twice the hard disk space). Sometimes, it just makes sense to store vital data on workstations (and back them up periodicaly).
OTOH, the question here was MS Office files, and if those are this big, you've got other problems.
Right now, the majority of space on chip is taken up by verious caches. A significant proportion of that space is taken up by wiring. Having much smaller wiring should allow much larger caches. A system with 8Mb on-chip cache (and a well-designed asynchronous algorythm for filling it) would hardly ever wait for the front-side-bus at all.
I don't know if any one person can understand the entire specs for an airplane (my guess would be yes), but I'm sure no one person can understand *everything* that goes into one -- metalurgy, fuel refinement, internals of the integrated chips that run the electronics, etc.
It all depends on what you count as parts of an airplane's design
Re:Just for comparison's sake...
on
Genome Surprise
·
· Score: 4, Insightful
Not to argue with your basic idea there, but how does culture determine the similarities then? The fact that most native Africans have dark skin, most Northern Europeans are relatively fair skinned, and most Asians are notably shorter than Native Americans? There has to be some genes doing something. Or some other mechanism we have yet to discover.
Dark-skinned is not a race. "Black" is a race -- and very few black people have actual black skin (much less black teeth!). Races are cultural constructs that may include genetic/physical characteristics in their definitions. Some physical characteristics are often considered (e.g. skin color) and some (earlobe attatchment, blood type) are generally ignored. Furthermore:
USA custom dictates that an individual with one great-great-grandfather from Nigeria and the rest from England is black, completely irrespective of any alleles the individual may actually express or carry.
USA custom acknowledges "Hispanic" as a race, even though it carries no genetic correlation. It is roughly defined as an individual who's native language is an American dialect of Spanish, or the decendant of such a person to the fourth generation. Except Jamacans, who might be black. And including Haitians and Brazilians, or something like that.
Mexican custom dictates that an individual with one nigerian parent and one Castillian is black, mulatto, or white depending on the individual's net income. (Disclaimer: I haven't been to Mexico -- this is the finding of some random ethnographer.)
What's probably most significant, though, is that the races which do correspond to genetic traits make no sense as biological characterizations. They don't match actual genetic difference groups at all. This is what is meant by the statement that races are purely cultural.
I notice there are http URLs for the schemas. If they contained the specifications for the parse tree, it could be quite informative (even if it just listed them -- complete context is helpful). Sadly, they simply get 404s (though the schemas.microsoft.com server exists). I wonder if that violates some sort of standard....
We're in the process of building community, from that we can construct a government.
Sounds kinda totalitarian to me. . .
Actually, it's strangely democratic. Seriously, the vast majority of successful Open Source projects have a single maintainer. X hasn't, and some might speculate that that's part of it's problem. I guess this has to be done to attract a large number of old X developers, but I really wonder if a benevolent dictator could make things work better (and if not, just use XFree86).
Wow, It's been a long time since something comparable happened. I guess the glibc/libc split is probably the closest. That settled out reasonably quickly, (though it left some freakish version numbers that still cause trouble). I suppose one can hope for something similar here.
X development has been somewhat slow, but it seems like the really big issue has always been drivers -- is there any way that new leadership can help get specs from manufacturers?
Well, IIRC PATRIOT I didn't define terrorism, but PATRIOT II defines it as violation of a law in order to effect government policy. I can't find offhand the ACLU calling for civil disobedience, but I wouldn't be surprised if they have. I know NAACP and AFL-CIO have, so they'll be terrorist groups if/when PATRIOT II passes.
As I read the legal documents (i.e. incompetantly) it seems that N2H@ got the case dismissed on the basis that they hadn't threatened to sue him. Hopefully this means that if he proceeds with the investigation and gets sued, this case will return to haunt N2H2. I can't actually think how that would happen, though.
Technology transfer doesn't have to steamroll cultures. Japan and Thailand managed to bring in a lot of outside technology without giving up their cultures. IIRC, Germany industrialized after the rest of Europe, mostly by copying outside inovations, and they remained perfectly German.
Interestingly, these examples have a common element -- these countries maintained their political independance during the time of technology transfer. India was colonized, but never totally controlled -- their culture was strained but not broken. Nigeria was conquered utterly, and their culture is in shambles.
Certainly some aspects of a culture are vitally linked with the technology. Economic systems, for example: you can't have a purely collectivist industrialized tribe or a feudal hunter-gatherer group (nor, I suspect, a capitalist information age society, but we'll see). But these aspects can change without devastating the culture. Consider 18th century Europe, or 19th century Japan.
Other parts of culture are tied to the government. Things like language and kinship structure. It doesn't much matter for those whether you're hunting and gathering or programming computers, but you're gonna have a hard time going against your government. This is where conquest really pinches.
And then there are things like religion, which can go on despite technology transfer and conquest, so long as the conquest doesn't include missionaries or other theocratic elements.
In short, giving technology doesn't automatically destroy cultures -- it all depends on hopw you do it.
P.S. Considering the strict nature of Structural Adjustment Programs, nations with heavy IMF/WB involvement should be viewed as conquered.
If we can exterminate an entire species, are we morally allowed to do it? Well we did it (almost) with the variola virus, but you could argue if a virus is alive. We'll soon be able to do it with mosquitoes, the tse-tse fly. Those are pests, but should they be erased from the face of earth? What about rats?
Right now, the interesting obligation runs the other way -- if we can refrain from exterminating a life-form, are we obligated to do it? What if there's a monetary expense? Who pays? What if there are human lives at stake?
IIRC, the only species we've intentionally exterminated is Smallpox, and we've actually kept two samples of it. It was very difficult. OTOH, we've steamrollered (often literally) thousands of species that happened to be in our way, sometimes before we knew they were there.
It seems to me that if exterminating a species that was in the wroung place at the wrong time is accidentaly is acceptable, exterminating one that threatens us after careful consideration must be.
The article says that the orbits they'v calculated are only preliminary. I wonder if, when other gravitational fields are taken into acount, they're really closed around Jupiter. The satellites are orbiting backward, which sounds a lot like they used to orbit the sun forwards a little clsoer iun than Jupiter.
I wouldn't be surprised if they become sun-orbiting asteroids within 100 years.
I realize this is a hideously Orwellian (Heelerian?) Catch-22, but it could be pure incompitance. When the ruling first appeared, many here questioned how such a law might be enforced when surely any ISP discovering child porn should report it to police so the server can be taken down.
It seems likely to me that they simply don't have a list, and they want to make it the ISP's problems. The best law enforcement agencies in the country can't stop kiddy-porn rings, so let's see if overworked sysadmins can! If it fails, at least we'll be able to pass the blame...
I think ISPs should simply declare that, to the best of their knowledge, there is no kiddy porn on the web, and only block things if they get complaints (then report the complainant as having viewed kiddy-porn.)
If you're using high temperatures to get the magnets to have an effect, then half the protons will go clockwise and half counterclockwise and you'll get nothing (because half were going forward brownianly and half backward). It may not cancel perfectly, but close enough.
Take a box, paint the inside black,and out a glass wall down the middle. Put a pocket of vacuum inside the glass. Now neither conduction of heat nor infra-red can cross the barrier. Figure the whole thing's room temperature.
Note that each region is filled with IR black-body radiation, but it gets re-absorbed and turned back into heat and everything's in equilibrium -- and nothing crosses the barrier.
Now stick into one side of it a IR photocell wired to a capaciter which discharges once a second through and LED. Now we've got a little visible light being generated from the IR on one side.
This light can cross the barrier -- we have a one-way transfer of energy (which will promptly become heat as the visible light hits the black paint).
Now rig up a traditional thermal difference engine, and exploit!
I haven't been able to find a flaw here. My guess is that I'm misunderstanding black-body radiation, 'cause it sure seems anti-entropic to me. Maybe someone else here can get it.
Magnetic fields only cause charged particles to go in circles if there's relative motion between the particles and the magnet. If you just stick a magnet next to a loop of wire, nothing happens. You have to *move* the magnet.
Similarly, you need to inject the plasma at high velocity. This needs to be done continously, at higher energy cost than what you get out.
Seriously, the FSF originally set out to make it so that you could run a completely free system, and now you can. I do it; so do many others I know.
The FSF's task now is to make sure it remains possible -- i.e. no SSSCA, no DMCA'ed.doc format, sane or relatively impotent patents, and a legal environment in which free software feels like a safe choice to managers. When we started out, the biggest threat was actually needing something that only proprietary software offered, but that's not the big threat now.
Sure, more software needs to be written, but we are writing it. The FSF looks to secure our most vulnerable points.
P.S. Debian essentially is the long-promised GNU system. The FSF dropped out of administering it pretty early, and it uses Linux not Hurd, but it is basically the promised GNU system.
You say the responce rate for e-mails cc'd to the boss is higher than otherwise, and suggest that this contradicts what else is said here. Mught I point out that you say you use this on people who haven't answered ordinary e-mails. What is being objected to is the assumption that someone won't get back to you wihut managerial involvement. The above slashdotters find it insulting and untrue to suppose they won't do anything on their own - but there are (presumably other) people out there who actually won't, and that's whomyou'edealing with.
Be fair, Sun has contributed to GNU/Linux. Openoffice isn't lip service (I have no idea how much they paid for it, nor how much total developer-time they put in, but the result is big, and no-one maid them). Their contributions to GNOME aren't lip service either. I recall Sun recently claimed to have contributed more code to GNU/Linux than any other single for-profit company -- I think it's true.
They've decided to hold onto Solaris, at least at the high end, and they're stumbling around how to do that. They want to support GNU/Linux, especially afgainst Microsoft, without getting eaten by it (hence their support on the desktop). They may not be the world's greatest friend, but they are a significant contributer.
Budget to purchase new sun hardware, maybe not (or, at least not the expensive items that make Sun's revenue) but that doesn't mean there's no effect. I'm currently writing some utilities that might be useful to anyoe working in a GNOME environment (like Solaris 9, IIRC). Testing them on Solaris is extra work for me. Fixing any problems is a lot of extra work. If Sun's good to us, I'll probably do it. If they go spewing FUD, I'll let them take care of it.
Now, they'll still have access to the code, and I'm sure their engineers can fix any problems easily enough, but enough people fealing this way will add up. And if the general fealing becomes that Open-source code works out-of-the-box (out-of-the-download?) on GNU/Linux but needs special work on Solaris, *that* could effect purchasing decisions.
As near as I can make out from their site, SCO is claiming that Linux is derived from Minix (which is sort of true) which was derived from BSD (again, a stretch) which came from AT&T Unix (true for once) which SCO now owns. Of course, no Minix code is present in Linux, so they're not claiming actual piracy (besides, with Linux's code available, they can hardly claim secret violations!), but maybe some mystical Unix-essence.
They're probably hoping IBM will buy them out for a couple million dollars in an effort to avoid bad publicity, and then the SCO managers can retirs in luxury. The sad thing is, it might even happen
You sure about those five nines? We just slashdotted a non-pornographic image on that server. Surely even the mighty pornographers fear the slashdt effect!
OTOH, it loaded very fast, maybe we need to get porn sites to mirror articles slashdot links to.
While I'll grant that adverstising is a waste of resources on the grand scale, and probably harmful in any context, that's not the only thing publishers do. Nor do I mean boxing and pressing. The primary thing publishers do is sort through to find what's worth publishing.
I'm sure I'm not the only slashdotter whose written games for personal amusement. Nor am I the only one whose distributed them to friends and gotten positive responces (I think honestly, but they might have just been being nice to me).
Now picture all those games coming unsorted through some sort of web portal. Combined with buggy games, games which only run on an SGI mainframe, games with trojan horses, over-used joke games (thermonuclearwar, the game that just pops up a dialog saying "You lose") and downright trolls (a game built around goatse.cx).
Now, I'm not saying publishers are the only way to strain this down to something acceptable. Gaming magazines can give reviews (though less than 1% of games would ever get reviewed at all); players would have favorite game authors; there could even be something like slashdot moderation (we all know how well that works -- actually, IMHO, it's one of the better forms I've seen).
I'm just saying that publishers can't just "get out of the way" -- they can only be replaced by something better.
A lot of the ideas here are based on an essay of Eric Flint's. He expounds in detail.
.sig: We go to war with Iraq to prevent them from building nukes and using them against us, destabilizing Pakistan, allowing Al-queada to get nukes, and use them against us -- oh the irony!
Slashdot Mirror
When you have several gigabytes of data to process in a non-CPU-intensive manner, you can't affor to save everything to a central server. The time taken up by NFS (AFS,SMB,whatever) is too high. You want to get done today (or tommorrow, but definately not next week!). Now, you could copy the data from the server, process it, and copy the results back, but this takes extra time (and requires twice the hard disk space). Sometimes, it just makes sense to store vital data on workstations (and back them up periodicaly).
OTOH, the question here was MS Office files, and if those are this big, you've got other problems.
Right now, the majority of space on chip is taken up by verious caches. A significant proportion of that space is taken up by wiring. Having much smaller wiring should allow much larger caches. A system with 8Mb on-chip cache (and a well-designed asynchronous algorythm for filling it) would hardly ever wait for the front-side-bus at all.
It all depends on what you count as parts of an airplane's design
Dark-skinned is not a race. "Black" is a race -- and very few black people have actual black skin (much less black teeth!). Races are cultural constructs that may include genetic/physical characteristics in their definitions. Some physical characteristics are often considered (e.g. skin color) and some (earlobe attatchment, blood type) are generally ignored. Furthermore:
What's probably most significant, though, is that the races which do correspond to genetic traits make no sense as biological characterizations. They don't match actual genetic difference groups at all. This is what is meant by the statement that races are purely cultural.
I notice there are http URLs for the schemas. If they contained the specifications for the parse tree, it could be quite informative (even if it just listed them -- complete context is helpful). Sadly, they simply get 404s (though the schemas.microsoft.com server exists). I wonder if that violates some sort of standard....
Actually, it's strangely democratic. Seriously, the vast majority of successful Open Source projects have a single maintainer. X hasn't, and some might speculate that that's part of it's problem. I guess this has to be done to attract a large number of old X developers, but I really wonder if a benevolent dictator could make things work better (and if not, just use XFree86).
Wow, It's been a long time since something comparable happened. I guess the glibc/libc split is probably the closest. That settled out reasonably quickly, (though it left some freakish version numbers that still cause trouble). I suppose one can hope for something similar here.
/. interview?
X development has been somewhat slow, but it seems like the really big issue has always been drivers -- is there any way that new leadership can help get specs from manufacturers?
Editors: can we get Keith for a
Oh, and, FSP? (first substantive post)
Well, IIRC PATRIOT I didn't define terrorism, but PATRIOT II defines it as violation of a law in order to effect government policy. I can't find offhand the ACLU calling for civil disobedience, but I wouldn't be surprised if they have. I know NAACP and AFL-CIO have, so they'll be terrorist groups if/when PATRIOT II passes.
As I read the legal documents (i.e. incompetantly) it seems that N2H@ got the case dismissed on the basis that they hadn't threatened to sue him. Hopefully this means that if he proceeds with the investigation and gets sued, this case will return to haunt N2H2. I can't actually think how that would happen, though.
O'Reilly's Definitive Guide to OSX is being delayed by hnical difficulties: the finished book keeps collapsing into a black hole.
I mean, if a nutshell guide is over 700 pages...
Technology transfer doesn't have to steamroll cultures. Japan and Thailand managed to bring in a lot of outside technology without giving up their cultures. IIRC, Germany industrialized after the rest of Europe, mostly by copying outside inovations, and they remained perfectly German.
Interestingly, these examples have a common element -- these countries maintained their political independance during the time of technology transfer. India was colonized, but never totally controlled -- their culture was strained but not broken. Nigeria was conquered utterly, and their culture is in shambles.
Certainly some aspects of a culture are vitally linked with the technology. Economic systems, for example: you can't have a purely collectivist industrialized tribe or a feudal hunter-gatherer group (nor, I suspect, a capitalist information age society, but we'll see). But these aspects can change without devastating the culture. Consider 18th century Europe, or 19th century Japan.
Other parts of culture are tied to the government. Things like language and kinship structure. It doesn't much matter for those whether you're hunting and gathering or programming computers, but you're gonna have a hard time going against your government. This is where conquest really pinches.
And then there are things like religion, which can go on despite technology transfer and conquest, so long as the conquest doesn't include missionaries or other theocratic elements.
In short, giving technology doesn't automatically destroy cultures -- it all depends on hopw you do it.
P.S. Considering the strict nature of Structural Adjustment Programs, nations with heavy IMF/WB involvement should be viewed as conquered.
Right now, the interesting obligation runs the other way -- if we can refrain from exterminating a life-form, are we obligated to do it? What if there's a monetary expense? Who pays? What if there are human lives at stake?
IIRC, the only species we've intentionally exterminated is Smallpox, and we've actually kept two samples of it. It was very difficult. OTOH, we've steamrollered (often literally) thousands of species that happened to be in our way, sometimes before we knew they were there.
It seems to me that if exterminating a species that was in the wroung place at the wrong time is accidentaly is acceptable, exterminating one that threatens us after careful consideration must be.
The article says that the orbits they'v calculated are only preliminary. I wonder if, when other gravitational fields are taken into acount, they're really closed around Jupiter. The satellites are orbiting backward, which sounds a lot like they used to orbit the sun forwards a little clsoer iun than Jupiter.
I wouldn't be surprised if they become sun-orbiting asteroids within 100 years.
It seems likely to me that they simply don't have a list, and they want to make it the ISP's problems. The best law enforcement agencies in the country can't stop kiddy-porn rings, so let's see if overworked sysadmins can! If it fails, at least we'll be able to pass the blame...
I think ISPs should simply declare that, to the best of their knowledge, there is no kiddy porn on the web, and only block things if they get complaints (then report the complainant as having viewed kiddy-porn.)
Now q and B are constant, but v_i is random. You get clockwise motion is v_i is pointed away from the axel -- which only happens half the time.
P.S. Why doesn't slashdot support the <sub> tag?
If you're using high temperatures to get the magnets to have an effect, then half the protons will go clockwise and half counterclockwise and you'll get nothing (because half were going forward brownianly and half backward). It may not cancel perfectly, but close enough.
Take a box, paint the inside black,and out a glass wall down the middle. Put a pocket of vacuum inside the glass. Now neither conduction of heat nor infra-red can cross the barrier. Figure the whole thing's room temperature.
Note that each region is filled with IR black-body radiation, but it gets re-absorbed and turned back into heat and everything's in equilibrium -- and nothing crosses the barrier.
Now stick into one side of it a IR photocell wired to a capaciter which discharges once a second through and LED. Now we've got a little visible light being generated from the IR on one side.
This light can cross the barrier -- we have a one-way transfer of energy (which will promptly become heat as the visible light hits the black paint).
Now rig up a traditional thermal difference engine, and exploit!
I haven't been able to find a flaw here. My guess is that I'm misunderstanding black-body radiation, 'cause it sure seems anti-entropic to me. Maybe someone else here can get it.
Umm, am I missing something? This one's easy...
Magnetic fields only cause charged particles to go in circles if there's relative motion between the particles and the magnet. If you just stick a magnet next to a loop of wire, nothing happens. You have to *move* the magnet.
Similarly, you need to inject the plasma at high velocity. This needs to be done continously, at higher energy cost than what you get out.
Seriously, the FSF originally set out to make it so that you could run a completely free system, and now you can. I do it; so do many others I know.
The FSF's task now is to make sure it remains possible -- i.e. no SSSCA, no DMCA'ed .doc format, sane or relatively impotent patents, and a legal environment in which free software feels like a safe choice to managers. When we started out, the biggest threat was actually needing something that only proprietary software offered, but that's not the big threat now.
Sure, more software needs to be written, but we are writing it. The FSF looks to secure our most vulnerable points.
P.S. Debian essentially is the long-promised GNU system. The FSF dropped out of administering it pretty early, and it uses Linux not Hurd, but it is basically the promised GNU system.
You say the responce rate for e-mails cc'd to the boss is higher than otherwise, and suggest that this contradicts what else is said here. Mught I point out that you say you use this on people who haven't answered ordinary e-mails. What is being objected to is the assumption that someone won't get back to you wihut managerial involvement. The above slashdotters find it insulting and untrue to suppose they won't do anything on their own - but there are (presumably other) people out there who actually won't, and that's whomyou'edealing with.
Be fair, Sun has contributed to GNU/Linux. Openoffice isn't lip service (I have no idea how much they paid for it, nor how much total developer-time they put in, but the result is big, and no-one maid them). Their contributions to GNOME aren't lip service either. I recall Sun recently claimed to have contributed more code to GNU/Linux than any other single for-profit company -- I think it's true.
They've decided to hold onto Solaris, at least at the high end, and they're stumbling around how to do that. They want to support GNU/Linux, especially afgainst Microsoft, without getting eaten by it (hence their support on the desktop). They may not be the world's greatest friend, but they are a significant contributer.
Budget to purchase new sun hardware, maybe not (or, at least not the expensive items that make Sun's revenue) but that doesn't mean there's no effect. I'm currently writing some utilities that might be useful to anyoe working in a GNOME environment (like Solaris 9, IIRC). Testing them on Solaris is extra work for me. Fixing any problems is a lot of extra work. If Sun's good to us, I'll probably do it. If they go spewing FUD, I'll let them take care of it.
Now, they'll still have access to the code, and I'm sure their engineers can fix any problems easily enough, but enough people fealing this way will add up. And if the general fealing becomes that Open-source code works out-of-the-box (out-of-the-download?) on GNU/Linux but needs special work on Solaris, *that* could effect purchasing decisions.
They're probably hoping IBM will buy them out for a couple million dollars in an effort to avoid bad publicity, and then the SCO managers can retirs in luxury. The sad thing is, it might even happen
OTOH, it loaded very fast, maybe we need to get porn sites to mirror articles slashdot links to.
I'm sure I'm not the only slashdotter whose written games for personal amusement. Nor am I the only one whose distributed them to friends and gotten positive responces (I think honestly, but they might have just been being nice to me).
Now picture all those games coming unsorted through some sort of web portal. Combined with buggy games, games which only run on an SGI mainframe, games with trojan horses, over-used joke games (thermonuclearwar, the game that just pops up a dialog saying "You lose") and downright trolls (a game built around goatse.cx).
Now, I'm not saying publishers are the only way to strain this down to something acceptable. Gaming magazines can give reviews (though less than 1% of games would ever get reviewed at all); players would have favorite game authors; there could even be something like slashdot moderation (we all know how well that works -- actually, IMHO, it's one of the better forms I've seen).
I'm just saying that publishers can't just "get out of the way" -- they can only be replaced by something better.
A lot of the ideas here are based on an essay of Eric Flint's. He expounds in detail.
.sig: We go to war with Iraq to prevent them from building nukes and using them against us, destabilizing Pakistan, allowing Al-queada to get nukes, and use them against us -- oh the irony!