wich is his point actually: just like linux is only a kernel and NOT an OS, when you say that "linux" runs on X device you say that an OS based on the linux kernel runs on several devices, wich is the same than saying that Mac OS X is BSD/Mach.
so gp is spot-on, and you are actually saying the same thing.
long: it depends. are they packing software written by other people? they are not developing anytyhing, so there's nothing to share. Are they runnning their own software over a linux system? There's no reasson to asume that this google developed software constitutes a derivative work and has to be GPLed, you can put propietary non-gpled software in a distro, like the binary blobs that go into ubuntu. Are they introducing changes to the underlying, already GPLed code? then yes, they would have to make source and rights available IF they distribute it.
Distro making is just the art of putting together software written by other people, GPLed or not, and passing it along. IF you happen to develope something, AND if you happen to use GPLed work as a foundation, making your work a derivative, then, in that particular case, yes: you have to share just like when you have to share any other GPL-derivative work.
but there's a big gaping hole in this particular application to the prisoners dilemma: there is NEED.
if i NEED to get this piece of software developed, i just can't risk the possibility of it not getting funded and then dropped. If i really need it, i will try to make sure that it stays afloat, and that introduces a bigger incentive than the pure free riding impulse to desert all other companies and hope someone cooperates and gets exploited by desertors.
In real life, the prisoners dilemma takes a much more complicated form as the possibilities of interaction are much more varied than the two prototypical cooperate/desert, and the incentive and payouts structure is much much much more finegrained and open to absolutes, wich finally transform the whole rational decission making process altogether.
well, that's the thing, precisely. His statement points to the negative aspect of FS on this particular thing, namely, the very real possibility that a project gets abandoned, instead of pointing out the even more real and evident positive aspect of FS in this very same aspect, namely the posibility of picking up support *even when and if the project gets abandoned*.
Putting it as he put it, it sounds like FUD. even if he personally didn't intended it so. And since this statements are probably sanctioned by MS PR, that accusation is not held against him personally: we are not assuming that MS employees are personally evil, but as long as they speak using their MS credentials, its MS speaking. and we do believe that MS, as an organization, is evil.
but the whole point is that in closed source scenarios, the likelihood of that auditing is even less. and the probability of a bug creeping thru the audit decreases as the number of simultaneous auditors increase. and to top it off, the auditing depends on a concerted decision to divert scarce resources to auditing. In open source development, the problem is out in the open, and that increases the number of people (potentially) trying to solve it.
the argument is still correct: the probability of arriving to a satisfactory solution increases as the number of people trying to solve it increases, not even accounting for the talent of the people involved, and this is *obvious*, its a pure mathematical truth, spurious counter examples not-withstanding.
and this is a perfect example. Fermilab staff try to break it for a year without getting nowhere, and after the problem goes "open-source", it is solved in one week.
(who cares anyway? OSS is good only as it is a prerequisite for FS. its technical atributes are irrelevant)
"slavery" is precisely defined as the act of giving away self-ownership, in a manner such that you lose power over the decission (ie: you cannot revert it, since you gave yourself away to someone else), and it's illegal in any legal system based on a system of personal rights.
"voluntary and time-limited" is not slavery, its just voluntary work. Calling voluntary work slavery is at least wrong (and probably fascist, racist, etc. although i couldn't care less).
And to GP: the criminal status of any activity is subject to several considerations, one of which can be consent. Normally, the consent is disregarded only in crimes against life, and highly considered in sexual crimes, where the "crime" is by definition the lack of consent in the consumation of the act deemed criminal.
In england in particular, consented assault is only criminal when the resulting injuries are extremely severe, and you can prove that this injuries escape some pre-arranged form of consent (like the limits imposed by the rules of a game).
so OP is correct: having engaged in violent consensual activity wouldn't (in most cases) be a crime, and you can not derive its criminal status simply from a photograph.
actually, the us already had its hand in paraguay AND bolivia when it was believed that there was oil in the chaco region, shared by the two countries.
this eventauly ended up in a war between paraguay and bolivia, with the US based Standard Oil supporting Bolivia, and the UK based Shell Oil supporting Paraguay.
This was the worst armed conflict in latin america in the 20th century. over 35.000 died. and of course, there was no fucking oil whatsoever.
and down here in the south, our fricking exports are paid in dollars and we are getting screwed big time. now our central bank has come out to buy dollars to keep the price at some reasonable price, and we all share the love paying for the comercial imbalances of the us. yay!
GP is bulshit. US dollar is at its lowest since 1995.
Apart from the reliability issues already put up by several of my fellow/.ers, there's another problem related to the different standard of proof required by different legal actions. ie, you requiere lesser standards of proof for a civil case, where in theory you are not violating the liberties of anyone, than in a criminal case, where the outcome might strip someone of very basic rights.
In this context, you need totally clean, perfect and irrefutable evidence to convict of piracy, than the one you need to prevent a private third party of violating your liberties by spying on you.
Now, the above has two twists:
1.- The possitions can be reversed, mediasentry has rights of their own, although juridic personalities (or whatever you call organizations) don't have the same rights as natural personalities (or whatever you call people).
2.- The above is nothing more than a more legalesse formulation of the argument about the reliability of IP addresses as proof. That's what standards of proof are all about.
The point is that you can have it in such a way that IP based evidence is inadmisible to prove copyright infringment, but admissible to proove C&D infringment. And this is no doulbe standard, and we are still much muhc much better than them.
ps: IANAL. pps: I'm not from the USA, so things might be different in your legal system... but AFAIK, that's what the "reasonable doubt" thingy you use is about.
because archive != packages and archive FORMATS != package SYSTEMS.
tar, bz2, rar, zip, etc are compression/archiving *formats*, to put a bunch of files and lump them together in a (probably compressed) big file, or set of files. to use it, you just need a program that can read and extract the files.
A package is much more than that: it contains the binary files that make up some piece of software, and is in this sense an "archive", but the important part of the packaging is the dependencies management. Any package manager acomplish its function because every package references other packages that it depends upon, and keeps track of this so you don't have to worry about satisfying dependencies and cleaning unneeded garbage.
This structure of dependencies is called a repository, and is the core and soul of a distribution (at least package-based ones), and what it makes it superior or inferior. That's what "debian-based" or "red-hat-based" stands for: based *on the repos* of debian or red-hat.
And of course, there are structural differences in the ways that distros are built, and this affects wich files and wich pieces of software end up in a given package, etc etc.
So, although you actually can "read" rpm with apt and vice versa (there are lots of tools for this), the whole point of having package managers in the first place makes this exercise futile, as you would still have to satisfy the dependencies, by hand.
it's not that the packages are locked into the managers. is that the managers "manage" packages with references to a repository, and this repositories are distro-based, to suit the needs of a specific user base.
This is of course, a Good Thing (tm), as it encourages innovation and diversity, and it doesn't present that many problems, as the use cases in which it turns out problematic (ie, needing an rpm-only program in a deb-based system) is very rare (eg, i remember only two times in my years with *buntu).
and even in this case, compiling is really not that hard...
so basically, we have to agree on a package management system, and we are set...
and we are half the way there, as we already have one, two or three mainstream package formats, wich can be used for a whole range of distros. call it whatever you like, in the end it all comes down to the package format. deb or rpm?
that bug's related to atieventsd. If you disable the daemon from loading, it wont corrupt your X server on logout. You will have to reboot, though. Disabling the daemon once it has already lanuched will still cause this issue with X on logout.
Allegedly ati fixed this in its march 5 release, but at least in my experience, the only way to keep X server running ok was to disable the events daemon.
oh, and i can play youtube videos in fullscreen with no problems, incidentally.
Well, did you try putting some linux on the same box? i bet that with xubuntu, or even some high-spec distro, you could get better performance. my brother has an old dell with kubuntu, and it runs like a dream.
I have used OSX and linux (kubuntu mainly) on the same old ibook g4, and can say that linux generally runs smoother, unless of course the box you ran osx on is an uber-god-monster, in which case any os would run fine...
good to know that osx x86 is able to handle old x86 boxes, though...
His is the first post that states asensible criteria for the selection of articles: demand for the topic based on search queries and page-visits.
So, Faitfull to the *definition* of encyclopedia, as we have previosuly discussed here, and since theres no cost in having articles, WP should include ALL articles, and delete only if the demand for any given article, based on search and visits, is negligible.
that's asuming that you need to carry as much weight of hydrogen than of fuel for each plane, right?
i would have thought that hydrogen is more effective in the relation between energy and mass, but i must confess that i haven't paid much attention to this all hydrogen thingy...
yeah, i know how your federeal system works. I know about the electoral college and all that. But just as you can have federal regulations for some things, you could have federal regulations for these things too.
Take a student federation, for example: it is composed of different student unions syndicated in one bigger confederation. This unions are autonomous and "sovereign" in almost every sense, but the election for confederate representatives are enforced on confederate regulations.
There is no contradiction between having a federal multi-state republic, and have a coherent, rational and unified ellection system for federal authorities. Take Germany, for example.
And about Chavez... I truly believe that Venezuela's ellection system is waaaaay clenaer and more transparent than the US, at least. Chavez lost his last election, remember? he lost, and even more, he accepted the loss, even when the difference was about 1%... In th US, that same difference usually implies weeks or even months of courts and uncertainty.
The drudge work of overseeing an election should be done by representatives of the different parties or interenst groups present in the counting process. That's how we do it over here, and i think that in any other transparent voting system: with "apoderados" (i dont know the english term, sorry), that give its aproval to the each and every disputed vote. That way, that responsability falls on the parties.
i don't think that you could offend any chilean with that post, but you are not understanding how this copuntry works. We in chile do not have tons of money, true. But if we decide that we need some techno gadget to improve the efficiency of critical systems, usually we find the money and the engineering to pull it off. Take our highway-tolls for example: they are the most advanced in the world, and i'm not kidding you (see here).
Besides that, the point is that in chile we *don't have* discrepancies. We get results with 99.9% acuracy at 9-10 pm on the elction day. We tally all votes in the week following the election, and the final discrepancy between preliminary, 99%-confidence results and final results is always less than 0.02%. that's NOT 98% accuracy. 2% of error in the most important decission a country can take is not acceptable, i would say. That's what surprises me: how can you, as a country, accept this!?
I think that the problem is not machine counting versus hand counting, although i think that hand-counting is the best system for straight, one question ballots. I think that your main problem is that you combine different ellections into one ballot.
I have always found it incredibly curious how, of all countries, the United States has such big problems for vote counting. I know that problems like these are everywhere to be found, and that the US hava a very atypical election machinery (with each state presenting votes as they see fit, and other decissions based on a per-county basis, etc etc), but all in all, it should be pretty obvious that you have some serious election problem.
Down in my country (i'm form Chile), the electoral system is incredible clean and efficient. Every vote is hand counted, and the aggregated results of the election are official one or two hours after the last table closes, with a certainty of about 99.9%... and it's not a technological wonder: just ordered hand counting, and coordinated recollection of results. i know, we are a small country, but the voting population is about 4 mill people... more than NH in any case.
And in the event that there's a problem (i don't remember any in the last 20 years), we can track each ballot to the specific table where it was counted and check it all the way down to the ballot.
And Chile is a country with a reputation for chaos and disorder. Should i be amazed for our electoral system, or be amazed for how crappy the united states' system is?
in other words... with all due respect (and i mean it, it's an honet question...), why do you have such a crappy system? wouldn't it be cheaper to implement a low-tech, efficient and accountable sytem rather than risking every election with a thrillion different systems for each district and all this eternal debate about who probably got more votes?
Actaully, although i'm not american, i too think that most reasonable alternative is to say that ford "invented" the automobile, or whoever introduced the first mass-produced comercially viable product, or, in any other industry, the same criteria to determine when an application is "ready for production" rather than prototypes and early drafts.
And in that sence, i think that the assumption of Ford as the "inventor" is preetty acurate.
Is either that, or eternal discussions like the one above about computers. Introduction to a market marks a clear and objective point of comparison. In any case, choosing this particular point above other is equally arbitry:)
Slashdot Mirror
wich is his point actually: just like linux is only a kernel and NOT an OS, when you say that "linux" runs on X device you say that an OS based on the linux kernel runs on several devices, wich is the same than saying that Mac OS X is BSD/Mach.
so gp is spot-on, and you are actually saying the same thing.
amen.
short answer: no.
long: it depends. are they packing software written by other people? they are not developing anytyhing, so there's nothing to share. Are they runnning their own software over a linux system? There's no reasson to asume that this google developed software constitutes a derivative work and has to be GPLed, you can put propietary non-gpled software in a distro, like the binary blobs that go into ubuntu. Are they introducing changes to the underlying, already GPLed code? then yes, they would have to make source and rights available IF they distribute it.
Distro making is just the art of putting together software written by other people, GPLed or not, and passing it along. IF you happen to develope something, AND if you happen to use GPLed work as a foundation, making your work a derivative, then, in that particular case, yes: you have to share just like when you have to share any other GPL-derivative work.
but there's a big gaping hole in this particular application to the prisoners dilemma: there is NEED.
:)
if i NEED to get this piece of software developed, i just can't risk the possibility of it not getting funded and then dropped. If i really need it, i will try to make sure that it stays afloat, and that introduces a bigger incentive than the pure free riding impulse to desert all other companies and hope someone cooperates and gets exploited by desertors.
In real life, the prisoners dilemma takes a much more complicated form as the possibilities of interaction are much more varied than the two prototypical cooperate/desert, and the incentive and payouts structure is much much much more finegrained and open to absolutes, wich finally transform the whole rational decission making process altogether.
even if i would generally agree with you
well, that's the thing, precisely. His statement points to the negative aspect of FS on this particular thing, namely, the very real possibility that a project gets abandoned, instead of pointing out the even more real and evident positive aspect of FS in this very same aspect, namely the posibility of picking up support *even when and if the project gets abandoned*.
Putting it as he put it, it sounds like FUD. even if he personally didn't intended it so. And since this statements are probably sanctioned by MS PR, that accusation is not held against him personally: we are not assuming that MS employees are personally evil, but as long as they speak using their MS credentials, its MS speaking. and we do believe that MS, as an organization, is evil.
well, snakes are preetty much deaf, they "hear" by feeling vibrations in the ground.
but the whole point is that in closed source scenarios, the likelihood of that auditing is even less. and the probability of a bug creeping thru the audit decreases as the number of simultaneous auditors increase. and to top it off, the auditing depends on a concerted decision to divert scarce resources to auditing. In open source development, the problem is out in the open, and that increases the number of people (potentially) trying to solve it.
the argument is still correct: the probability of arriving to a satisfactory solution increases as the number of people trying to solve it increases, not even accounting for the talent of the people involved, and this is *obvious*, its a pure mathematical truth, spurious counter examples not-withstanding.
and this is a perfect example. Fermilab staff try to break it for a year without getting nowhere, and after the problem goes "open-source", it is solved in one week.
(who cares anyway? OSS is good only as it is a prerequisite for FS. its technical atributes are irrelevant)
now THAT's funny.
you are wrong.
slavery is banned in all liberal legislations.
"slavery" is precisely defined as the act of giving away self-ownership, in a manner such that you lose power over the decission (ie: you cannot revert it, since you gave yourself away to someone else), and it's illegal in any legal system based on a system of personal rights.
"voluntary and time-limited" is not slavery, its just voluntary work. Calling voluntary work slavery is at least wrong (and probably fascist, racist, etc. although i couldn't care less).
And to GP: the criminal status of any activity is subject to several considerations, one of which can be consent. Normally, the consent is disregarded only in crimes against life, and highly considered in sexual crimes, where the "crime" is by definition the lack of consent in the consumation of the act deemed criminal.
In england in particular, consented assault is only criminal when the resulting injuries are extremely severe, and you can prove that this injuries escape some pre-arranged form of consent (like the limits imposed by the rules of a game).
so OP is correct: having engaged in violent consensual activity wouldn't (in most cases) be a crime, and you can not derive its criminal status simply from a photograph.
actually, the us already had its hand in paraguay AND bolivia when it was believed that there was oil in the chaco region, shared by the two countries.
this eventauly ended up in a war between paraguay and bolivia, with the US based Standard Oil supporting Bolivia, and the UK based Shell Oil supporting Paraguay.
This was the worst armed conflict in latin america in the 20th century. over 35.000 died. and of course, there was no fucking oil whatsoever.
plumetting, actually.
and down here in the south, our fricking exports are paid in dollars and we are getting screwed big time. now our central bank has come out to buy dollars to keep the price at some reasonable price, and we all share the love paying for the comercial imbalances of the us. yay!
GP is bulshit. US dollar is at its lowest since 1995.
Apart from the reliability issues already put up by several of my fellow /.ers, there's another problem related to the different standard of proof required by different legal actions. ie, you requiere lesser standards of proof for a civil case, where in theory you are not violating the liberties of anyone, than in a criminal case, where the outcome might strip someone of very basic rights.
In this context, you need totally clean, perfect and irrefutable evidence to convict of piracy, than the one you need to prevent a private third party of violating your liberties by spying on you.
Now, the above has two twists:
1.- The possitions can be reversed, mediasentry has rights of their own, although juridic personalities (or whatever you call organizations) don't have the same rights as natural personalities (or whatever you call people).
2.- The above is nothing more than a more legalesse formulation of the argument about the reliability of IP addresses as proof. That's what standards of proof are all about.
The point is that you can have it in such a way that IP based evidence is inadmisible to prove copyright infringment, but admissible to proove C&D infringment. And this is no doulbe standard, and we are still much muhc much better than them.
ps: IANAL.
pps: I'm not from the USA, so things might be different in your legal system... but AFAIK, that's what the "reasonable doubt" thingy you use is about.
because archive != packages and archive FORMATS != package SYSTEMS.
tar, bz2, rar, zip, etc are compression/archiving *formats*, to put a bunch of files and lump them together in a (probably compressed) big file, or set of files. to use it, you just need a program that can read and extract the files.
A package is much more than that: it contains the binary files that make up some piece of software, and is in this sense an "archive", but the important part of the packaging is the dependencies management. Any package manager acomplish its function because every package references other packages that it depends upon, and keeps track of this so you don't have to worry about satisfying dependencies and cleaning unneeded garbage.
This structure of dependencies is called a repository, and is the core and soul of a distribution (at least package-based ones), and what it makes it superior or inferior. That's what "debian-based" or "red-hat-based" stands for: based *on the repos* of debian or red-hat.
And of course, there are structural differences in the ways that distros are built, and this affects wich files and wich pieces of software end up in a given package, etc etc.
So, although you actually can "read" rpm with apt and vice versa (there are lots of tools for this), the whole point of having package managers in the first place makes this exercise futile, as you would still have to satisfy the dependencies, by hand.
it's not that the packages are locked into the managers. is that the managers "manage" packages with references to a repository, and this repositories are distro-based, to suit the needs of a specific user base.
This is of course, a Good Thing (tm), as it encourages innovation and diversity, and it doesn't present that many problems, as the use cases in which it turns out problematic (ie, needing an rpm-only program in a deb-based system) is very rare (eg, i remember only two times in my years with *buntu).
and even in this case, compiling is really not that hard...
so basically, we have to agree on a package management system, and we are set...
and we are half the way there, as we already have one, two or three mainstream package formats, wich can be used for a whole range of distros. call it whatever you like, in the end it all comes down to the package format. deb or rpm?
mmmm, rotten egg stinking bacon. yammie!
and that "one-atom" thickness sure makes a hell of a sexy boob!
uuuuh i can feel your nipple neutron there baby!
that bug's related to atieventsd. If you disable the daemon from loading, it wont corrupt your X server on logout. You will have to reboot, though. Disabling the daemon once it has already lanuched will still cause this issue with X on logout.
Allegedly ati fixed this in its march 5 release, but at least in my experience, the only way to keep X server running ok was to disable the events daemon.
oh, and i can play youtube videos in fullscreen with no problems, incidentally.
Well, did you try putting some linux on the same box? i bet that with xubuntu, or even some high-spec distro, you could get better performance. my brother has an old dell with kubuntu, and it runs like a dream.
I have used OSX and linux (kubuntu mainly) on the same old ibook g4, and can say that linux generally runs smoother, unless of course the box you ran osx on is an uber-god-monster, in which case any os would run fine...
good to know that osx x86 is able to handle old x86 boxes, though...
plis mod parent up!
His is the first post that states asensible criteria for the selection of articles: demand for the topic based on search queries and page-visits.
So, Faitfull to the *definition* of encyclopedia, as we have previosuly discussed here, and since theres no cost in having articles, WP should include ALL articles, and delete only if the demand for any given article, based on search and visits, is negligible.
that's asuming that you need to carry as much weight of hydrogen than of fuel for each plane, right?
i would have thought that hydrogen is more effective in the relation between energy and mass, but i must confess that i haven't paid much attention to this all hydrogen thingy...
yeah, i know how your federeal system works. I know about the electoral college and all that. But just as you can have federal regulations for some things, you could have federal regulations for these things too.
Take a student federation, for example: it is composed of different student unions syndicated in one bigger confederation. This unions are autonomous and "sovereign" in almost every sense, but the election for confederate representatives are enforced on confederate regulations.
There is no contradiction between having a federal multi-state republic, and have a coherent, rational and unified ellection system for federal authorities. Take Germany, for example.
And about Chavez... I truly believe that Venezuela's ellection system is waaaaay clenaer and more transparent than the US, at least. Chavez lost his last election, remember? he lost, and even more, he accepted the loss, even when the difference was about 1%... In th US, that same difference usually implies weeks or even months of courts and uncertainty.
The drudge work of overseeing an election should be done by representatives of the different parties or interenst groups present in the counting process. That's how we do it over here, and i think that in any other transparent voting system: with "apoderados" (i dont know the english term, sorry), that give its aproval to the each and every disputed vote. That way, that responsability falls on the parties.
i don't think that you could offend any chilean with that post, but you are not understanding how this copuntry works. We in chile do not have tons of money, true. But if we decide that we need some techno gadget to improve the efficiency of critical systems, usually we find the money and the engineering to pull it off. Take our highway-tolls for example: they are the most advanced in the world, and i'm not kidding you (see here).
Besides that, the point is that in chile we *don't have* discrepancies. We get results with 99.9% acuracy at 9-10 pm on the elction day. We tally all votes in the week following the election, and the final discrepancy between preliminary, 99%-confidence results and final results is always less than 0.02%. that's NOT 98% accuracy. 2% of error in the most important decission a country can take is not acceptable, i would say. That's what surprises me: how can you, as a country, accept this!?
I think that the problem is not machine counting versus hand counting, although i think that hand-counting is the best system for straight, one question ballots. I think that your main problem is that you combine different ellections into one ballot.
I have always found it incredibly curious how, of all countries, the United States has such big problems for vote counting. I know that problems like these are everywhere to be found, and that the US hava a very atypical election machinery (with each state presenting votes as they see fit, and other decissions based on a per-county basis, etc etc), but all in all, it should be pretty obvious that you have some serious election problem.
Down in my country (i'm form Chile), the electoral system is incredible clean and efficient. Every vote is hand counted, and the aggregated results of the election are official one or two hours after the last table closes, with a certainty of about 99.9%... and it's not a technological wonder: just ordered hand counting, and coordinated recollection of results. i know, we are a small country, but the voting population is about 4 mill people... more than NH in any case.
And in the event that there's a problem (i don't remember any in the last 20 years), we can track each ballot to the specific table where it was counted and check it all the way down to the ballot.
And Chile is a country with a reputation for chaos and disorder. Should i be amazed for our electoral system, or be amazed for how crappy the united states' system is?
in other words... with all due respect (and i mean it, it's an honet question...), why do you have such a crappy system? wouldn't it be cheaper to implement a low-tech, efficient and accountable sytem rather than risking every election with a thrillion different systems for each district and all this eternal debate about who probably got more votes?
Actaully, although i'm not american, i too think that most reasonable alternative is to say that ford "invented" the automobile, or whoever introduced the first mass-produced comercially viable product, or, in any other industry, the same criteria to determine when an application is "ready for production" rather than prototypes and early drafts.
:)
And in that sence, i think that the assumption of Ford as the "inventor" is preetty acurate.
Is either that, or eternal discussions like the one above about computers. Introduction to a market marks a clear and objective point of comparison. In any case, choosing this particular point above other is equally arbitry