you completely missed my point. indexing really offers nothing to the overwhelming majority of machines running linux because the number of users listed in/etc/passwd is so small.
Not odd at all... you can have the source for free - since anyone else can give it you as well. But if they do the work of building it, they want you to pay for the time and effort you've saved (which apparently on Windows, seems like quite a bit).
so, there are a lot of us in the following position, no doubt: we run a router (linksys, whatever) that gets DNS from our ISP. lets assume that the ISP is patched. our local machines use the router for DNS. do we need to patch the router? are its DNS request services even accessible to the external network? can it be compromised in the same way that the ISP DNS could be?
i have been wondering this ever since news of this problem broke, and i have still not seen a clear answer.
"In Rainbows" and perhaps other Radiohead albums have been (and continue to be) available from emusic.com, for up to 50% less than the iTunes price, DRM-free, lifetime-replaceable.
so basically, like, meh.
society Foo is NOT free. it significantly restricts the actions of MURDERERS, RAPISTS and CHILD MOLESTERS. thus it is clearly not free.
laws are not nothing more than anti freedom licensing, and has significantly diminished all our freedoms.
society Bar is a truly free place. it allows anyone to beat anyone else over the head with a cast iron pipe for no reason other than they enjoy doing it. laws that stop people doing what they enjoy is a huge step backwards, IMHO.
your karma drops when you use regular/. articles to try to launch a meta-discussion about/. itself. is this good or bad? i don't know. but it parallels lots of other things in life, both online and elsewhere.
Not talking to the police is the first -- and almost only -- rule in dealing with the police
That wouldn't have helped a teenager under my care who was recently (correctly) charged with stealing from his employer. His cooperation with the police and honesty about what he had done transformed his situation from one where he was facing life with a criminal record (along with an under-age drinking charge that could have relieved him of his drivers license for a while) into a relatively mild case that will likely result in a few hours community service.
Small town police and big city detectives are not the same thing. It can pay to differentiate.
the prize was not US$1000. it started out very small. Knuth did indeed pay out, and indeed doubled it, several times. From wikipedia: "The award per bug started at $2.56 (one "hexadecimal dollar"[24]) and doubled every year until it was frozen at its current value of $327.68. This has not made Knuth poor, however, as there have been very few bugs claimed. In addition, people have been known to frame a check proving they found a bug in TeX instead of cashing it."
spoken like a true parallel programming neophyte. 14 years ago, i was programming on a 64 core system. a couple of years before that, a 32 core system. when you don't have enough perspective on this matter, you're easily susceptible to some crazy idea that "eight! i mean think about that, eight cores is quite a bit". you're even more susceptible to the idea that parallel systems are going to follow some moore-ish curve, when they've already been in and out of fashion (and up and down in scale) several times since the idea was first floated.
parallel systems and parallel programming are not a panacea, they are not trivial to use (and probably never will be), but they have their uses. the current bubble in "multicore" systems is not the first time they've become popular at some level, and it probably won't be the last.
"OS running real user apps in real world condition"
Lack of precision is sometimes roughly equivalent to handwaving. Your list of "user" apps is not necessarily my list. Your list of "conditions" is almost certainly not my list either. I agree that there is work to be done in this area and that the current behaviour is not ideal for many kinds of desktop/office use patterns. But your original description of what happens, what can happen and what cannot happen was just wrong.
i suggest you read the output of man memlock. you clearly don't know enough about linux (or POSIX) to be making generic hand waving comments that appear to be intended to authoritative.
when you're done with memlock, check into SCHED_FIFO scheduling too. oh, and/etc/security/limits.conf while you're there. the problems with multimedia "performance" on linux are mostly distro-related: distro's do not generally ship in a way that lets ordinary users run apps that request the use of these facilities. media-centric distros (Ubuntu Studio) or overlays (Planet CCRMA) fix this.
Not bad. Not bad until you got to the part about housing.
US housing costs are dramatically cheaper (on average) than those in western Europe. The primary reason for the difference is that housing costs in the US reflect the fact that land on which to build is cheap, so the cost of buying an existing house has to compete with the fact that you could, if you were willing, simply build a new one. This option is generally unimaginable for inhabitants of most of Europe, where land prices make this option absurd. As a result, house/apartment/rental costs there are not competing with the "i'll do it myself" option, and can climb to levels contained only by median salaries.
Your inevitablity stance on a global economy is also a little sad. Things like the "global economy" don't just happen. They happen because a specific (if large) set of vested interests arrange/push for it to happen. In this case, owners of capital who stand to see huge benefits from the free flow of their property, have pushed hard for it while telling everyone that the whole world will benefit from it. They have resisted similarly free flow of labor, while relying on the fact that moving labor around is much harder than moving capital. It was never inevitable - its the result of power and money seeking more power and more money, just like so much of human history has been.
A union that spans only a single company is almost completely useless. Employers gain power over employees from the fact that, despite being competitive in the consumer marketplace, their interests are generally aligned when it comes to their operating practices. Likewise unions only gain any kind of power to negotiate by recognizing that the interests of employees spans different companies and in some cases, even different industries. That is, they prevent an employer from saying to the in-company union "we don't want to negotiate with you, you're all fired", because the employer knows that terms for the replacement set of employees will be identical (and don't kid yourself, more and more people worldwide work in jobs for which they are completely replaceable.)
Unions got you the weekend, the 40 hour week, paid vacation and health benefits. Yes, US unions abdicated most of their power in silly per-company deals and forgot about the wider political arena, which is why their power is a fraction of that in Europe where they have remained political organizations with stated goals and interests in national issues. But then again, I guess since you think per-company unions are the only form that make sense, you're a part of that sellout too?
amazing. finally i've found someone who has done real double blind tests of lossy audio data compression versus high quality originals in real spaces, and has proven conclusively that at least 99% of the listening audience can tell the difference. could you post a URL where i can read the details of your testing, so that i can post it whenever some fool shows up and claims that every double blind test to date has shown that virtually no-one can tell the difference? i'd also love to read your work on DSD and 96kHz double blind testing too, because i've been looking for material to argue against all the tests that have shown no discernible difference for at least 90% of the listening audience. please, lets put these losers where they belong, and make sure the world fully understands just how much of a step backward is going on here!
we can't get them impeached when they break the laws, violate the constitution, torture, engage in warmaking, arrest without probable cause, hold people incommunicado without hearings for extended periods of time, make a huge industry out of imprisoning the population for personal choices about what intoxicants they prefer...
as much as i agree with you about the behaviour of this (and other) administrations, i think you're being unnecessarily cynical. the reason they can get away with these things is that there is substantial disagreement among the population about whether they have actually done anything wrong. the low approval ratings for the current president do not equate to a belief that he should be impeached. do you seriously believe for one moment that if people were actually riled up in large numbers about these issues that impeachment (or something like it) would still fail to happen? we, in our collective apathy, get the government we (collectively) deserve...
what you say is true, but nevertheless, it has taken a substantial, long term effort to get this patent revoked. it didn't work the first time around (c/o tim o'reilly) and even now, the prior art gathered took some real work to dig up. it just wasn't the case in 1994 that you could buy stuff from online stores without going through the whole checkout thing. it should never have been submitted as a patent, it should never have been awarded a patent, but i stand by my point that it was not "obvious" in the sense that the problem(s) with the ">1 click" process were only just beginning to be identified. the customer base for any online store at that time were happy enough not be using archie and gopher:)
I helped to start Amazon (I was the 2nd employee there). I've spoken out against the 1 click patent in the past. However, this comment "And it only took many many years to remove what would have been obvious to the most incompetent web developer" is not the reason why the patent should be permanently rejected. 1 click shopping was "new" at the time - if it was obvious, we would have done it right from the beginning on the web site. The issue with 1 click is not whether or not it was obvious to a web developer. It is whether or not business method patents that fundamentally simply map a practice in the non-online world ("put this on my account") to the online world ("1 click") should be permitted.
I don't believe that they should, and I am glad to see the patent struck down.
what is being discussed is called "scheduler activations" within the CS community (or was). its an old idea. i did some work on a real-world (hah) implemention back in the early 1990's when i worked at UWashington. google it. Solaris actually added this design at least 10 years ago (plus or minus 2 years). its a very cool OS design, but can also be hard to get the implementation right; it also requires both kernel and userspace implementations.
the basic idea is that the kernel doesn't try to decide which threads within a task/process should run. as long as the process is scheduled to have access to a CPU, whenever its about to block (e.g. on disk i/o) or to be granted a processor from another task, the kernel tells the user space scheduler what is going on. scheduling is then done in user space, where maximal knowledge about the applications internal design and thread priorities can be easily accessed.
there are several papers on this design, ranging from Tom Anderson's "original" through reports on various implementation efforts. it was certainly fun trying to write a user space context switch routine that has to be reentrant itself, not to mention trying to deal with priority inversion issues. i think sun simply worked around the latter problem with some design assumptions/limitations, but i don't know for sure.
i very much doubt that you can produce a single peer-reviewed citation for even one clause, let alone sentence, in your description. hand-waving retrospective analysis to justify the status quo is fun, but as a method of arriving at the truth, it leaves much to be desired.
Slashdot Mirror
you completely missed my point. indexing really offers nothing to the overwhelming majority of machines running linux because the number of users listed in /etc/passwd is so small.
% wc -l /etc/passwd
39
%
absolutely nothing. of course, downloaders have to trust the download source for that to work.
Not odd at all ... you can have the source for free - since anyone else can give it you as well. But if they do the work of building it, they want you to pay for the time and effort you've saved (which apparently on Windows, seems like quite a bit).
so, there are a lot of us in the following position, no doubt: we run a router (linksys, whatever) that gets DNS from our ISP. lets assume that the ISP is patched. our local machines use the router for DNS. do we need to patch the router? are its DNS request services even accessible to the external network? can it be compromised in the same way that the ISP DNS could be? i have been wondering this ever since news of this problem broke, and i have still not seen a clear answer.
"In Rainbows" and perhaps other Radiohead albums have been (and continue to be) available from emusic.com, for up to 50% less than the iTunes price, DRM-free, lifetime-replaceable. so basically, like, meh.
society Foo is NOT free. it significantly restricts the actions of MURDERERS, RAPISTS and CHILD MOLESTERS. thus it is clearly not free. laws are not nothing more than anti freedom licensing, and has significantly diminished all our freedoms. society Bar is a truly free place. it allows anyone to beat anyone else over the head with a cast iron pipe for no reason other than they enjoy doing it. laws that stop people doing what they enjoy is a huge step backwards, IMHO.
your karma drops when you use regular /. articles to try to launch a meta-discussion about /. itself. is this good or bad? i don't know. but it parallels lots of other things in life, both online and elsewhere.
So much for the stale, dull, sour-faced, monkish slashdot community modding you down to -1, eh?
Not talking to the police is the first -- and almost only -- rule in dealing with the police
That wouldn't have helped a teenager under my care who was recently (correctly) charged with stealing from his employer. His cooperation with the police and honesty about what he had done transformed his situation from one where he was facing life with a criminal record (along with an under-age drinking charge that could have relieved him of his drivers license for a while) into a relatively mild case that will likely result in a few hours community service.
Small town police and big city detectives are not the same thing. It can pay to differentiate.
the prize was not US$1000. it started out very small. Knuth did indeed pay out, and indeed doubled it, several times. From wikipedia: "The award per bug started at $2.56 (one "hexadecimal dollar"[24]) and doubled every year until it was frozen at its current value of $327.68. This has not made Knuth poor, however, as there have been very few bugs claimed. In addition, people have been known to frame a check proving they found a bug in TeX instead of cashing it."
spoken like a true parallel programming neophyte. 14 years ago, i was programming on a 64 core system. a couple of years before that, a 32 core system. when you don't have enough perspective on this matter, you're easily susceptible to some crazy idea that "eight! i mean think about that, eight cores is quite a bit". you're even more susceptible to the idea that parallel systems are going to follow some moore-ish curve, when they've already been in and out of fashion (and up and down in scale) several times since the idea was first floated.
parallel systems and parallel programming are not a panacea, they are not trivial to use (and probably never will be), but they have their uses. the current bubble in "multicore" systems is not the first time they've become popular at some level, and it probably won't be the last.
"OS running real user apps in real world condition"
Lack of precision is sometimes roughly equivalent to handwaving. Your list of "user" apps is not necessarily my list. Your list of "conditions" is almost certainly not my list either. I agree that there is work to be done in this area and that the current behaviour is not ideal for many kinds of desktop/office use patterns. But your original description of what happens, what can happen and what cannot happen was just wrong.
"linux swap paradigm".
i suggest you read the output of man memlock. you clearly don't know enough about linux (or POSIX) to be making generic hand waving comments that appear to be intended to authoritative.
when you're done with memlock, check into SCHED_FIFO scheduling too. oh, and /etc/security/limits.conf while you're there. the problems with multimedia "performance" on linux are mostly distro-related: distro's do not generally ship in a way that lets ordinary users run apps that request the use of these facilities. media-centric distros (Ubuntu Studio) or overlays (Planet CCRMA) fix this.
Not bad. Not bad until you got to the part about housing.
US housing costs are dramatically cheaper (on average) than those in western Europe. The primary reason for the difference is that housing costs in the US reflect the fact that land on which to build is cheap, so the cost of buying an existing house has to compete with the fact that you could, if you were willing, simply build a new one. This option is generally unimaginable for inhabitants of most of Europe, where land prices make this option absurd. As a result, house/apartment/rental costs there are not competing with the "i'll do it myself" option, and can climb to levels contained only by median salaries.
Your inevitablity stance on a global economy is also a little sad. Things like the "global economy" don't just happen. They happen because a specific (if large) set of vested interests arrange/push for it to happen. In this case, owners of capital who stand to see huge benefits from the free flow of their property, have pushed hard for it while telling everyone that the whole world will benefit from it. They have resisted similarly free flow of labor, while relying on the fact that moving labor around is much harder than moving capital. It was never inevitable - its the result of power and money seeking more power and more money, just like so much of human history has been.
A union that spans only a single company is almost completely useless. Employers gain power over employees from the fact that, despite being competitive in the consumer marketplace, their interests are generally aligned when it comes to their operating practices. Likewise unions only gain any kind of power to negotiate by recognizing that the interests of employees spans different companies and in some cases, even different industries. That is, they prevent an employer from saying to the in-company union "we don't want to negotiate with you, you're all fired", because the employer knows that terms for the replacement set of employees will be identical (and don't kid yourself, more and more people worldwide work in jobs for which they are completely replaceable.)
Unions got you the weekend, the 40 hour week, paid vacation and health benefits. Yes, US unions abdicated most of their power in silly per-company deals and forgot about the wider political arena, which is why their power is a fraction of that in Europe where they have remained political organizations with stated goals and interests in national issues. But then again, I guess since you think per-company unions are the only form that make sense, you're a part of that sellout too?
amazing. finally i've found someone who has done real double blind tests of lossy audio data compression versus high quality originals in real spaces, and has proven conclusively that at least 99% of the listening audience can tell the difference. could you post a URL where i can read the details of your testing, so that i can post it whenever some fool shows up and claims that every double blind test to date has shown that virtually no-one can tell the difference? i'd also love to read your work on DSD and 96kHz double blind testing too, because i've been looking for material to argue against all the tests that have shown no discernible difference for at least 90% of the listening audience. please, lets put these losers where they belong, and make sure the world fully understands just how much of a step backward is going on here!
we can't get them impeached when they break the laws, violate the constitution, torture, engage in warmaking, arrest without probable cause, hold people incommunicado without hearings for extended periods of time, make a huge industry out of imprisoning the population for personal choices about what intoxicants they prefer...
as much as i agree with you about the behaviour of this (and other) administrations, i think you're being unnecessarily cynical. the reason they can get away with these things is that there is substantial disagreement among the population about whether they have actually done anything wrong. the low approval ratings for the current president do not equate to a belief that he should be impeached. do you seriously believe for one moment that if people were actually riled up in large numbers about these issues that impeachment (or something like it) would still fail to happen? we, in our collective apathy, get the government we (collectively) deserve ...
no, but you could always subscribe to Ardour and help keep alive the development of foremost libre software digital audio workstation.
what you say is true, but nevertheless, it has taken a substantial, long term effort to get this patent revoked. it didn't work the first time around (c/o tim o'reilly) and even now, the prior art gathered took some real work to dig up. it just wasn't the case in 1994 that you could buy stuff from online stores without going through the whole checkout thing. it should never have been submitted as a patent, it should never have been awarded a patent, but i stand by my point that it was not "obvious" in the sense that the problem(s) with the ">1 click" process were only just beginning to be identified. the customer base for any online store at that time were happy enough not be using archie and gopher :)
I helped to start Amazon (I was the 2nd employee there). I've spoken out against the 1 click patent in the past. However, this comment "And it only took many many years to remove what would have been obvious to the most incompetent web developer" is not the reason why the patent should be permanently rejected. 1 click shopping was "new" at the time - if it was obvious, we would have done it right from the beginning on the web site. The issue with 1 click is not whether or not it was obvious to a web developer. It is whether or not business method patents that fundamentally simply map a practice in the non-online world ("put this on my account") to the online world ("1 click") should be permitted.
I don't believe that they should, and I am glad to see the patent struck down.
and no doubt you've done double blind ABX tests with this $300 cable to verify the differences that you believe are perceptible?
what is being discussed is called "scheduler activations" within the CS community (or was). its an old idea. i did some work on a real-world (hah) implemention back in the early 1990's when i worked at UWashington. google it. Solaris actually added this design at least 10 years ago (plus or minus 2 years). its a very cool OS design, but can also be hard to get the implementation right; it also requires both kernel and userspace implementations.
the basic idea is that the kernel doesn't try to decide which threads within a task/process should run. as long as the process is scheduled to have access to a CPU, whenever its about to block (e.g. on disk i/o) or to be granted a processor from another task, the kernel tells the user space scheduler what is going on. scheduling is then done in user space, where maximal knowledge about the applications internal design and thread priorities can be easily accessed.
there are several papers on this design, ranging from Tom Anderson's "original" through reports on various implementation efforts. it was certainly fun trying to write a user space context switch routine that has to be reentrant itself, not to mention trying to deal with priority inversion issues. i think sun simply worked around the latter problem with some design assumptions/limitations, but i don't know for sure.
#!/usr/bin/perl
chdir ("/PATH/TO/YOUR/MUSIC/LIBRARY");
while () {
if (/class="download"/) {
@tags = split (/[]/);
$tags[3] =~ s/a href=//;
$tags[3] =~ s/"//g;
system ("wget --load-cookies
}
}
exit 0;
i very much doubt that you can produce a single peer-reviewed citation for even one clause, let alone sentence, in your description. hand-waving retrospective analysis to justify the status quo is fun, but as a method of arriving at the truth, it leaves much to be desired.