Sure you can. If you are checking reverse resolution how do you run two domains on a single IP? No, what you do is check the -domain- the sender claims to have and its MX record. Update the MX record dynamically to run an SMTP server on a dynamic IP. (It's not totally safe but it'll work well enough for the types who would do it).
It's not hard. If you get mail from x@y.com then you contact y.com to verify it's real; indeed, you will need only make sure the connection is made from y.com in some cases (or all cases if so-designed). There is overhead there, but spam is more overhead. Do it this way and you can build up a domain blacklist for spammers.
But using the domain name system for addresses is pretty lame, considering it tends to make them unreliable (existentially).
It should also be noted that there is significant overlap; the old kernels are not completely unrelated to the new ones, and patches are both back-ported and forward-ported between them. The development of all trees benefits from the development of the other trees, in varying degrees. This also means that maintaining old kernels is not as much effort as maintaining the new ones (which is not to say that less effort is required of the people involved - but certainly fewer people are required to be involved).
Someone came to him with a "story idea". He asked for details and the guy said, "Amnesia". He replied, "ok, what's your idea?" They guy was confused and said again, "Amnesia" as if that were, in and of itself, a story idea.
Seems more likely the guy just forgot his story idea.
Uhh, I think there are actually about three or four hundred lines in between "make menuconfig" and "make bzImage".
You also have to manually upgrade all the packages to 2.6-supporting versions (by reading the README to find out what they are), and download and install the kernel source. Even if you wish to configure and compile the kernel yourself, Debian's package system will do this for you.
But the real issues involved in upgrading a kernel have nothing to do with how to install it. That is easy. LVM has been deprecated by a new version, devfs has been deprecated, ide-scsi has been deprecated, etc.: there are real issues to resolve that require changes -outside- the kernel and knowledge of the underlying issues.
Spelling is not an exact science either; the "correct" spelling of a word is a mixture of usage and history and it is not necessarily clear what to do when the two are inconsistent with each other or within themselves.
But this is all quite beside the point. The question was how can humans be inaccurate when humans define accuracy. That question is silly. Humans define accuracy, but they are not by definition accurate.
Please, Intel cannot force anyone out of the market by making incompatible CPUs. Every CPU they release introduces incompatibities! Along with every CPU released by every competitor. And compilers generate code for both their processors, and all their competitors'. And microsoft creates its software for both their processors and all their competitors'. It is not an evil practice; if you have the market significance to actually physically manufacture a general purpose CPU then you have the market significance to get software support for it (this is far, far cheaper). I've never heard of any case of Intel having exclusive support for any important software (not since DOS days anyway). People are just looking at Intel and seeing Microsoft, when really the practices of the two companies are not comparable, not on this point at least.
Intel is -not- a monopoly[1], and "embrace and extend" for instruction sets doesn't make any sense in today's CPU market, because the semantics of new instructions must be public (by convention and due to market realities), and so besides the fact that competitors are freely capable of implementing them, compilers are capable of generating equivalent code that doesn't use them. Intel is -not- trying to create a market of programs that won't run on AMD processors. The standard practice for use of new processor extensions in software is always -not- to break compatibility with older processors. Note that all Microsoft products are compatible with AMD and that Visual C++ as well as the rest of the Microsoft compiler suite generates code that is compatible with AMD. AMD -and- the other, smaller x86 competitors, mind you.
[1] This is not to say that none of their success is based on inertia. But come on - what software can you run on Intel that you can't run on AMD? Such software - badly written, unimportant programs - are the extent of Intel's monopoly. Right now, AMD makes slightly better processors, which is not enough to overcome Intel inertia; but if AMD made massively better processors, Intel would have no power in the market to force them not to be adopted; AMD would become top dog. The CPU market doesn't work like the software market.
Well, the math isn't wrong, but there is an unstated assumption, that whenever the two disagree, it will be possible to check and determine who is right. What multiple judges allows is better -detection- of errors, not in itself correction of errors. A majority scheme with an odd number of judges would indeed allow correction as well.
However, you've only actually increased the probability of error in any sense if you count a disagreement as an error - but there is no more reason to count a disagreement as an error than there is to count it as a correct judgment. It is, in fact, distinct from either. A disagreement implies no conclusion, so it cannot be right or wrong.
OK, but that's still braindead. If your house is on fire, and you don't pay or haven't payed to put it out, then what? You let it burn down the rest of the block, destroy all nearby power lines, sewage pipes, roads, trees? Firefighting in almost any location is a public good. Indeed, am I supposed to let my house burn up because my cheap neighbor got the slowest firefighting service in town and fell asleep smoking?
People who are actually responsible for fires are already legally liable for the damages they cause. Pushing liability further makes no sense[1], nor does making firefighting service optional.
[1] Free market people seem to love holding people responsible for things outside their control, perhaps because they want credit for their own privilege.
First of all, on any IM network, this is (for practical purposes) a client issue. Second, ICQ and AOL's clients and the popular free alternatives for their services also offer this.
If every individual human has an accuracy of 99.983%, then two independent humans have an accuracy of 1 -.00017^2 or 99.99999711%. This would allow ample accuracy to judge the computer, except that it's not true[1]. A better answer is the one you suggest: humans must judge spam from subject/author alone, whereas computers get to look at the whole message. Humans reading the whole message, and possibly even following included links, responding, etc., can be assumed to have full accuracy, within epistemic bounds. Indeed, merely re-checking your work, etc. - being consciously more diligent than the average spam-sorter - should insure your accuracy is better than average.
As for how accuracy was actually judged in this particular study, I suppose you would have to read the article for that. I haven't, myself...
[1] It assumes the probability of error is equal for every message, which is obviously not true (i.e., that error is random rather than systematic). The real accuracy of two humans in concert is surely much lower; OTOH, it is still sure to be much, much higher than the accuracy of a single human.
Unfortunately, if you build a module against a tree that does not have the same configuration as the binary kernel that will load it, it sometimes will not work. Sometimes, even adding the module to the existing kernel will require recompiling bzImage, else the module will not work (which is highly counter-intuitive). This is unfortunate, but I don't think the linux developers have made any attempt to allow modules to be compiled piecewise in general (it wouldn't be very hard), and there doesn't seem any way to find out which modules or base kernels will work without just trying it. Additionally, even if it does load, in at least one instance I've had a loaded module not work for this reason; it was a third-party module, but Linux makes no guarantee that its own modules will work under these circumstances either.
I've talked about the general issue and reasons I disagree with this perspective, elsewhere in the thread. I will only add here that, although it may seem perverse, I would actually prefer that it be difficult to "keep crime in check despite increasing poverty". In general, when it is easier to stop crime by ever-more-powerful law enforcement than by ameliorating the social causes of crime, I anticipate evils far greater than common crime, and I fear any technology which brings us further into that world.
It's true that, as a practical matter, privacy in public places cannot be obtained. That is no reason to say it is not desirable, or that greater privacy in public should not be encouraged. A camera on a street corner might be compared to an anonymous passer-by observing your day-to-day routine. Or it might be compared to an ominous figure lurking in the bushes, following you from a block or two behind, mysterously present every day. Unless we know how the cameras are being used, which we cannot, we do not know which comparison is more apt. The reality likely varies depending on the individual being taped.
One significant difference between public spaces with cameras and without is that, in general, in public spaces the observer is also public; he cannot hide from you any more than you from him. That seems to be a good check on particularly odious police monitoring; it is legal, for sure, but since it can be observed by anyone, the police are still checked by public opinion. There is no public opinion of secret police activity, though, and all monitoring via camera falls into this category.
Well, not everyone leads a boring life, and not every interesting life is criminal. For example, political activists are already closely monitored by the US government in legal and illegal ways. As a recent real-world example, I live in Connecticut and as you may have heard, we are having some problems with our governor accepting bribes, kickbacks, etc. Recently there was an open meeting of citizens seeking to hasten his removal from office, and a uniformed police officer showed up, gave his card to some activists whom he addressed by name (people who had certainly never met him), and generally spoke as much as possible, in an attempt to disrupt the meeting. Naturally, he was just trying to scare people by proving to them that they are being watched. But there is good reason for that to be scary, and it is likely that this information is being gathered for purposes beyond small-time intimidation tactics.
When the government knows what you're doing, even when it's legal, it can treat you differently for doing it, even when it's legal. This may take the form of petty harrassment, selective enforcement of commonly ignored laws, or something even more ominous. Obviously, you're right, we can't practically prevent the government from knowing about a certain amount of legal activity -- but we should not openly invite them to monitor all legal activity. Maybe that 10pm walk is to a political meeting; maybe it's to your gay lover's apartment; maybe it's to an AA meeting -- but if you're not breaking the law, it's none of the government's business.
First of all, the author isn't violating the GPL unless he's copying somebody else's GPL'd code. Otherwise, it's simply not possible. But I know what you mean, and you're wrong about that, too.
From the GPL FAQ:
If the libraries that you link with fall within the following exception
in the GPL:
However, as a special exception, the source code distributed need not
include anything that is normally distributed (in either source or
binary form) with the major components (compiler, kernel, and so on) of
the operating system on which the executable runs, unless that
component itself accompanies the executable.
then you don't have to do anything special to use them; the requirement
to distribute source code for the whole program does not include those
libraries, even if you distribute a linked executable containing them.
Thus, if the libraries you need come with major parts of a proprietary
operating system, the GPL says people can link your program with them
without any conditions.
If you want your program to link against a library not covered by that
exception, you need to add your own exception, wholly outside of the
GPL. This copyright notice and license notice give permission to link
with the program FOO:
Copyright (C) yyyy <name of copyright holder>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
In addition, as a special exception, <name of copyright
holder> gives permission to link the code of this program with
the FOO library (or with modified versions of FOO that use the
same license as FOO), and distribute linked combinations including
the two. You must obey the GNU General Public License in all
respects for all of the code used other than FOO. If you modify
this file, you may extend this exception to your version of the
file, but you are not obligated to do so. If you do not wish to
do so, delete this exception statement from your version.
Only the copyright holders for the program can legally authorize this
exception. If you wrote the whole program yourself, then assuming your
employer or school does not claim the copyright, you are the copyright
holder--so you can authorize the exception. But if you want to use parts
of other GPL-covered programs by other authors in your code, you cannot
authorize the exception for them. You have to get the approval of the
copyright holders of those programs.
When other people modify the program, they do not have to make the same
exception for their code--it is their choice whether to do so.
So, (1) the exemption includes the Java subsystem since at least one operating system is regularly distributed with it (2) Java executables don't static link against class libraries, and therefore don't link against a particular implementation with a particular license anyway, and (3) you're allowed to link against proprietary libs, so long as you know what you're doing (you don't need to LGPL anything). Of course, we need GPL'd Java libraries, but that's no reason to wait for them before writing anything in Java.
OTOH, Java shouldn't be used for a host of other reasons.
There is apparently a widespread myth that the FSF is giving out free legal services to GPL copyright holders. Nonsense! If the FSF owns a copyright, they -can- sue; but they have made no promise to protect any copyright assigned to them, and generally speaking do not have the capability to do so. The FSF is on the order of six people, including four programmers. They're not IBM. They don't have a legal task force. Eben Moglen is probably not licensed to practice in your state, assuming you're even American, and he's a full-time professor of law, -not- a full-time lawyer for FSF, so I don't know that he would even have the capacity to represent them in a lawsuit they -needed- to fight, let alone every petty copyright case thrown them. He hasn't represented them in a suit so far.
They are -already- in violation of the law. You can pursue that if they do not meet your terms. So, you -can- force them to release the code, or to give you money, or whatever, at the threat of a suit.
Also, you cannot "get in trouble" for mistakenly claiming to a company that it is in violation of the law. That is both absurd and unprecedented.
(If you file a police report or file for an injunction in court or something like that, then you can get in trouble. But only for lying; not for making a mistake of fact, and certainly not for making a mistake of law).
No. When you sue someone in a court that doesn't have jurisdiction over that someone, the court just throws it out. It doesn't offer some sort of meaningless default win for failure to appear.
Cooperation on the Internet works on the basis of social pressure, not on legislation. Legislation will only cause the Internet to fragment and "route around" the stupidity.
Be realistic. Legislation is an easy way to generate social pressure, far easier in general than rational argument. When the two come up against each other, there is certainly no reason to assume a priori that legislation will lose. Indeed, "social" pressure on the internet has nothing to do with society; net standards are determined by a small group of industry insiders, entry to which usually requires the favor (and often employ) of large tech companies. Since tech companies need government contracts and gov't support to survive, the government can, if it chooses, directly dictate practices to them. Generally they would meet no resistence, since they would be sure not to harm the business no matter how they might harm the user. But government pressure is ultimately impossible to resist anyway, if it is sufficiently determined.
Slashdot Mirror
Sure you can. If you are checking reverse resolution how do you run two domains on a single IP? No, what you do is check the -domain- the sender claims to have and its MX record. Update the MX record dynamically to run an SMTP server on a dynamic IP. (It's not totally safe but it'll work well enough for the types who would do it).
It's not hard. If you get mail from x@y.com then you contact y.com to verify it's real; indeed, you will need only make sure the connection is made from y.com in some cases (or all cases if so-designed). There is overhead there, but spam is more overhead. Do it this way and you can build up a domain blacklist for spammers.
But using the domain name system for addresses is pretty lame, considering it tends to make them unreliable (existentially).
Uh, duh, apollo is the moon. Hence, the Apollo mission...
Another ignorant capitalist saying Marx was against personal property.
Clue: communism is about social ownership of CAPITAL, the -means of production-. Not bicycles, but the factories that manufacture them.
It should also be noted that there is significant overlap; the old kernels are not completely unrelated to the new ones, and patches are both back-ported and forward-ported between them. The development of all trees benefits from the development of the other trees, in varying degrees. This also means that maintaining old kernels is not as much effort as maintaining the new ones (which is not to say that less effort is required of the people involved - but certainly fewer people are required to be involved).
Seems more likely the guy just forgot his story idea.
Uhh, I think there are actually about three or four hundred lines in between "make menuconfig" and "make bzImage".
You also have to manually upgrade all the packages to 2.6-supporting versions (by reading the README to find out what they are), and download and install the kernel source. Even if you wish to configure and compile the kernel yourself, Debian's package system will do this for you.
But the real issues involved in upgrading a kernel have nothing to do with how to install it. That is easy. LVM has been deprecated by a new version, devfs has been deprecated, ide-scsi has been deprecated, etc.: there are real issues to resolve that require changes -outside- the kernel and knowledge of the underlying issues.
Spelling is not an exact science either; the "correct" spelling of a word is a mixture of usage and history and it is not necessarily clear what to do when the two are inconsistent with each other or within themselves.
But this is all quite beside the point. The question was how can humans be inaccurate when humans define accuracy. That question is silly. Humans define accuracy, but they are not by definition accurate.
Please, Intel cannot force anyone out of the market by making incompatible CPUs. Every CPU they release introduces incompatibities! Along with every CPU released by every competitor. And compilers generate code for both their processors, and all their competitors'. And microsoft creates its software for both their processors and all their competitors'. It is not an evil practice; if you have the market significance to actually physically manufacture a general purpose CPU then you have the market significance to get software support for it (this is far, far cheaper). I've never heard of any case of Intel having exclusive support for any important software (not since DOS days anyway). People are just looking at Intel and seeing Microsoft, when really the practices of the two companies are not comparable, not on this point at least.
Intel is -not- a monopoly[1], and "embrace and extend" for instruction sets doesn't make any sense in today's CPU market, because the semantics of new instructions must be public (by convention and due to market realities), and so besides the fact that competitors are freely capable of implementing them, compilers are capable of generating equivalent code that doesn't use them. Intel is -not- trying to create a market of programs that won't run on AMD processors. The standard practice for use of new processor extensions in software is always -not- to break compatibility with older processors. Note that all Microsoft products are compatible with AMD and that Visual C++ as well as the rest of the Microsoft compiler suite generates code that is compatible with AMD. AMD -and- the other, smaller x86 competitors, mind you.
[1] This is not to say that none of their success is based on inertia. But come on - what software can you run on Intel that you can't run on AMD? Such software - badly written, unimportant programs - are the extent of Intel's monopoly. Right now, AMD makes slightly better processors, which is not enough to overcome Intel inertia; but if AMD made massively better processors, Intel would have no power in the market to force them not to be adopted; AMD would become top dog. The CPU market doesn't work like the software market.
Well, the math isn't wrong, but there is an unstated assumption, that whenever the two disagree, it will be possible to check and determine who is right. What multiple judges allows is better -detection- of errors, not in itself correction of errors. A majority scheme with an odd number of judges would indeed allow correction as well.
However, you've only actually increased the probability of error in any sense if you count a disagreement as an error - but there is no more reason to count a disagreement as an error than there is to count it as a correct judgment. It is, in fact, distinct from either. A disagreement implies no conclusion, so it cannot be right or wrong.
OK, but that's still braindead. If your house is on fire, and you don't pay or haven't payed to put it out, then what? You let it burn down the rest of the block, destroy all nearby power lines, sewage pipes, roads, trees? Firefighting in almost any location is a public good. Indeed, am I supposed to let my house burn up because my cheap neighbor got the slowest firefighting service in town and fell asleep smoking?
People who are actually responsible for fires are already legally liable for the damages they cause. Pushing liability further makes no sense[1], nor does making firefighting service optional.
[1] Free market people seem to love holding people responsible for things outside their control, perhaps because they want credit for their own privilege.
Another poster's analogy was incredibly apt. Computer spellcheckers also have higher accuracy than human ones. Think about it.
First of all, on any IM network, this is (for practical purposes) a client issue. Second, ICQ and AOL's clients and the popular free alternatives for their services also offer this.
If every individual human has an accuracy of 99.983%, then two independent humans have an accuracy of 1 - .00017^2 or 99.99999711%. This would allow ample accuracy to judge the computer, except that it's not true[1]. A better answer is the one you suggest: humans must judge spam from subject/author alone, whereas computers get to look at the whole message. Humans reading the whole message, and possibly even following included links, responding, etc., can be assumed to have full accuracy, within epistemic bounds. Indeed, merely re-checking your work, etc. - being consciously more diligent than the average spam-sorter - should insure your accuracy is better than average.
As for how accuracy was actually judged in this particular study, I suppose you would have to read the article for that. I haven't, myself...
[1] It assumes the probability of error is equal for every message, which is obviously not true (i.e., that error is random rather than systematic). The real accuracy of two humans in concert is surely much lower; OTOH, it is still sure to be much, much higher than the accuracy of a single human.
Unfortunately, if you build a module against a tree that does not have the same configuration as the binary kernel that will load it, it sometimes will not work. Sometimes, even adding the module to the existing kernel will require recompiling bzImage, else the module will not work (which is highly counter-intuitive). This is unfortunate, but I don't think the linux developers have made any attempt to allow modules to be compiled piecewise in general (it wouldn't be very hard), and there doesn't seem any way to find out which modules or base kernels will work without just trying it. Additionally, even if it does load, in at least one instance I've had a loaded module not work for this reason; it was a third-party module, but Linux makes no guarantee that its own modules will work under these circumstances either.
I've talked about the general issue and reasons I disagree with this perspective, elsewhere in the thread. I will only add here that, although it may seem perverse, I would actually prefer that it be difficult to "keep crime in check despite increasing poverty". In general, when it is easier to stop crime by ever-more-powerful law enforcement than by ameliorating the social causes of crime, I anticipate evils far greater than common crime, and I fear any technology which brings us further into that world.
It's true that, as a practical matter, privacy in public places cannot be obtained. That is no reason to say it is not desirable, or that greater privacy in public should not be encouraged. A camera on a street corner might be compared to an anonymous passer-by observing your day-to-day routine. Or it might be compared to an ominous figure lurking in the bushes, following you from a block or two behind, mysterously present every day. Unless we know how the cameras are being used, which we cannot, we do not know which comparison is more apt. The reality likely varies depending on the individual being taped.
One significant difference between public spaces with cameras and without is that, in general, in public spaces the observer is also public; he cannot hide from you any more than you from him. That seems to be a good check on particularly odious police monitoring; it is legal, for sure, but since it can be observed by anyone, the police are still checked by public opinion. There is no public opinion of secret police activity, though, and all monitoring via camera falls into this category.
Well, not everyone leads a boring life, and not every interesting life is criminal. For example, political activists are already closely monitored by the US government in legal and illegal ways. As a recent real-world example, I live in Connecticut and as you may have heard, we are having some problems with our governor accepting bribes, kickbacks, etc. Recently there was an open meeting of citizens seeking to hasten his removal from office, and a uniformed police officer showed up, gave his card to some activists whom he addressed by name (people who had certainly never met him), and generally spoke as much as possible, in an attempt to disrupt the meeting. Naturally, he was just trying to scare people by proving to them that they are being watched. But there is good reason for that to be scary, and it is likely that this information is being gathered for purposes beyond small-time intimidation tactics.
When the government knows what you're doing, even when it's legal, it can treat you differently for doing it, even when it's legal. This may take the form of petty harrassment, selective enforcement of commonly ignored laws, or something even more ominous. Obviously, you're right, we can't practically prevent the government from knowing about a certain amount of legal activity -- but we should not openly invite them to monitor all legal activity. Maybe that 10pm walk is to a political meeting; maybe it's to your gay lover's apartment; maybe it's to an AA meeting -- but if you're not breaking the law, it's none of the government's business.
well, again, a mistake of law is not a lie, and not the kind of thing that can get you in trouble anyway.
OTOH, Java shouldn't be used for a host of other reasons.
There is apparently a widespread myth that the FSF is giving out free legal services to GPL copyright holders. Nonsense! If the FSF owns a copyright, they -can- sue; but they have made no promise to protect any copyright assigned to them, and generally speaking do not have the capability to do so. The FSF is on the order of six people, including four programmers. They're not IBM. They don't have a legal task force. Eben Moglen is probably not licensed to practice in your state, assuming you're even American, and he's a full-time professor of law, -not- a full-time lawyer for FSF, so I don't know that he would even have the capacity to represent them in a lawsuit they -needed- to fight, let alone every petty copyright case thrown them. He hasn't represented them in a suit so far.
They are -already- in violation of the law. You can pursue that if they do not meet your terms. So, you -can- force them to release the code, or to give you money, or whatever, at the threat of a suit.
Also, you cannot "get in trouble" for mistakenly claiming to a company that it is in violation of the law. That is both absurd and unprecedented.
(If you file a police report or file for an injunction in court or something like that, then you can get in trouble. But only for lying; not for making a mistake of fact, and certainly not for making a mistake of law).
No. When you sue someone in a court that doesn't have jurisdiction over that someone, the court just throws it out. It doesn't offer some sort of meaningless default win for failure to appear.
Be realistic. Legislation is an easy way to generate social pressure,
far easier in general than rational argument. When the two come up
against each other, there is certainly no reason to assume a priori
that legislation will lose. Indeed, "social" pressure on the internet
has nothing to do with society; net standards are determined by a small
group of industry insiders, entry to which usually requires the favor
(and often employ) of large tech companies. Since tech companies need
government contracts and gov't support to survive, the government can,
if it chooses, directly dictate practices to them. Generally they would
meet no resistence, since they would be sure not to harm the business
no matter how they might harm the user. But government pressure
is ultimately impossible to resist anyway, if it is sufficiently
determined.