While CR does good, unbiased reports on the boring everyday items such as blenders and vacuum cleaners, their car reviews are awful and hold no water except with people who like to drive cars that have no soul...
I dunno, I find them a good source of concrete relative comparisons rather than gushing hyperbole like most advertising-driven reviews from other sources. Even for cars. Yes, there are lots of subjective factors that go into car buying (or blender buying, for that matter), but subjective reviews by third parties (even if they were directed by a desire to maximize advertising revenue rather than honesty) aren't a good guide to those factors most of the time, anyway (the self-awareness to know what objective factors influence your subjective response to something allows objective reviews to sometimes be a good guide to that, and beyond that you've just got to try it yourself.)
I think you mean the distribution OF THE MEAN of the results.
No, I don't. That doesn't even make sense -- the mean has no distribution, its a single value.
What I mean is the is the distribution of the results. Just like I said.
Well, maybe not since these... composite securities... are composed of very connected parts.
I addressed that. Maybe you need to go reread what I wrote, paying particular attention to the two sentences using the phrase "to the extent that...", and the final sentence: "Assessing the degree to which two different investments are independent in their risks is, AFAIK, still more art than science to start with, and when the people doing the assessment often have financial interests (even if only indirectly) in promoting the sales of the packaged investments, well, the results are likely to represent those interests more than any rational assessment of reality."
But those three sentences are the whole post. If you missed them, what exactly are you responding to?
I'm no MBA but I would imagine that the rating of any composite security should be the lowest rating of the most risky component.
To the extent that different investments in a portfolio (which is what a "composite security" is, in essence, a prepackaged portfolio) have independent risks, there is a leveling effect (this is why, e.g., when you roll two dice, the distribution of the results is tighter proportionate to the range than when you roll one, and tighter still when you roll three, etc.)
OTOH, to the extent they tend to vary together, they don't level each other. Assessing the degree to which two different investments are independent in their risks is, AFAIK, still more art than science to start with, and when the people doing the assessment often have financial interests (even if only indirectly) in promoting the sales of the packaged investments, well, the results are likely to represent those interests more than any rational assessment of reality.
if access to the source mattered, he wouldn't be running Vista.
That doesn't follow. First of all, access to the source might "matter" for this application, but not the O/S (some people care about OSS for practical reasons that vary by the particular application, not out of pure ideology.) Further, access to the source could matter in both cases, but other concerns could apply to the O/S that outweighed it, and those concerns might not apply (or might not apply as strongly) to the application in question. Finally, the O/S may not be within the control of the poster at all.
You see using a database like SQLite, either with its normal console or the free GUI tools available as "overkill", so you want the OP to whip up a custom XML form and manipulate it with C++?
Those inevitably suck for almost anything where the data isn't guaranteed to be 100% consistently formatted. If a column is zip codes, you're probably OK. If it's a company name, say, you're good until the first time you need to store "Acme, Inc.". How do you escape commas? Do you quote the whole column? If so, then how do you escape quotes?
As long as you aren't taking data from some other unreliable source (which doesn't seem to be an issue here) that may not be doing the same thing, following section 2 of RFC 4180 seems to address all those issues cleanly. Of course, while its nice to have a readable format, its often nice also not to have to edit it directly, especially when quoting is an issue: that's why I'd recommend using something like Ruby + FasterCSV + (if you need reporting) Ruport for this; writing up a simple interface, either console or with any GUI toolkit you want, adapted to the use here should be fairly simple. I'd probably use SQLite via Sequel, rather than CSV via FasterCSV, if I was doing it for myself -- it may be overkill, but Sequel abstracts away the complexity you don't need fairly nicely, and its not like the resource use in this application is going to be substantial.
they have now asked for direct read-only access to our Oracle database, to be able to run ad-hoc queries without consulting us. As a DBA, my heart sinks at the thought of amateurs pawing through my database. Unfortunately, 'because you are stupid' is not considered a valid business reason to reject their request. So can any Slashdotters assist me in building my case to restrict access?
Instead of looking for reasons to say "no", you should put together a reasonable, first-blush estimate of what it would take you to say "yes" (in terms of putting together the necessary security, administrative controls, etc. to reasonably allow this without mixing customer data, allowing customers to starve the system, etc.), figure out what you'd have to charge the customers to make it worthwhile for you to do that, and tell them that they can have what they want -- if they pay the price (for which you will be able to provide the justification for the cost -- not necessarily a deatiled breakdown, but an overview.)
Then, they'll either say "no", and you don't have the hassle, or they'll say "yes" on terms that are worthwhile to you.
It sounds like you say that out of a bad experience with GPL'd software.
Well, no. What makes you think that?
However I don't think that is the only possible outcome.
I'm not describing an outcome at all, so I don't understand this response.
The anti-tivoization clause was probably controversial enough to encourage limiting its extent to a class of products. Is this off the point, or should I understand that the GPLv3 would seem more "free" to you if this clause applied to all products, and that the GPLv2 is a perfectly "free" license in your view?
I would say, rather, that the GPLv2 would seem to be closer to represent a coherent ideological view of "freedom" being advanced through the mechanism of a license, where the GPLv3 seems to be a compromise between ideologues and industrialists with very different reasons for supporting a "sticky" (to avoid the emotionally-load "v" word) open-source licensing system. Neither one is a perfect match for my idea of "freedom", though they (and other open-source licenses, including less-sticky ones) are, IMO, socially useful. As is simply making things available in the public domain. Developing a community and maintaining the open-licensed (or public domain) version is the important thing from the standpoint of keeping software useful and free, the particular license among the major kinds of open source license is probably less important. SQLite is public domain, plenty of major projects are under BSD/MIT-style licenses, and plenty are under various versions of the GPL and its cousins (LGPL, Affero, etc.) While the particular licensing model or copyright status may be important for some uses, I don't think any of them is generally superior; public domain is probably the "most free" by any reasonable definition, but that doesn't always make it the best one to use.
Again, the idea is to advance the community's interests, not a particular person's.
Advancing a particular parties concept of the community interest is not advancing the community interest. Individuals choosing to use offer licensing under the GPL represent only themselves. Anyway, plenty of people issue software under Free licenses, including the GPL, not out of ideology but out of a rational assessment that that best serves their financial interests.
The question is, do you think that using/modifying/redistributing software should be considered a fundamental right (possibly in connection with free speech itself, think non-disclosure agreements)?
Whether I do or not is irrelevant; the authors of the GPL clearly do not, otherwise many provisions of the GPL would be inexplicable, particularly the distinct treatment of consumer vs. business products in the anti-tivoization provisions of the GPLv3.
The GPL does not say that you must give your software away free of charge
No, it says you must give one particular form of software (the source code) away free of any additional charges. The "free as in beer" thing is the source code, and its "free as in beer" rather than "free as in speech", because, unlike free speech, what you get is given by another actor given to advance their own (sometimes financial, sometimes not) interests, not a fundamental, inalienable right that isn't dependent on a license from someone else.
I mean I like how Ubuntu can install things from the add/remove panel but what if I want Apache on my installation and they don't offer it?
In declining order of preference (not that Apache is a good example, what Linux distro doesn't have Apache in its official repo?):
(1) You find a third-party repository that has packages for it, and add that to the repository list for your system package manager, and install from there (you get an install that's easier than most Windows GUI installs, plus you get updates as they are posted to the repo); (2) You find a package (.deb for Ubuntu or other debian-based systems, RPM for Redhat, etc.) for your OS, and install from that (this is usually easier than a Windows installer, but no updates unless the application has its own updating functionality); (3) You download the source packages and build instructions from the project's homepage and build the software yourself, usually, this is fairly straightforward, though you have to manage dependencies manually, and you're still in a better position than on Windows, since at least the tools you need to build are almost certainly Free, free, and usually in the standard repositories for your OS [probably the base install, too].
I'm sure the server Distro has it but what of it.
If the "server distro" has it, its in the Ubuntu standard repositories, and you can just click on it in your package manager and install it. The different versions of Ubuntu (desktop vs. server, ubuntu vs. xubuntu vs. kubuntu vs. edubuntu) are all most remixes of material that is in the standard repositories once you install any of them, you can start from any and end up with any of the features of any of the others, essentially, just by adding packages that are in the standard repositories.
Leaving aside the clear paranoia that is causing you to characterise the NSA as "malicious", they would have to be not only malicious but downright stupid to put backdoors into open-source code.
Especially in open-source code designed to enable systems to meet certification requirements for sensitive US government use.
Eh. The GPL, propaganda aside, is already "free as in 'free (not as in speech)' beer", not "free as in speech". If it was "free as in speech", it wouldn't require a license.
Surely you don't think I can make a GPL program, and charge $10M for the code?
Sure you can. The GPL is not a sales contract, and so it does not set the terms of initial sale. It is a license. If you make a program and choose to license it under the GPL, you can still charge anything you want for it (though, after the first copy, you may be competing with the people you gave it to, who can distribute it, but only for free.) The GPL doesn't bind the creator at all; it is a gratuitous license.
Now, if you create a derived work from someone else's GPL-licensed program, and only have the right to distribute that at all because of the GPL, then you do face some restrictions. But even then you can charge anything you want for the software in whatever form you distribute it, you just can't charge anything extra (beyond, in certain cases, your own reasonable reproduction costs) for the source code beyond what you charged for the binary/object form.
In fact except for the treatment of prisoners and civilians (and casualties) in war, there isn't anything in any of the conventions.
That's a pretty big "except"; and the treatment of civilians part includes protections of civilian property against reprisals, germane to the immediate discussion.
Did you know that they really don't protect civilians under "contemporary" conditions ?
As it is quite impossible to know something that is not true, no.
It specifically states that if "the enemy" (anyone whom you're at war with) does not clearly identify itself (which is defined to mean military bases OUTSIDE of population centers and CLEARLY uniformed troops) that civilians, enemy troops AND casualties are fair game ?
This is not even remotely true. No such limitation exists on, e.g., the protection of civilians in Geneva (IV), particularly Article II thereof.
Besides, there isn't a single warring faction in the world today, except the United States (and Israel, Turkey and "maybe" China (insofar you call Tibet a war, besides I doubt you will find China respecting Geneva in Africa)), that even pretend to respect the Geneva conventions.
On the contrary, no nation that is a signatory of the Conventions, as most are, except the United States, has had high government officials state that it is not bound by the Geneva Conventions in regard to a particular international armed conflict.
Lots of other warring parties don't respect Geneva : the islamist government of Sudan, Egypt (in it's south), Iran, Pakistan,...
Since you seem to be referring to internal rather than international armed conflicts, which are generally outside of the scope of the Geneva Conventions, I think you are missing the point. There are certainly examples of states that have failed to do much to seem to adhere to the Conventions in international armed conflicts, but you don't seem to be focussing on those.
Never mind civilian computers being fair game. These conventions date from immediately after WWII (not that anyone really thinks Hitler would have respected them if they existed, in fact he would probably have used them to his advantage, but hey, one can hope, right ?)
One might note that the 1949 Geneva Conventions were essentially updates of older conventions (the first Geneva Convention of 1949 was an update of the Geneva Convention of 1864, the second Geneva Convention of 1949 was an update of the Hague Convention (X) of 1907, the third Geneva Convention of 1949 was an update of the Geneva Convention of 1929, and the fourth Geneva Convention of 1949 was, in part, an update of parts of the Hague Convention (IV) of 1907.) One might also note that various Axis military officers and officials were tried, convicted, and either imprisoned or executed for violations of those prior conventions and other provisions of international law. No law that exists will, without enforcement, merely by existing stop all possible violations.
Also let's not forget, article 29(3) of the Human Rights
Presumably, you are referring to the Universal Declaration of Human Rights, which is simply a declaration of principles. The actual substantive implementation of those principles is in other treaties.
In other words, anyone attempting to abolish the human rights treaty (one obvious party would be islamists) does not have any human rights.
Well, no, a more accurate interpretation would be that those attempting to violate the principles of the United Nations as delineated in the Charter (which waging aggressive war does but abolishing the UDHR itself may not), in so doing, exceed the boundaries of the rights declared by the UDHR; that does not mean that they lose any of the rights they have under
First, they use a legal theory that only a fool would pursue and that is, indeed, known for having been pursued foolishly only to be dismissed with a very clear finding by the judge in a U.S. court.
Munich is not, last I checked, governed by U.S. law. Neither the statute law, nor the legal system, is the same, so while a decision by a U.S. trial court might have some minimal, tangential bearing, its hardly a valid basis for claiming the legal strategy here is something "only a fool would pursue".
Then, they pursue this case when complying with the terms of the GPL would cost them nothing, which is the mark of a lawyer who isn't considering his client's best interest.
Actually, no. Even if assuming your description is correct and Skype would forego nothing of value by doing this, what course they pursue isn't a sign of that, since lawyers ultimately (if they are doing their job) pursue the course their client decides on, within certain ethical boundaries, even if it is not what the lawyer would advise.
It would probably be easier to have a sensible discussion of that if anywhere in TFA or even in the post to which TFA linked as its source there was any indication of the particular legal argument Skype was making.
Of course, even if we had that, the odds of a sensible discussion of German anti-trust law on Slashdot when the GPL is involved would be low.
Doesn't the GPL do the exact *opposite*?
No, the GPL does not do the opposite of violating anti-trust law, which would be enforcing anti-trust law.
The GPL in some ways lowers certain barriers to entry in markets, which would seem to broadly align with the policy goals notionally served by anti-trust laws. But they also impose other restrictions; whether those conflict with laws governing restraint of trade in any particular jurisdictions would be the kind of question that would require knowing the applicable laws in the jurisdiction.
Were I extremely pedantic I could claim OOo3 no longer supports ISO standard ODF... but that would be silly.
That would not seem to be extremely pedantic so much as extremely wrong. While OOo3 does support ODF 1.2 which is not yet an ISO standard, it does not do so exclusively. There is a drop down in the Options UI to select the ODF version to use.
The motivation for widespread parallel programming seems to be that there is this upcoming glut of multicore PC chips that will get wasted if we all don't start writing concurrent programs.
Another reason is distributed programming. Often (as is the case with the Actor model) the same solutions that address problems with traditional threading on one machine (and thus make local parallel programs that can take advantage of multicore CPUs cleaner to develop) also apply directly to distributed programming across many machines.
But, in any case, I think its pretty clear that multicore chips are going to become more common and the number of cores on the average machine are going to be increasing for a while, so there certainly will be some increase in the applications for local parallelism. How far that will go is hard to say.
This is true, but TFA is still marginally useful in that the comments present a link to a more useful article which reflects the same general viewpoint as TFA but does make an argument and present alternatives:
I'm all for getting rid of threads, but what are you going to replace them with? Traditional functional languages may be the most obvious solution, but they're also among the most impractical of solutions.
"Traditional functional languages" are not a replacement for threads; functional languages may or may feature concurrency and if they do they can feature just about any concurrency model. A concurrency model with shared-nothing processes communicating through "mailboxes" (the Actor model) is a frequently recommended replacement, and some of the languages that embrace that as their concurrency model are functional languages (e.g., Erlang and Oz), but some are not (e.g. E, Io.) One advantage of the actor model is that it naturally extends to non-local distributed computing as well as local concurrency.
I haven't studied Erlang yet, but threads (or more generally concurrency) done securely would require mandatory locking of all data.
Erlang concurrency works through lightweight processes that don't share staet and can communicate only through special dedicated communications channels (or at least that's the high-level view; on a low level, I assume, though I haven't looked at the implementation, that in fact the communication channels have to be implemented as shared state with locking enforced by the language implementation.) Consequently, "thread-safety" is rendered into a non-issue.
Turning over the entire leadership team to corporate pawns and stripping out everything that makes the platform special is not going to help.
The OLPC project may now have a long life ahead in its new rule, supported "charitable" corporate donations, operating basically as a notionally charitable marketing firm for certain large commercial software firms.
So the change may help, just not help the people the OLPC project was started to help.
Slashdot Mirror
I dunno, I find them a good source of concrete relative comparisons rather than gushing hyperbole like most advertising-driven reviews from other sources. Even for cars. Yes, there are lots of subjective factors that go into car buying (or blender buying, for that matter), but subjective reviews by third parties (even if they were directed by a desire to maximize advertising revenue rather than honesty) aren't a good guide to those factors most of the time, anyway (the self-awareness to know what objective factors influence your subjective response to something allows objective reviews to sometimes be a good guide to that, and beyond that you've just got to try it yourself.)
F = ma
HTH
No, I don't. That doesn't even make sense -- the mean has no distribution, its a single value.
What I mean is the is the distribution of the results. Just like I said.
I addressed that. Maybe you need to go reread what I wrote, paying particular attention to the two sentences using the phrase "to the extent that...", and the final sentence: "Assessing the degree to which two different investments are independent in their risks is, AFAIK, still more art than science to start with, and when the people doing the assessment often have financial interests (even if only indirectly) in promoting the sales of the packaged investments, well, the results are likely to represent those interests more than any rational assessment of reality."
But those three sentences are the whole post. If you missed them, what exactly are you responding to?
To the extent that different investments in a portfolio (which is what a "composite security" is, in essence, a prepackaged portfolio) have independent risks, there is a leveling effect (this is why, e.g., when you roll two dice, the distribution of the results is tighter proportionate to the range than when you roll one, and tighter still when you roll three, etc.)
OTOH, to the extent they tend to vary together, they don't level each other. Assessing the degree to which two different investments are independent in their risks is, AFAIK, still more art than science to start with, and when the people doing the assessment often have financial interests (even if only indirectly) in promoting the sales of the packaged investments, well, the results are likely to represent those interests more than any rational assessment of reality.
That doesn't follow. First of all, access to the source might "matter" for this application, but not the O/S (some people care about OSS for practical reasons that vary by the particular application, not out of pure ideology.) Further, access to the source could matter in both cases, but other concerns could apply to the O/S that outweighed it, and those concerns might not apply (or might not apply as strongly) to the application in question. Finally, the O/S may not be within the control of the poster at all.
You see using a database like SQLite, either with its normal console or the free GUI tools available as "overkill", so you want the OP to whip up a custom XML form and manipulate it with C++?
As long as you aren't taking data from some other unreliable source (which doesn't seem to be an issue here) that may not be doing the same thing, following section 2 of RFC 4180 seems to address all those issues cleanly. Of course, while its nice to have a readable format, its often nice also not to have to edit it directly, especially when quoting is an issue: that's why I'd recommend using something like Ruby + FasterCSV + (if you need reporting) Ruport for this; writing up a simple interface, either console or with any GUI toolkit you want, adapted to the use here should be fairly simple. I'd probably use SQLite via Sequel, rather than CSV via FasterCSV, if I was doing it for myself -- it may be overkill, but Sequel abstracts away the complexity you don't need fairly nicely, and its not like the resource use in this application is going to be substantial.
Instead of looking for reasons to say "no", you should put together a reasonable, first-blush estimate of what it would take you to say "yes" (in terms of putting together the necessary security, administrative controls, etc. to reasonably allow this without mixing customer data, allowing customers to starve the system, etc.), figure out what you'd have to charge the customers to make it worthwhile for you to do that, and tell them that they can have what they want -- if they pay the price (for which you will be able to provide the justification for the cost -- not necessarily a deatiled breakdown, but an overview.)
Then, they'll either say "no", and you don't have the hassle, or they'll say "yes" on terms that are worthwhile to you.
Well, no. What makes you think that?
I'm not describing an outcome at all, so I don't understand this response.
I would say, rather, that the GPLv2 would seem to be closer to represent a coherent ideological view of "freedom" being advanced through the mechanism of a license, where the GPLv3 seems to be a compromise between ideologues and industrialists with very different reasons for supporting a "sticky" (to avoid the emotionally-load "v" word) open-source licensing system. Neither one is a perfect match for my idea of "freedom", though they (and other open-source licenses, including less-sticky ones) are, IMO, socially useful. As is simply making things available in the public domain. Developing a community and maintaining the open-licensed (or public domain) version is the important thing from the standpoint of keeping software useful and free, the particular license among the major kinds of open source license is probably less important. SQLite is public domain, plenty of major projects are under BSD/MIT-style licenses, and plenty are under various versions of the GPL and its cousins (LGPL, Affero, etc.) While the particular licensing model or copyright status may be important for some uses, I don't think any of them is generally superior; public domain is probably the "most free" by any reasonable definition, but that doesn't always make it the best one to use.
Advancing a particular parties concept of the community interest is not advancing the community interest. Individuals choosing to use offer licensing under the GPL represent only themselves. Anyway, plenty of people issue software under Free licenses, including the GPL, not out of ideology but out of a rational assessment that that best serves their financial interests.
Whether I do or not is irrelevant; the authors of the GPL clearly do not, otherwise many provisions of the GPL would be inexplicable, particularly the distinct treatment of consumer vs. business products in the anti-tivoization provisions of the GPLv3.
No, it says you must give one particular form of software (the source code) away free of any additional charges. The "free as in beer" thing is the source code, and its "free as in beer" rather than "free as in speech", because, unlike free speech, what you get is given by another actor given to advance their own (sometimes financial, sometimes not) interests, not a fundamental, inalienable right that isn't dependent on a license from someone else.
In declining order of preference (not that Apache is a good example, what Linux distro doesn't have Apache in its official repo?):
(1) You find a third-party repository that has packages for it, and add that to the repository list for your system package manager, and install from there (you get an install that's easier than most Windows GUI installs, plus you get updates as they are posted to the repo);
(2) You find a package (.deb for Ubuntu or other debian-based systems, RPM for Redhat, etc.) for your OS, and install from that (this is usually easier than a Windows installer, but no updates unless the application has its own updating functionality);
(3) You download the source packages and build instructions from the project's homepage and build the software yourself, usually, this is fairly straightforward, though you have to manage dependencies manually, and you're still in a better position than on Windows, since at least the tools you need to build are almost certainly Free, free, and usually in the standard repositories for your OS [probably the base install, too].
If the "server distro" has it, its in the Ubuntu standard repositories, and you can just click on it in your package manager and install it. The different versions of Ubuntu (desktop vs. server, ubuntu vs. xubuntu vs. kubuntu vs. edubuntu) are all most remixes of material that is in the standard repositories once you install any of them, you can start from any and end up with any of the features of any of the others, essentially, just by adding packages that are in the standard repositories.
Especially in open-source code designed to enable systems to meet certification requirements for sensitive US government use.
Eh. The GPL, propaganda aside, is already "free as in 'free (not as in speech)' beer", not "free as in speech". If it was "free as in speech", it wouldn't require a license.
Sure you can. The GPL is not a sales contract, and so it does not set the terms of initial sale. It is a license. If you make a program and choose to license it under the GPL, you can still charge anything you want for it (though, after the first copy, you may be competing with the people you gave it to, who can distribute it, but only for free.) The GPL doesn't bind the creator at all; it is a gratuitous license.
Now, if you create a derived work from someone else's GPL-licensed program, and only have the right to distribute that at all because of the GPL, then you do face some restrictions. But even then you can charge anything you want for the software in whatever form you distribute it, you just can't charge anything extra (beyond, in certain cases, your own reasonable reproduction costs) for the source code beyond what you charged for the binary/object form.
Anti-trust laws are about as uniquely American as anti-murder laws.
That's a pretty big "except"; and the treatment of civilians part includes protections of civilian property against reprisals, germane to the immediate discussion.
As it is quite impossible to know something that is not true, no.
This is not even remotely true. No such limitation exists on, e.g., the protection of civilians in Geneva (IV), particularly Article II thereof.
On the contrary, no nation that is a signatory of the Conventions, as most are, except the United States, has had high government officials state that it is not bound by the Geneva Conventions in regard to a particular international armed conflict.
Since you seem to be referring to internal rather than international armed conflicts, which are generally outside of the scope of the Geneva Conventions, I think you are missing the point. There are certainly examples of states that have failed to do much to seem to adhere to the Conventions in international armed conflicts, but you don't seem to be focussing on those.
One might note that the 1949 Geneva Conventions were essentially updates of older conventions (the first Geneva Convention of 1949 was an update of the Geneva Convention of 1864, the second Geneva Convention of 1949 was an update of the Hague Convention (X) of 1907, the third Geneva Convention of 1949 was an update of the Geneva Convention of 1929, and the fourth Geneva Convention of 1949 was, in part, an update of parts of the Hague Convention (IV) of 1907.) One might also note that various Axis military officers and officials were tried, convicted, and either imprisoned or executed for violations of those prior conventions and other provisions of international law. No law that exists will, without enforcement, merely by existing stop all possible violations.
Presumably, you are referring to the Universal Declaration of Human Rights, which is simply a declaration of principles. The actual substantive implementation of those principles is in other treaties.
Well, no, a more accurate interpretation would be that those attempting to violate the principles of the United Nations as delineated in the Charter (which waging aggressive war does but abolishing the UDHR itself may not), in so doing, exceed the boundaries of the rights declared by the UDHR; that does not mean that they lose any of the rights they have under
Munich is not, last I checked, governed by U.S. law. Neither the statute law, nor the legal system, is the same, so while a decision by a U.S. trial court might have some minimal, tangential bearing, its hardly a valid basis for claiming the legal strategy here is something "only a fool would pursue".
Actually, no. Even if assuming your description is correct and Skype would forego nothing of value by doing this, what course they pursue isn't a sign of that, since lawyers ultimately (if they are doing their job) pursue the course their client decides on, within certain ethical boundaries, even if it is not what the lawyer would advise.
It would probably be easier to have a sensible discussion of that if anywhere in TFA or even in the post to which TFA linked as its source there was any indication of the particular legal argument Skype was making.
Of course, even if we had that, the odds of a sensible discussion of German anti-trust law on Slashdot when the GPL is involved would be low.
No, the GPL does not do the opposite of violating anti-trust law, which would be enforcing anti-trust law.
The GPL in some ways lowers certain barriers to entry in markets, which would seem to broadly align with the policy goals notionally served by anti-trust laws. But they also impose other restrictions; whether those conflict with laws governing restraint of trade in any particular jurisdictions would be the kind of question that would require knowing the applicable laws in the jurisdiction.
That would not seem to be extremely pedantic so much as extremely wrong. While OOo3 does support ODF 1.2 which is not yet an ISO standard, it does not do so exclusively. There is a drop down in the Options UI to select the ODF version to use.
Another reason is distributed programming. Often (as is the case with the Actor model) the same solutions that address problems with traditional threading on one machine (and thus make local parallel programs that can take advantage of multicore CPUs cleaner to develop) also apply directly to distributed programming across many machines.
But, in any case, I think its pretty clear that multicore chips are going to become more common and the number of cores on the average machine are going to be increasing for a while, so there certainly will be some increase in the applications for local parallelism. How far that will go is hard to say.
This is true, but TFA is still marginally useful in that the comments present a link to a more useful article which reflects the same general viewpoint as TFA but does make an argument and present alternatives:
http://www.eecs.berkeley.edu/Pubs/TechRpts/2006/EECS-2006-1.pdf
"Traditional functional languages" are not a replacement for threads; functional languages may or may feature concurrency and if they do they can feature just about any concurrency model. A concurrency model with shared-nothing processes communicating through "mailboxes" (the Actor model) is a frequently recommended replacement, and some of the languages that embrace that as their concurrency model are functional languages (e.g., Erlang and Oz), but some are not (e.g. E, Io.) One advantage of the actor model is that it naturally extends to non-local distributed computing as well as local concurrency.
Erlang concurrency works through lightweight processes that don't share staet and can communicate only through special dedicated communications channels (or at least that's the high-level view; on a low level, I assume, though I haven't looked at the implementation, that in fact the communication channels have to be implemented as shared state with locking enforced by the language implementation.) Consequently, "thread-safety" is rendered into a non-issue.
The OLPC project may now have a long life ahead in its new rule, supported "charitable" corporate donations, operating basically as a notionally charitable marketing firm for certain large commercial software firms.
So the change may help, just not help the people the OLPC project was started to help.