GPLv3's handling of license violations seems quite clear to me; it's quite explicit. I haven't read v2 recently, so I can't really comment on the situation there.
The GPLv3 is more explicit, though some of the same ambiguities remain as regard the effect of automatic revocation under section 8 followed by receipt of the license again under section 10. Normally, if your rights under a contract are revoked, permanently or provisionally, per its terms, and you get an offer of the same contract terms from the same offeror again, you are free to accept it. Under Section 10 of GPL v3, you automatically get a license whenever you receive software under the GPL. That you previously had a license which was revoked to the same software (or an earlier version of it) that were terminated due to breach would seem not to prevent you from accepting the license offered (by modifying or propagating the work, see Section 9) and having full rights associated with the license.
Every registered voter has a public / private key. Votes are digitally signed by the voters. Then after the election (or during), the signed messages are posted online. Voters would be able to see that their vote counted in the right direction, and unless someone else knows your private key, nobody would be able to tell who you voted for
Someone issued your public/private key combo, and probably required your identity when they provided it to you. That someone knows your private key.
A judge can (and will!) strike various parts of the agreement if he finds them to be in conflict with either the law or the intent of the agreement.
No, he won't.
He can.
The GPL deliberately omits the severability clause commonly found in contracts.
The absence of a severability clause does not necessarily stop a court from altering the terms of a contract in all cases. However, there may be limits on how far a court can reform a contract without invalidating it, which having a severability clause (which is an express agreement of the parties to accept alterations, in effect) may extend, and a clause expressly voiding the contract if it its terms are unenforceable (a clause the GPLv2 also lacks) may restrict.
This means that if any part of it is contrary to local law, the entire thing is void in that jurisdiction.
No, though an express clause stating that if any term is unenforceable the entire agreement is void might have that effect. It might not, either, and even if it did some of the terms even if ineffective as contract terms might have lingering effects under other theories other than contract per se.
Firstly there IS a print icon, its in the menu that appears when you click the Office icon top left, notice how on first start this button glows to attract your attention and there's even a friendly little pop-up that explains it's purpous, save, print, load, all of those functions not directly linked to editing the document
The one gripe I have with Office 2007 UI, is that some of the obvious things weren't put into the Quick Access toolbar by default. The office button as the one spot for obscure configuration and conversions, etc., is fine, but save/load/print ought to be out in the main interface.
But, compared to the 2003 UI, 2007 is a lot better.
Word: I am starting a new section of a document, and want to make the section header a different font. (something like \subsection{Justification} in TeX/LaTeX) Ok, in Word 2003, i needed to highlight the text, and then pick the font / font size that i had used previously.
In either Word 2003 or Word 2007, the easiest way to do this, particularly if you want consistent is to select an appropriate style from the Task Pane (2003) or the style selector on the Home pane of the Ribbon (2007). It is streamlined in 2007, though not as big of a difference as you suggest. You seem to be comparing manually changing the individual attributes in 2003 to using styles in 2007, but styles and a quick-access method of applying them is not new in 2007, though where it is located is new.
Another one is in Excel. I want to make a 3 series line plot in a worksheet. Ok, i can highlight the 3 different sets of cells. In Excel 2003, i can go throught he chart wizard and try to remember which style is what, etc. Or in 2007, I just click on the chart style i want directly from the ribbon.
In Excel 2003, there is an item on the Chart Toolbar that lets you do this, too, though again its streamlined in 2007 with the Ribbon. Again, you seem to be comparing the easy way in 2007 to the hard way (rather than the most comparable way) in 2003.
OTOH, since the easy way to do lots of things in 2003 is hidden away such that you need to be actively told about it or hunt your way through options at least once to find the easy way, the apples to oranges comparison isn't entirely inappropriate.
The GPL has had sufficient legal testing to show that it will hold up in court.
As I recall, there are only a handful of cases in a variety of jurisdictions that have gone to a final result even in a trial court, even fewer, if any, appellate court decisions, and no final decisions at all in some important jurisdictions (like, any in the United States) on the validity of the GPL.
That's a rather low bar for "sufficient legal testing".
This is far from pedantry. Using terms suggesting that someone would be convicted or held liable for a "GPL violation" suggests that a court must uphold the GPL as valid for such a lawsuit to be successful.
Its very odd to argue about what somebody "would" be held liable for when we have a case in which someone was held liable, and if the reporting is wrong, reference to the actual case and its actual holding should fairly unequivocally resolve the issue. Arguing that something would not happen based on a legal theory (and one for which no authority is offered) as a rebuttal to a news article claiming that a court has, in fact, done the thing in question is, well, somewhat dubious.
On the contrary, a standard legal action against someone not in compliance with the terms of the GPL would be a simple copyright infringement case
Strange, then, that in Progress Software v. MySQL AB, MySQL AB alleged a breach of the GPL rather than a "bare" copyright violation. Can you cite any case in which only a bare copyright violation was alleged, or in which a court took the perspective that that was the right way to analyze the issue? Or are you referring to a "standard" that applies only in your imagination, not in the real world?
It seems to me that the more usual case is to argue in the alternative, from the outset, that either no permission existed from the beginning, or that subsequent breach of the license giving permission terminated the permission (providing evidence that, if the agreement was formed, the defendants action did, in fact, violate it) and that subsequent distribution was a copyright violation, not to just allege a copyright violation without specific argument to breach of the license. IIRC, that's been the case in at least one of the German cases that have gone to final results, though, not reading German, I've only read third-party accounts and/or unofficial translations of material from those cases.
You *can* violate the terms of the GPL, if you do you lose your rights to distribute (since it is what gives you the rights).
This may be true (under paragraph 4 of the GPLv2), though its questionable what legal effect it has (since the GPLv2 is an open offer that you receive anew every time you receive software distributed under the license, and which you may accept, under its express terms, simply by redistributing the software.)
However you would not be convicted of doing so. Instead, you would be convicted of distributing without a license in violation of copyright (since you had no license to distribute).
Actually, it seems more likely that a breach after (perhaps simultaneously, in the event the initial distribution is noncompliant) acceptance of the GPLv2 (under paragraph 5, by distribution), would be found by the court, that the court would (perhaps) apply the remedy for breach in paragraph 4 which terminates the license from the moment of the breach, and then possibly consider subsequent distribution as a copyright violation.
OTOH, since under the GPLv2's own terms, the GPL is an offer, and one that one receives the GPL as a new offer everytime one received GPL licensed software, and can be accepted by distribution, its not entirely clear what affect termination has (certainly, the FSF has argued in various fora for it to constitute a permanent ban at least as regards the software improperly distributed, and providing the copyright holder doesn't "forgive" the past breach, but I don't think there is any case law supporting that view, and it is at least arguably contrary to the plain language of the license.)
Certainly, the main case I've seen on the issue (Progress Software v. MySQL AB) approached the issue from the perspective of a GPL breach not a "bare" copyright violation as MySQL's cause of action, though it never reached (before settlement) whether there was a breach and what legal effect the breach had and what remedies were available.
Noone would be convicted of "breaching the GPL". The GPL is not an EULA. If you violate the terms of the GPL, you are (re)distributing without a license permitting you to do so (since the GPL, which you violated, is the only thing that gives you permission to do so), which is a copyright violation, not a GPL violation.
This is a popular open-source myth, but doesn't seem true under the little US case law applying the GPL (consider Progress Software v. MySQL AB), or under the express terms of the GPLv2 (which make distribution, not merely compliant distribution, an acceptance of the GPL, which seem to constrain action for noncompliant distribution to action for breach of the license, not for a "bare" copyright violation.) It might be an accurate statement of German law, but I see no reason to believe that it is.
I wish articles would get the specifics right.
What evidence is that there article got the specifics wrong?
Equivalents to the first or second generation Palms should exist at well under the $50 mark, but they don't.
Low cost personal handheld electronic organizers with the kind of application the people in the market for those things want bundled in, but without the ability to load custom apps, exist at and below that price point. If ones that allow loading apps don't (I'm not sure whether or not they do), its because there is inadequate demand for it without the power of more modern PDAs, I suspect.
What kind of steps could the LKML folks take to reconnect with the average user? What steps could Linux communities take to connect their average users with the LKML?
It sounds like it would help to have a linux desktop support/advocacy community of some kind that could address desktop complaints of the type pointed to, identify things that are reasonably likely to relate to kernel issues, and work with users to get to produce useful bug reports to LKML (or, for non-kernel issues, point them to the right application vendor or other party to address the issue.) Enterprise users are likely to have their own facilities that can act as that kind of filter between non-technical end users and developers, and thus are more able to frame their issues in a way appropriate for kernel developers to act on (and avoid presenting issues that aren't within the kernel developer's domain).
I don't know if there is an existing group that tries to do this or would be a natural fit for it, or where the resources could plausibly come from to establish it, but I think that seems to be what is needed.
Yes; you'll note (presuming that you read the posts you respond to) that the omissions I point to from the webpage that you neglected to mention are neither embedded, nor the Enterprise editions of 2003, simply the non-embedded versions of XP and Vista that are listed on the Wikipedia page you cited but that you missed both in your guess and in your review of what you missed in your guess.
Outside of those exclusions, you claimed you missed only two versions of Vista, and the Server 2003 versions (though, as you note, the latter aren't entirely outside the exclusion).
But your mistake on Vista versions was more wrong than you claimed, and you missed a number of non-embedded XP versions that you neglected to mention.
As its ~$175 and its goal it is to reduce price while keeping the same or improved performance with every subsequent revision. Technological improvements are needed to make that happen. While they don't have to be "Moore's Law" improvements (# of transistors per unit area) precisely, they are the same general type of improvements. And, frankly, I don't think the XO and similar systems will reduce the demand for high-end computers. If anything, making basic computers (and experience with them) universally available will probably increase demand for high-end computers.
A case in point is the article on Harry Potter and the deathly hallows which as has been mentioned earlier in slashdot itself , is full of spoilers, posted inside a day of the book having been released. A lot of people who stumbled onto that article while looking for details on the book must have felt cheated.
So, you are saying, that people felt cheated when they deliberately searched for details of a book and found, horror of horrors, details of the book?
I used to love Wikipedia, but that incident made me realise it's nothing more than a starting point to get a very basic idea of a subject and then move on.
That's basically the essence of an encyclopedia, so what you've essentially said is "I used to love Wikipedia, until I realized that it is exactly what it claims to be."
I agree, but he is not talking comparison so much as absolute performance.
Since failure or success in a market is relative, absolute rather than relative performance would be a bad explanation for failure in the market.
Operating systems take too long to boot, period.
I agree with that. And, given that most OSes do much better, in terms of time, at restoring from "hibernation", you'd think this would be easy to fix—when you make a system change that affects startup, on the next startup, the system takes goes through a full startup and, before permitting user interaction, saves its state as if it were hibernating; in the future, until the next startup-affecting system change, it restores from that state image as if from hibernation, and continues on its merry way.
And, according to Microsoft, "Vista Home" does not, contrary to your list, identify a single version, the versions are "Vista Home Basic" and "Vista Home Premium", and you missed "Vista Business". Actually, the Wikipedia page you link to also points that out, so your description of what you missed is wrong even based on the source you used. You also missed, for Windows XP, again from the page you cite, Windows XP Starter, Windows XP Home Edition N, Windows XP Professional Edition N, Windows XP Professional Edition x64, Windows XP Tablet PC Edition, and, oddly enough, what was probably the most popular bundled version of XP in the last year or so before Vista, Windows XP Media Center Edition.
Did you even read the page you cited to claim that you only missed two Vista versions and the Server 2k3 versions?
One of the strengths of Linux is also its biggest weakness. If someone has a computer and for some strange reason needs to install an OS, which Linux distro do they choose? I've run Linux for years and I still can't name all the available distros. I doubt ANYONE can.
So? I can't name all the different kinds of laundry detergent (or even Tide laundry detergent), but its not hard for me to find one that works well enough and use it.
Why would I need to be able to name all of the Linux distros in order to use one? I don't need to use all of them.
Another problem is the MS dominance over the OS market. It's hard to buy a computer without Windows and even harder to purchase one with Linux preinstalled. Your average computer user is not going to purchase a computer that won't run (because of no OS) and even if they did, when they go to the store pick up an OS, all they see is Windows.
I'm pretty to sure that, e.g., Fry's sells both Linux pre-installed systems and retail-box versions of some of the commercial Linux distros. Dell sells systems with Linux pre-installed. And, of course, plenty of computers come with MacOS pre-installed, which, while it isn't Linux, is certainly "without Windows".
Linux users need to stick to a Distro that works, is easy, is well known, and comes as an option to be preinstalled on computers from the majority of manufacturers, even if it is along side Windows or as a bootable DVD thrown into the box.
Since no distribution exists that "comes as an option to be preinstalled on computers from the majority of manufacturers", and its clear that none will until the demand for Linux from consumers is great enough to force that option, you seem to be saying that to become popular, Linux must first be popular. (I disagree that Linux users have to stick to one distro for this to be the case; as long as there is application compatibility across distros, it wouldn't matter if each hardware manufacturer offered a different distro.)
But there is a kernel of truth there in that Linux won't acheive desktop competitiveness (in the market sense rather than the quality sense), until lots of people get exposed to Linux other than by choosing to buy it individually, so that they don't have to take the risk of what they perceive as "the unknown". Which means that the only way Linux will compete is the same way MS got its desktop dominance back in the days of DOS: providing enterprises a reason to standardize on it so that its what people need to use for work.
he's saying the Linux kernel is so bloated with enterprise level crap, and is so optimized for the server role, that it performs poorly on the desktop.
IME, it doesn't "perform poorly" on the desktop in the first place, at least compared to Windows XP (running on the exact same box), and where it does perform poorly, it seems to be mostly with audio/video (sometimes with other content, for similar reasons) and be primarily related to the use of substitute applications that aren't the ones the content was principally developed and tested for, which doesn't seem likely to be fundamentally a kernel problem.
How many OLPC staff were there handling the implementation? How many trained people per school will a full roll-out require? What hiccups occurred, what's being done to fix them? What is being done about Internet connections for the laptops (without which, bitfrost is not very able to deactivate stolen machines)? It's great that a few classrooms have had OLPC days, what are the results?
The pilot sites, except for Thailand where participation has ended, are ongoing and expanding, and intended to be absorbed into the full deployment of the systems in the countries where it is being piloted. For example, the Nigerian pilot site was initially 3 grade levels and expanding to 6; the Uruguay pilot is a full school.
Responses from parents, students, and teachers have been positive, and some issues have been identified and resolved through the process, so I'd say its worked. If you want more details, you probably need to go through the OLPC website yourself and, failing that, ask them.
"Seriously, ALMOST beating OS X's 6% market share when you are a predatory monopolist who has been cramming Vista down vendor's throats for six+ months now isn't something to be proud of."
If nearly beating Apple's market share after just six months is nothing to be proud of, what does that say for Linux?
One might note that the poster you are responding to only said ALMOST beating OS X's 6% market share was nothing to be proud of if you are a predatory monopolist. Since when did Linux have an OS monopoly?
So they are saying that someone doesn't qualify despite the fact that there are no rules as to what qualifies?
No. They are saying that there are rules, since the law explicitly requires experience. They are asking that the court interpret and apply that requirement to the specific case, and they are saying that they believe that under any reasonable interpretation, the incumbent falls outside of it.
Its not really uncommon at all for less than every word in a statute to be specifically defined in that statute.
If Google wins the auction, why would they need the Government to makes some rules about who can access it?
There won't be just one winner of the auction. Google is saying they'll bid only if the rules are that every winner must provide certain types of open access. They are basically attacking the premise that not requiring open access serves the public interest, at a minimum, by increasing the willingness of purchasers to bid for the spectrum.
Slashdot Mirror
The GPLv3 is more explicit, though some of the same ambiguities remain as regard the effect of automatic revocation under section 8 followed by receipt of the license again under section 10. Normally, if your rights under a contract are revoked, permanently or provisionally, per its terms, and you get an offer of the same contract terms from the same offeror again, you are free to accept it. Under Section 10 of GPL v3, you automatically get a license whenever you receive software under the GPL. That you previously had a license which was revoked to the same software (or an earlier version of it) that were terminated due to breach would seem not to prevent you from accepting the license offered (by modifying or propagating the work, see Section 9) and having full rights associated with the license.
Someone issued your public/private key combo, and probably required your identity when they provided it to you. That someone knows your private key.
He can.
The absence of a severability clause does not necessarily stop a court from altering the terms of a contract in all cases. However, there may be limits on how far a court can reform a contract without invalidating it, which having a severability clause (which is an express agreement of the parties to accept alterations, in effect) may extend, and a clause expressly voiding the contract if it its terms are unenforceable (a clause the GPLv2 also lacks) may restrict.
No, though an express clause stating that if any term is unenforceable the entire agreement is void might have that effect. It might not, either, and even if it did some of the terms even if ineffective as contract terms might have lingering effects under other theories other than contract per se.
Well, presumably, with a different name to avoid confusion with the existing operating system "Microsoft Windows Fundamentals for Legacy PCs".
The one gripe I have with Office 2007 UI, is that some of the obvious things weren't put into the Quick Access toolbar by default. The office button as the one spot for obscure configuration and conversions, etc., is fine, but save/load/print ought to be out in the main interface.
But, compared to the 2003 UI, 2007 is a lot better.
In either Word 2003 or Word 2007, the easiest way to do this, particularly if you want consistent is to select an appropriate style from the Task Pane (2003) or the style selector on the Home pane of the Ribbon (2007). It is streamlined in 2007, though not as big of a difference as you suggest. You seem to be comparing manually changing the individual attributes in 2003 to using styles in 2007, but styles and a quick-access method of applying them is not new in 2007, though where it is located is new.
In Excel 2003, there is an item on the Chart Toolbar that lets you do this, too, though again its streamlined in 2007 with the Ribbon. Again, you seem to be comparing the easy way in 2007 to the hard way (rather than the most comparable way) in 2003.
OTOH, since the easy way to do lots of things in 2003 is hidden away such that you need to be actively told about it or hunt your way through options at least once to find the easy way, the apples to oranges comparison isn't entirely inappropriate.
As I recall, there are only a handful of cases in a variety of jurisdictions that have gone to a final result even in a trial court, even fewer, if any, appellate court decisions, and no final decisions at all in some important jurisdictions (like, any in the United States) on the validity of the GPL.
That's a rather low bar for "sufficient legal testing".
Its very odd to argue about what somebody "would" be held liable for when we have a case in which someone was held liable, and if the reporting is wrong, reference to the actual case and its actual holding should fairly unequivocally resolve the issue. Arguing that something would not happen based on a legal theory (and one for which no authority is offered) as a rebuttal to a news article claiming that a court has, in fact, done the thing in question is, well, somewhat dubious.
Strange, then, that in Progress Software v. MySQL AB, MySQL AB alleged a breach of the GPL rather than a "bare" copyright violation. Can you cite any case in which only a bare copyright violation was alleged, or in which a court took the perspective that that was the right way to analyze the issue? Or are you referring to a "standard" that applies only in your imagination, not in the real world?
It seems to me that the more usual case is to argue in the alternative, from the outset, that either no permission existed from the beginning, or that subsequent breach of the license giving permission terminated the permission (providing evidence that, if the agreement was formed, the defendants action did, in fact, violate it) and that subsequent distribution was a copyright violation, not to just allege a copyright violation without specific argument to breach of the license. IIRC, that's been the case in at least one of the German cases that have gone to final results, though, not reading German, I've only read third-party accounts and/or unofficial translations of material from those cases.
This may be true (under paragraph 4 of the GPLv2), though its questionable what legal effect it has (since the GPLv2 is an open offer that you receive anew every time you receive software distributed under the license, and which you may accept, under its express terms, simply by redistributing the software.)
Actually, it seems more likely that a breach after (perhaps simultaneously, in the event the initial distribution is noncompliant) acceptance of the GPLv2 (under paragraph 5, by distribution), would be found by the court, that the court would (perhaps) apply the remedy for breach in paragraph 4 which terminates the license from the moment of the breach, and then possibly consider subsequent distribution as a copyright violation.
OTOH, since under the GPLv2's own terms, the GPL is an offer, and one that one receives the GPL as a new offer everytime one received GPL licensed software, and can be accepted by distribution, its not entirely clear what affect termination has (certainly, the FSF has argued in various fora for it to constitute a permanent ban at least as regards the software improperly distributed, and providing the copyright holder doesn't "forgive" the past breach, but I don't think there is any case law supporting that view, and it is at least arguably contrary to the plain language of the license.)
Certainly, the main case I've seen on the issue (Progress Software v. MySQL AB) approached the issue from the perspective of a GPL breach not a "bare" copyright violation as MySQL's cause of action, though it never reached (before settlement) whether there was a breach and what legal effect the breach had and what remedies were available.
This is a popular open-source myth, but doesn't seem true under the little US case law applying the GPL (consider Progress Software v. MySQL AB), or under the express terms of the GPLv2 (which make distribution, not merely compliant distribution, an acceptance of the GPL, which seem to constrain action for noncompliant distribution to action for breach of the license, not for a "bare" copyright violation.) It might be an accurate statement of German law, but I see no reason to believe that it is.
What evidence is that there article got the specifics wrong?
Low cost personal handheld electronic organizers with the kind of application the people in the market for those things want bundled in, but without the ability to load custom apps, exist at and below that price point. If ones that allow loading apps don't (I'm not sure whether or not they do), its because there is inadequate demand for it without the power of more modern PDAs, I suspect.
It sounds like it would help to have a linux desktop support/advocacy community of some kind that could address desktop complaints of the type pointed to, identify things that are reasonably likely to relate to kernel issues, and work with users to get to produce useful bug reports to LKML (or, for non-kernel issues, point them to the right application vendor or other party to address the issue.) Enterprise users are likely to have their own facilities that can act as that kind of filter between non-technical end users and developers, and thus are more able to frame their issues in a way appropriate for kernel developers to act on (and avoid presenting issues that aren't within the kernel developer's domain).
I don't know if there is an existing group that tries to do this or would be a natural fit for it, or where the resources could plausibly come from to establish it, but I think that seems to be what is needed.
Yes; you'll note (presuming that you read the posts you respond to) that the omissions I point to from the webpage that you neglected to mention are neither embedded, nor the Enterprise editions of 2003, simply the non-embedded versions of XP and Vista that are listed on the Wikipedia page you cited but that you missed both in your guess and in your review of what you missed in your guess.
Outside of those exclusions, you claimed you missed only two versions of Vista, and the Server 2003 versions (though, as you note, the latter aren't entirely outside the exclusion).
But your mistake on Vista versions was more wrong than you claimed, and you missed a number of non-embedded XP versions that you neglected to mention.
As its ~$175 and its goal it is to reduce price while keeping the same or improved performance with every subsequent revision. Technological improvements are needed to make that happen. While they don't have to be "Moore's Law" improvements (# of transistors per unit area) precisely, they are the same general type of improvements. And, frankly, I don't think the XO and similar systems will reduce the demand for high-end computers. If anything, making basic computers (and experience with them) universally available will probably increase demand for high-end computers.
So, you are saying, that people felt cheated when they deliberately searched for details of a book and found, horror of horrors, details of the book?
That's basically the essence of an encyclopedia, so what you've essentially said is "I used to love Wikipedia, until I realized that it is exactly what it claims to be."
Since failure or success in a market is relative, absolute rather than relative performance would be a bad explanation for failure in the market.
I agree with that. And, given that most OSes do much better, in terms of time, at restoring from "hibernation", you'd think this would be easy to fix—when you make a system change that affects startup, on the next startup, the system takes goes through a full startup and, before permitting user interaction, saves its state as if it were hibernating; in the future, until the next startup-affecting system change, it restores from that state image as if from hibernation, and continues on its merry way.
And, according to Microsoft, "Vista Home" does not, contrary to your list, identify a single version, the versions are "Vista Home Basic" and "Vista Home Premium", and you missed "Vista Business". Actually, the Wikipedia page you link to also points that out, so your description of what you missed is wrong even based on the source you used. You also missed, for Windows XP, again from the page you cite, Windows XP Starter, Windows XP Home Edition N, Windows XP Professional Edition N, Windows XP Professional Edition x64, Windows XP Tablet PC Edition, and, oddly enough, what was probably the most popular bundled version of XP in the last year or so before Vista, Windows XP Media Center Edition.
Did you even read the page you cited to claim that you only missed two Vista versions and the Server 2k3 versions?
So? I can't name all the different kinds of laundry detergent (or even Tide laundry detergent), but its not hard for me to find one that works well enough and use it.
Why would I need to be able to name all of the Linux distros in order to use one? I don't need to use all of them.
I'm pretty to sure that, e.g., Fry's sells both Linux pre-installed systems and retail-box versions of some of the commercial Linux distros. Dell sells systems with Linux pre-installed. And, of course, plenty of computers come with MacOS pre-installed, which, while it isn't Linux, is certainly "without Windows".
Since no distribution exists that "comes as an option to be preinstalled on computers from the majority of manufacturers", and its clear that none will until the demand for Linux from consumers is great enough to force that option, you seem to be saying that to become popular, Linux must first be popular. (I disagree that Linux users have to stick to one distro for this to be the case; as long as there is application compatibility across distros, it wouldn't matter if each hardware manufacturer offered a different distro.)
But there is a kernel of truth there in that Linux won't acheive desktop competitiveness (in the market sense rather than the quality sense), until lots of people get exposed to Linux other than by choosing to buy it individually, so that they don't have to take the risk of what they perceive as "the unknown". Which means that the only way Linux will compete is the same way MS got its desktop dominance back in the days of DOS: providing enterprises a reason to standardize on it so that its what people need to use for work.
IME, it doesn't "perform poorly" on the desktop in the first place, at least compared to Windows XP (running on the exact same box), and where it does perform poorly, it seems to be mostly with audio/video (sometimes with other content, for similar reasons) and be primarily related to the use of substitute applications that aren't the ones the content was principally developed and tested for, which doesn't seem likely to be fundamentally a kernel problem.
I wasn't aware Linux had failed on the desktop. Has it peaked and begun a clear terminal decline without ever reaching mass adoption?
Yes.
The pilot sites, except for Thailand where participation has ended, are ongoing and expanding, and intended to be absorbed into the full deployment of the systems in the countries where it is being piloted. For example, the Nigerian pilot site was initially 3 grade levels and expanding to 6; the Uruguay pilot is a full school.
Responses from parents, students, and teachers have been positive, and some issues have been identified and resolved through the process, so I'd say its worked. If you want more details, you probably need to go through the OLPC website yourself and, failing that, ask them.
One might note that the poster you are responding to only said ALMOST beating OS X's 6% market share was nothing to be proud of if you are a predatory monopolist. Since when did Linux have an OS monopoly?
No. They are saying that there are rules, since the law explicitly requires experience. They are asking that the court interpret and apply that requirement to the specific case, and they are saying that they believe that under any reasonable interpretation, the incumbent falls outside of it.
Its not really uncommon at all for less than every word in a statute to be specifically defined in that statute.
There won't be just one winner of the auction. Google is saying they'll bid only if the rules are that every winner must provide certain types of open access. They are basically attacking the premise that not requiring open access serves the public interest, at a minimum, by increasing the willingness of purchasers to bid for the spectrum.