There's a *huge* difference between a refurbished machine with an obsolete space-heater processor and a new machine with a SOTA low-energy processor. I'm not trying to defend Wal-Mart here, nor do I have anything specifically against refurbished stuff, but it's simply not an apples to apples comparison.
I personally don't, I guess I haven't watched the right documentaries yet, to tell me what to think, or something.
Another way to read this is simply that you haven't educated yourself enough to form a valid opinion.
Sure, lots of people complain about lots of things. Many of those people are blatantly incorrect, but some of them are right. The thing is, there's no way to tell the difference until you look into it. If you assume that someone is wrong without actually bothering to understand the issue, you're the one who's blatantly incorrect.
I'm not specifically talking about Wal-mart here. I haven't looked into it enough to have a valid opinion. But I'm not going to arbitrarily assume that people who are going out of their way to complain about an issue are just gibbering idiots.
Go through the tabs and click the "expand all" button in each of them.
Firefox hasn't frozen, it's just processing AJAX really slowly.
Personally, I see this as a bug in firefox (it should do javascript and user events in separate threads so it stays responsive), but a faster machine helps quite a bit.
Sounds right. Software patents expire once the technique involved is blatantly obsolete. The only exception that I know if is the RSA patent - that just set back computer security by 20 years outright.
Sure, as long as they pay the appropriate patent license fees to all the necessary parties.
This is why you hear complaints about software patents - they make it so that some standard file formats can only be legally implemented in software as payware.
I'm not so sure. One of the things I've heard is that GPLv3 will create hardware vender lockout. Because of clauses in v3 they won't move to v3. It's hard enough the get hardware venders to release drivers for Linux, with v3 they won't period.
The GPLv3 doesn't actually contain any clauses that would make that response appropriate. The only reason we'd see behavior like that is if they believe the anti-GPLv3 FUD. Companies with software patents may want to avoid distributing GPLv3 code, but that doesn't prevent them from releasing drivers or specs in any way - they can simply license their drivers permissively, with a proprietary license, or with the clever "GPLv2 or later".
They aren't the good guys. In fact, they are the biggest patent abusing bastards in the world. But... they'd take a revenue hit if people were afraid of deploying their favorite commodity UNIX (i.e. GNU/Linux), so they're likely to step in and maul anyone who actually attacks it with software patents.
And yes, they can win a patent war against Microsoft.
I already mentioned wireless devices (but is a company going to even create a blob for GPL3 code if there is more uncertainty that the blob could open them up to patent issues, federal regulation issues, etc?).
When it comes to driver blobs, the GPLv3 is no different than GPLv2 - they're not allowed but the developer can make an explicit exception. If the Linux kernel moved to GPLv3, no wireless device manufacturers would need to change their policies at all. Linus would probably want to make his binary module exception explicit, but that's an existing problem with the current license. Binary firmware would remain completely fine.
The only legitimate issue with GPLv3 seems to be the case of embedded devices with regulatory limitations on end user modification. This is only relevant in a few very specific areas, and there are a number of ways to work around the problem. Personally, I can't see any reason to make a big deal about this - just add an "additional permission" to your code if you're really worried about it.
As for an updating being needed, how about trying to make it a compatible update rather than setting out from the start to make it an incompatible one? Because you'd be wasting your opportunity to use the peak popularity of your software for your ideology?
It is impossible to make a new copyleft license with new requirements that is compatible with the old version. Again, bringing the GPL's patent text in to line with every other major "open source" license required an incompatible change.
I'll keep using GPL2 and probably will release future stuff I write under the GPL2. I really hope some of the major players fork the GNU tools to keep them at GPL2. If not, I'm sure we're going to see more attention going to projects like busybox and ulibc that the FSF doesn't control. In fact, I'd really savor it if the FSF made themselves totally irrelevant by willfully fracturing the community.
Even if there is a GPLv2 fork of some of these projects, it's extremely unlikely that it will become the dominant version. Companies like Sun and IBM put a lot of work into getting the patent clauses right in GPLv3 - they're not going to want to screw around with an obsolete license if they don't have to.
There is a tradition in many newsgroups and other Internet discussion forums that once such a comparison is made, the thread is finished and whoever mentioned the Nazis has automatically "lost" whatever debate was in progress. This principle is itself frequently referred to as Godwin's Law.
Through the Tivoization clause, it bans use in embedded medical devices, for example.
You say "for example", but that is the *only* example - and that only because of FDA regulations that are innately in conflict with the idea of free software.
The compatibility issue for source code exchange does exist, but it's an unavoidable problem that occurs when a copyleft license is updated. The recommended use of the GPLv2 was to say "GPLv2 or later" for exactly this reason. And the update was necessary, if for no other reason than to bring the GPL into line with other licenses on the issue of patent protection.
If you were really attached to the GPLv2 specifically, then it really sucks - but that is no longer the license for FSF copyrighted code. If you want to use new FSF code, you're going to have to play by the FSF rules.
The licensing issues are a nightmare, especially because such an important project is having a license changed rammed through in a two week time frame for political reasons instead of technical ones.
Licensing decisions are usually made for legal reasons rather than technical ones. The GPLv3 is being adopted because it fixes a number of legal weaknesses in GPLv2 - that's why it was developed, and that's why it's being adopted.
However, the result is... yes, GPL2 and GPL3 programs can run side by side on the same box and even be shipped on the same CD/DVD. Outside of that though, they are totally incompatible and that drives a wedge into the community.
They're exactly as incompatible as any other sets of code with two licenses. This is like saying that GPLv2 and the MPL are incompatible - yea, they are, and the community has been coping with that sort of problem for years.
There are other countries where democracy works great. Smaller countries, in Europe, with proportional representation, where a couple hundred thousand votes for a political party gets them multiple members of parliament.
Here in the USA though, I have a sneaking suspicion that our "you can have 49% of millions of voters and still get nothing" system of elections thwarts the whole idea of functional democracy pretty effectively.
I'm really starting to wonder if these posts are anti-GPLv3 FUD, or the result of misunderstandings created by previous anti-GPLv3 FUD...
This is very simple: If you are not personally planning on redistributing GPLv3 software in such a way that the recipient can't modify it, you don't need to worry about problems from the license. In fact, as an end user you should welcome GPLv3 because it defends you better from patent problems.
At this point, I'm personally willing to suspend Godwins law for discussions about American politics. The more that people look at fascism and the USA next to each other, the more likely it is that we'll be able to fix some of the disturbing similarities.
You comments are cultural rather than rational. The whole idea that you don't want to work for the benefit of others is a standard lower-class American belief - that mindset will get you from "poor" to "upper middle class" in the USA (or at least provide comfort when you fail), so it's competitively advantageous for you to have it.
Other cultures promote different mindsets, many of which are also sustainable in practice. I don't know if there is a mindset that would work in a democratic communist system, but I wouldn't want to simply assume that things are impossible simply because they wouldn't mesh cleanly with my cultural background.
The thing that makes democracy work in a society like the US is accountability.
That's the optimistic view.
It seems to me that maintaining a functional system of accountability for any official of the federal government in a country of more than 300 million people may be impossible - especially if you have an economy structured the way the US economy is.
The highest estimate for civilian deaths in Iraq that I've heard is 600,000. If we consider the existence of the war in Afganistan, that implies that we *could* get to an estimate 1,000,000 using the method you describe.
Just to be completely clear: The GPLv3 does not cause any compatibility issue for different programs working together. Your car analogy is nonsensical, because the GNU code and the Linux kernel can happily be under different licenses and still work together fine.
The GPL3 is less about using software how you want to, as long as you keep the source open, and more a means of forcing developers into a political ideology.
This sounds like unsubstantiated FUD. The GPLv3 doesn't force anyone into a political ideology anymore than GPLv2 did. If you have some specific, practical problem with a specific part of the GPLv3 then we can have a reasoned discussion, but your comments are as vague and non-specific as possible.
Slashdot Mirror
There's a *huge* difference between a refurbished machine with an obsolete space-heater processor and a new machine with a SOTA low-energy processor. I'm not trying to defend Wal-Mart here, nor do I have anything specifically against refurbished stuff, but it's simply not an apples to apples comparison.
That's horrible logic.
Another way to read this is simply that you haven't educated yourself enough to form a valid opinion.
Sure, lots of people complain about lots of things. Many of those people are blatantly incorrect, but some of them are right. The thing is, there's no way to tell the difference until you look into it. If you assume that someone is wrong without actually bothering to understand the issue, you're the one who's blatantly incorrect.
I'm not specifically talking about Wal-mart here. I haven't looked into it enough to have a valid opinion. But I'm not going to arbitrarily assume that people who are going out of their way to complain about an issue are just gibbering idiots.
MadTux.org has been selling a very similar box with no OS for a while:& products_id=229
http://store.madtux.org/product_info.php?cPath=57
Personally, I see this as a bug in firefox (it should do javascript and user events in separate threads so it stays responsive), but a faster machine helps quite a bit.
I distinctly remember having a K6-III 700mhz. I'm not sure that they got to 900mhz, but it wasn't that far off.
Sounds right. Software patents expire once the technique involved is blatantly obsolete. The only exception that I know if is the RSA patent - that just set back computer security by 20 years outright.
Sure, as long as they pay the appropriate patent license fees to all the necessary parties.
This is why you hear complaints about software patents - they make it so that some standard file formats can only be legally implemented in software as payware.
Those aren't license terms. Those are notices. You don't have to accept them - they simply are.
Whole teams of lawyers spent almost two years on this license. If there were blatant mechanical problems with it, they'd have noticed long ago.
Have you actually read GPLv3? It doesn't have a web services clause in it. They split that out into a completely separate license.
The GPLv3 doesn't actually contain any clauses that would make that response appropriate. The only reason we'd see behavior like that is if they believe the anti-GPLv3 FUD. Companies with software patents may want to avoid distributing GPLv3 code, but that doesn't prevent them from releasing drivers or specs in any way - they can simply license their drivers permissively, with a proprietary license, or with the clever "GPLv2 or later".
There is a player you're forgetting about: IBM.
They aren't the good guys. In fact, they are the biggest patent abusing bastards in the world. But... they'd take a revenue hit if people were afraid of deploying their favorite commodity UNIX (i.e. GNU/Linux), so they're likely to step in and maul anyone who actually attacks it with software patents.
And yes, they can win a patent war against Microsoft.
Knowing the kernel guys, you'd get a lot of people moving their code to "GPLv2 or GPLv3" rather than "GPLv2 or later" - which is fine.
There's a whole *section* in GPLv3 to cover this. From section 9:
Stop spreading misinformation.
When it comes to driver blobs, the GPLv3 is no different than GPLv2 - they're not allowed but the developer can make an explicit exception. If the Linux kernel moved to GPLv3, no wireless device manufacturers would need to change their policies at all. Linus would probably want to make his binary module exception explicit, but that's an existing problem with the current license. Binary firmware would remain completely fine.
The only legitimate issue with GPLv3 seems to be the case of embedded devices with regulatory limitations on end user modification. This is only relevant in a few very specific areas, and there are a number of ways to work around the problem. Personally, I can't see any reason to make a big deal about this - just add an "additional permission" to your code if you're really worried about it.
It is impossible to make a new copyleft license with new requirements that is compatible with the old version. Again, bringing the GPL's patent text in to line with every other major "open source" license required an incompatible change.
Even if there is a GPLv2 fork of some of these projects, it's extremely unlikely that it will become the dominant version. Companies like Sun and IBM put a lot of work into getting the patent clauses right in GPLv3 - they're not going to want to screw around with an obsolete license if they don't have to.
From that article:
That's the usage I was using.
You say "for example", but that is the *only* example - and that only because of FDA regulations that are innately in conflict with the idea of free software.
The compatibility issue for source code exchange does exist, but it's an unavoidable problem that occurs when a copyleft license is updated. The recommended use of the GPLv2 was to say "GPLv2 or later" for exactly this reason. And the update was necessary, if for no other reason than to bring the GPL into line with other licenses on the issue of patent protection.
If you were really attached to the GPLv2 specifically, then it really sucks - but that is no longer the license for FSF copyrighted code. If you want to use new FSF code, you're going to have to play by the FSF rules.
Licensing decisions are usually made for legal reasons rather than technical ones. The GPLv3 is being adopted because it fixes a number of legal weaknesses in GPLv2 - that's why it was developed, and that's why it's being adopted.
They're exactly as incompatible as any other sets of code with two licenses. This is like saying that GPLv2 and the MPL are incompatible - yea, they are, and the community has been coping with that sort of problem for years.
There are other countries where democracy works great. Smaller countries, in Europe, with proportional representation, where a couple hundred thousand votes for a political party gets them multiple members of parliament.
Here in the USA though, I have a sneaking suspicion that our "you can have 49% of millions of voters and still get nothing" system of elections thwarts the whole idea of functional democracy pretty effectively.
I'm really starting to wonder if these posts are anti-GPLv3 FUD, or the result of misunderstandings created by previous anti-GPLv3 FUD...
This is very simple: If you are not personally planning on redistributing GPLv3 software in such a way that the recipient can't modify it, you don't need to worry about problems from the license. In fact, as an end user you should welcome GPLv3 because it defends you better from patent problems.
At this point, I'm personally willing to suspend Godwins law for discussions about American politics. The more that people look at fascism and the USA next to each other, the more likely it is that we'll be able to fix some of the disturbing similarities.
You comments are cultural rather than rational. The whole idea that you don't want to work for the benefit of others is a standard lower-class American belief - that mindset will get you from "poor" to "upper middle class" in the USA (or at least provide comfort when you fail), so it's competitively advantageous for you to have it.
Other cultures promote different mindsets, many of which are also sustainable in practice. I don't know if there is a mindset that would work in a democratic communist system, but I wouldn't want to simply assume that things are impossible simply because they wouldn't mesh cleanly with my cultural background.
That's the optimistic view.
It seems to me that maintaining a functional system of accountability for any official of the federal government in a country of more than 300 million people may be impossible - especially if you have an economy structured the way the US economy is.
The highest estimate for civilian deaths in Iraq that I've heard is 600,000. If we consider the existence of the war in Afganistan, that implies that we *could* get to an estimate 1,000,000 using the method you describe.
Just to be completely clear: The GPLv3 does not cause any compatibility issue for different programs working together. Your car analogy is nonsensical, because the GNU code and the Linux kernel can happily be under different licenses and still work together fine.
This sounds like unsubstantiated FUD. The GPLv3 doesn't force anyone into a political ideology anymore than GPLv2 did. If you have some specific, practical problem with a specific part of the GPLv3 then we can have a reasoned discussion, but your comments are as vague and non-specific as possible.