Agree 100%. My impression is that the BT patent was generic enough such that internet web servers/clients would have infringed. Prior art would have gotten the patent thrown out altogether, but I'm sure that would have taken much longer.
Sometimes, I think the court doesn't really decide on the basis of law, but what it either consciously or subconsciously thinks society will actually tolerate. That's why you can have "bad decisions" (Dredd Scott, etc.) that eventually get replaced with different decisions.
Whereas in math, you can't have a "bad proof" that sticks around too long.
Yeah, it seems odd that this was presented as a new idea considering they taught it to us in introductory statistics.
They use it on surveys when they ask embarrassing questions, something like you flip a coin and if it comes up one way they can lie and if it comes up the other they have to tell the truth. But the surveyer doesn't really know what they are going to answer so they needn't be embarrassed or worry about the cops coming for them.
People who worry about "government regulation" are usually worried about regulations that directly affect private citizens or the private economy. Nader's argument go to the point of government regulating itself, by examining its own practices (or lack thereof) and deciding if the effect of those practices implies changes in their own requirements... security, cost, etc. with some introspection by the government would result in them choosing open source more often than they do now. Simple enlightened self interest. I did get the thrust of Nader's argument that this should also be done to reign in Microsoft. That was probably a mistake on his part... OMB doesn't have any interest in controlling the private economy, and in fact, and in particular with a republican administration, they would avoid any appearance of trying to do so. But aside from that I think he was right on target. Just encourage enlightened self interest in how the gov mandates procurement for itself...
Yeah, the study is pretty much bullshit. And you know the Microsofts of the world will pick up on this to spread FUD by labelling all open source work as one or two guys in a garage.
Looking at the report, I agree 100%. The study is fatally flawed. The only way to really determine the core developers on a project is to lurk on the mailing list, or ask someone. No one keeps the developer lists on SourceForge up to date (if they're ever accurate at all).
You'd be surprised how a company will jump when a very specific order comes down from the judge. Of course, there's always the alternative of the state contracting someone else to make the fixes.
I don't think there would be that much code which could be deemed worthy of hiding for "obvious" security reasons. That which is obviously an issue could be fixed a piece at a time, whether by Microsoft or a third party.
Remember, none of this code has faced public scrutiny before. To put it all out there at once is a mistake, unless the point is just to give Microsoft a hard time. If the point is to get the code out as quickly as reasonable, but still safely, a little caution is necessary.
But there's no reason to accept Microsoft's assertion that they just can't release code certain code at all because of security concerns. Depending on such security is no security at all. True security is mathematically provable, and secrecy of implementation/equations whatever doesn't affect the soundness of the mathematics.
"lock down" their OS is a totally open-ended target. We both know that they'd never be done, in 4 years, or 100 years.
Better that if they required to open up their source or APIs that a commission or special master take into account issues with the source code. If a code review shows serious flaws, I can see the code being held back until they were fixed. Anything that didn't seem flawed could be released immediately.
It isn't a binary choice here "all or nothing." If forced to release code or API docs, "a piece at a time" is certainly workable.
The creative commons has been designed to work in the area of copyright.
We all hate software patents, but they're not going away. We have to work within the system. Anyone have ideas on how to carve out a commons in the area of e-patents?
Here's a thought: form a Free Patent Foundation which would owns the patents. Regular folks submit patentable mechanisms to this foundation which does the patent search, defends the patent, and collects reasonable royalties on implementations which use it. Pay back some amount of the money procured from proprietary sources back to the submitter.
However, free implementations would pay no royalties at all, they would just be bound to reference the patent in the code. Thoughts?
By your reasoning, we shouldn't have any government roads since that would compete with toll roads.
No military since that would compete with mercenaries.
And your argument about that companies tax dollars is pretty ridiculous too, since they likely only put in a miniscule fraction of the money that went into that product. The government does owe any corporation the protection of its industry, whether from publicly funded software or their competitors, or foreign competitors.
We should have access to the IP our dollars produce. Is that so hard to understand?
I don't see your point. If this scary foreign government wanted that software they could find a proxy corporation to license it.
The point is that consortium's code is quasi-open now, it's available at a relatively low cost to your average business or major government. But that cost isn't so low to the average citizen
Gov't should require (for non-classified material):
GPL it to us for free. Proprietary license it to the corporations for a fee.
That way the consortium can get funding, corporations can hide away inside their code that reasonable people can use for free, and we reasonable types can get our software we paid for. And the software will fluorish quite well in the Free software community, improving the software without the gov't spending a dime more in cases where the code is of general utility.
Well, theoretically I see your point in cases where the government recoups monies/budget allotments by selling software to other agencies.
I'm not a bureaucrat though so the logic doesn't really fly with me. We're taxpayers... our money pays for that code to be developed. So in the best case, as an American, I should get full benefit out of the software. Now, an exception should be made for classified software, but other than that, it's our stuff.
We own it, unless there's a compelling reason to the contrary, we should get it. I don't buy the FUD about hackers/terrorists either. Please. For one, this article is about software specifically designed to be of very high quality. Second, the widespread belief among open source types like myself is that the more eyes that are on the code, the more likely the vulnerabilities will be discovered and ultimately patched. And, hackers find the vulnerabilities in closed source software all the time... the only way to prevent exploitation of vulnerabilities is to remove them as quickly as possible.
And I think a BSDish or X11 license would be a fine choice for the government to use since it doesn't take any political stance. (Although it would be great if they (we) chose the GPL.)
It's simply astonishing though that unclassified information which is produced a taxpayer expense, and by nature doesn't cost much to distribute, isn't given away for free.
Non-classified government funded software should be Free for public use, I guess is the point here. Are there any ongoing lobbying efforts to this effect?
Hmmm... I think it's probably a testing issue. It would be kinda difficult to try all combinations of CD's which aren't REALLY CD's.
Firmware is designed to respond to normal data -- that likely even includes random or corrupted data. But a disk designed to f**k with the firmware's mind, with the express purpose of confusing it, well, all bets are off. I doubt it ever even crossed the engineers mind to handle an evil CD-ish thing, nor should it have. He/she was designing to handle a given specification, and these disks break it.
I think this will eventually happen, mainly because Microsoft will see it as a way to get rid of Free Software (Microsoft can afford insurance and the lobbyists necessary to turn the law in their favor).
It makes sense to think NOW about what we as the free software community would want that law to look.
We'd need it to be strictly defined so that only the distribution vendors could be sued. Makers of say Apache, should never have a non-beta version, milestones, ok, but never "released." Users should be protected from beta versions at all. The distribution vendors, with the wherewithal to test (and responsibility!), should decide whether the package in its current form should be considered releasable.
Debian should probably never release either.
The distribution vendors will become much more careful about blessing any particular component by adding it to their distro (a good thing, since quality will go up). Additionally, they will have to budget for liability insurance. Cost of doing business, welcome to the world outside of the software industry.
The lawmakers and the courts will have to work out the reasonable equivalent of software malpractice. EFF, FSF, Red Hat, etc. will need to lobby to protect the free and open software movement.
Hidenori wasn't particularly worried about DMCA. From reading wine-devel, it was more fear about potential patent infringement and EULAs.
The DMCA creates the same kind of worries X 10 but mainly for Americans (and those who travel to US).
Hmmm... I see Codeweavers business plan as pretty straightforward... sell products with full support for a profit. And from what I've heard they are doing quite well with the CrossOver products.
Now Transgaming is the oddball here... no one likes to pay a subscription fee for software. But I hope it works for them. Need more good open source business models.
As for Lindows, their model is pretty much like Codeweavers.
But it doesn't really matter. All three are valuable, contributing members of the wine community. It would be much better for us plain old folks if there were only one tree. All three businesses are keeping stuff proprietary but all three consistently give back in varying degrees to the main tree (though TransGaming seems to like the "code trading" idea that Alexandre really doesn't like). Thus I have a hard time getting as upset at any of the parties as you seem to be.
I'd draw the line where where the hardware is abstracted and rudimentary system objects are defined. So, you can use all of your devices and rudimentary objects in a generic way (sound cards, video cards, printers, windows, fonts, etc.) but you wouldn't have software that would interpret an MP3, or an OGG. Those can be part of add-on libraries, or part of the installer for a given application.
Well, if you want to reason by platitude, I respond by saying that just because an idea is old and established doesn't make it wrong or irrelevant.
There's a real difference between the operating system and the apps. The typical (modern!) engineer is going to say that boundary is somewhere between the kernel/system services and the apps you run. I don't think anyone would consider a media player or web browser a system service or part of the kernel.
It's totally separable, they should be kept separate for a variety of reasons, and Microsoft knows it can be done. Yeah umpteen apps can and should depend on helper dynamically linked helper libraries that aren't really part of the OS. It's perfectly reasonable to embed COM services exposed by another app. That still doesn't make the COM server a part of the OS, just a dependency of your app.
Slashdot Mirror
Version numbers are arbitrary.
I guess Debian should have skipped straight from 2 to 7...
Agree 100%. My impression is that the BT patent was generic enough such that internet web servers/clients would have infringed. Prior art would have gotten the patent thrown out altogether, but I'm sure that would have taken much longer.
Sometimes, I think the court doesn't really decide on the basis of law, but what it either consciously or subconsciously thinks society will actually tolerate. That's why you can have "bad decisions" (Dredd Scott, etc.) that eventually get replaced with different decisions.
Whereas in math, you can't have a "bad proof" that sticks around too long.
Isn't human language fun?
The Free BIOS projects I know of are
OpenBIOS and
LinuxBIOS
-- John
Yeah, it seems odd that this was presented as a new idea considering they taught it to us in introductory statistics.
They use it on surveys when they ask embarrassing questions, something like you flip a coin and if it comes up one way they can lie and if it comes up the other they have to tell the truth. But the surveyer doesn't really know what they are going to answer so they needn't be embarrassed or worry about the cops coming for them.
-- John.
People who worry about "government regulation" are usually worried about regulations that directly affect private citizens or the private economy. Nader's argument go to the point of government regulating itself, by examining its own practices (or lack thereof) and deciding if the effect of those practices implies changes in their own requirements... security, cost, etc. with some introspection by the government would result in them choosing open source more often than they do now. Simple enlightened self interest.
I did get the thrust of Nader's argument that this should also be done to reign in Microsoft. That was probably a mistake on his part... OMB doesn't have any interest in controlling the private economy, and in fact, and in particular with a republican administration, they would avoid any appearance of trying to do so. But aside from that I think he was right on target. Just encourage enlightened self interest in how the gov mandates procurement for itself...
Yeah, the study is pretty much bullshit. And you know the Microsofts of the world will pick up on this to spread FUD by labelling all open source work as one or two guys in a garage.
Looking at the report, I agree 100%. The study is fatally flawed. The only way to really determine the core developers on a project is to lurk on the mailing list, or ask someone. No one keeps the developer lists on SourceForge up to date (if they're ever accurate at all).
Ah, the bravery of being out of range.
Nevertheless, I can't imagine real pilots having the same positive impression of these devices. Human judgement counts for something.
Will these things end up making less of these "mistaken target bombings" or is that all just garbage-in-garbage-out intelligence snafus?
You'd be surprised how a company will jump when a very specific order comes down from the judge. Of course, there's always the alternative of the state contracting someone else to make the fixes.
I don't think there would be that much code which could be deemed worthy of hiding for "obvious" security reasons. That which is obviously an issue could be fixed a piece at a time, whether by Microsoft or a third party.
Remember, none of this code has faced public scrutiny before. To put it all out there at once is a mistake, unless the point is just to give Microsoft a hard time. If the point is to get the code out as quickly as reasonable, but still safely, a little caution is necessary.
But there's no reason to accept Microsoft's assertion that they just can't release code certain code at all because of security concerns. Depending on such security is no security at all. True security is mathematically provable, and secrecy of implementation/equations whatever doesn't affect the soundness of the mathematics.
Actually, I don't agree.
"lock down" their OS is a totally open-ended target. We both know that they'd never be done, in 4 years, or 100 years.
Better that if they required to open up their source or APIs that a commission or special master take into account issues with the source code. If a code review shows serious flaws, I can see the code being held back until they were fixed. Anything that didn't seem flawed could be released immediately.
It isn't a binary choice here "all or nothing." If forced to release code or API docs, "a piece at a time" is certainly workable.
Agreed... the most security on their Shared Source stuff is at the level of a non-disclosure agreement.
I think the judge will see through this ploy.
The creative commons has been designed to work in the area of copyright.
We all hate software patents, but they're not going away. We have to work within the system. Anyone have ideas on how to carve out a commons in the area of e-patents?
Here's a thought: form a Free Patent Foundation which would owns the patents. Regular folks submit patentable mechanisms to this foundation which does the patent search, defends the patent, and collects reasonable royalties on implementations which use it. Pay back some amount of the money procured from proprietary sources back to the submitter.
However, free implementations would pay no royalties at all, they would just be bound to reference the patent in the code.
Thoughts?
Lose the tude, pal, your logic ain't that solid.
By your reasoning, we shouldn't have any government roads since that would compete with toll roads.
No military since that would compete with mercenaries.
And your argument about that companies tax dollars is pretty ridiculous too, since they likely only put in a miniscule fraction of the money that went into that product. The government does owe any corporation the protection of its industry, whether from publicly funded software or their competitors, or foreign competitors.
We should have access to the IP our dollars produce. Is that so hard to understand?
I don't see your point. If this scary foreign government wanted that software they could find a proxy corporation to license it.
The point is that consortium's code is quasi-open now, it's available at a relatively low cost to your average business or major government. But that cost isn't so low to the average citizen
We could also use the dual license idea.
Gov't should require (for non-classified material):
GPL it to us for free.
Proprietary license it to the corporations for a fee.
That way the consortium can get funding, corporations can hide away inside their code that reasonable people can use for free, and we reasonable types can get our software we paid for.
And the software will fluorish quite well in the Free software community, improving the software without the gov't spending a dime more in cases where the code is of general utility.
Well, theoretically I see your point in cases where the government recoups monies/budget allotments by selling software to other agencies.
I'm not a bureaucrat though so the logic doesn't really fly with me. We're taxpayers... our money pays for that code to be developed. So in the best case, as an American, I should get full benefit out of the software. Now, an exception should be made for classified software, but other than that, it's our stuff.
We own it, unless there's a compelling reason to the contrary, we should get it.
I don't buy the FUD about hackers/terrorists either. Please. For one, this article is about software specifically designed to be of very high quality. Second, the widespread belief among open source types like myself is that the more eyes that are on the code, the more likely the vulnerabilities will be discovered and ultimately patched. And, hackers find the vulnerabilities in closed source software all the time... the only way to prevent exploitation of vulnerabilities is to remove them as quickly as possible.
I read that too.
And I think a BSDish or X11 license would be a fine choice for the government to use since it doesn't take any political stance. (Although it would be great if they (we) chose the GPL.)
It's simply astonishing though that unclassified information which is produced a taxpayer expense, and by nature doesn't cost much to distribute, isn't given away for free.
Non-classified government funded software should be Free for public use, I guess is the point here. Are there any ongoing lobbying efforts to this effect?
Hmmm... I think it's probably a testing issue. It would be kinda difficult to try all combinations of CD's which aren't REALLY CD's.
Firmware is designed to respond to normal data -- that likely even includes random or corrupted data. But a disk designed to f**k with the firmware's mind, with the express purpose of confusing it, well, all bets are off. I doubt it ever even crossed the engineers mind to handle an evil CD-ish thing, nor should it have. He/she was designing to handle a given specification, and these disks break it.
I think this will eventually happen, mainly because Microsoft will see it as a way to get rid of Free Software (Microsoft can afford insurance and the lobbyists necessary to turn the law in their favor).
It makes sense to think NOW about what we as the free software community would want that law to look.
We'd need it to be strictly defined so that only the distribution vendors could be sued. Makers of say Apache, should never have a non-beta version, milestones, ok, but never "released." Users should be protected from beta versions at all. The distribution vendors, with the wherewithal to test (and responsibility!), should decide whether the package in its current form should be considered releasable.
Debian should probably never release either.
The distribution vendors will become much more careful about blessing any particular component by adding it to their distro (a good thing, since quality will go up). Additionally, they will have to budget for liability insurance. Cost of doing business, welcome to the world outside of the software industry.
The lawmakers and the courts will have to work out the reasonable equivalent of software malpractice. EFF, FSF, Red Hat, etc. will need to lobby to protect the free and open software movement.
Other ideas?
Hidenori wasn't particularly worried about DMCA. From reading wine-devel, it was more fear about potential patent infringement and EULAs. The DMCA creates the same kind of worries X 10 but mainly for Americans (and those who travel to US).
Hmmm... I see Codeweavers business plan as pretty straightforward... sell products with full support for a profit. And from what I've heard they are doing quite well with the CrossOver products.
Now Transgaming is the oddball here... no one likes to pay a subscription fee for software. But I hope it works for them. Need more good open source business models.
As for Lindows, their model is pretty much like Codeweavers.
But it doesn't really matter. All three are valuable, contributing members of the wine community. It would be much better for us plain old folks if there were only one tree. All three businesses are keeping stuff proprietary but all three consistently give back in varying degrees to the main tree (though TransGaming seems to like the "code trading" idea that Alexandre really doesn't like). Thus I have a hard time getting as upset at any of the parties as you seem to be.
He can't stop everyone, really. He is able to pull it only at the sufferance of the rest of the wine maintainer (Alexandre Julliard).
I personally, agree with Alexandre in acquiescing to Hidenori's wishes... which were clear "please do not use my codes anymore."
I'd draw the line where where the hardware is abstracted and rudimentary system objects are defined. So, you can use all of your devices and rudimentary objects in a generic way (sound cards, video cards, printers, windows, fonts, etc.) but you wouldn't have software that would interpret an MP3, or an OGG. Those can be part of add-on libraries, or part of the installer for a given application.
Well, if you want to reason by platitude, I respond by saying that just because an idea is old and established doesn't make it wrong or irrelevant.
There's a real difference between the operating system and the apps. The typical (modern!) engineer is going to say that boundary is somewhere between the kernel/system services and the apps you run. I don't think anyone would consider a media player or web browser a system service or part of the kernel.
It's totally separable, they should be kept separate for a variety of reasons, and Microsoft knows it can be done. Yeah umpteen apps can and should depend on helper dynamically linked helper libraries that aren't really part of the OS. It's perfectly reasonable to embed COM services exposed by another app. That still doesn't make the COM server a part of the OS, just a dependency of your app.