"A research firm serving the mobile phone industry has published an 18-page whitepaper about open source licensing. Entitled 'GPLv2 vs. GPLv3', the paper examines the meteoric rise of open source software, and the forces that shaped each license, before concluding with an extremely detailed point-by-point comparison."
Post a Comment
Member since:
2005-07-28
Fans: 0
Member since:2005-08-12
Fans: 2
Member since:2005-07-06
Fans: 11
Member since:2006-07-15
Fans: 5
Member since:2005-07-15
Fans: 2
>The term freedom is never debatable. A golden cage is 'freedom' up to the borders.
But there are natural borders which normally everyone accepts. This border is drawn at the point where the freedom of other persons began.
Nobody would really argue that people should have the power to restrict other people freedom. I say "power" because in my eyes this is the point where "freedom for myself" convert into "my power over other people".
I'm all in favor of freedom for everyone but not in favor of power for everyone.
Edited 2007-09-29 23:09
Member since:2005-07-06
Fans: 11
Member since:2006-10-25
Fans: 1
Anarchy aside, all freedoms are based on limits and restrictions of others freedoms.
Your freedom to own property, for example, depends on restricting other people's freedom to take that property from you.
Your freedom of speech is nothing more than a restriction on others "freedom" to suppress your speech.
And of course, you are free to choose whichever license best suits your project.
Of a more interesting subject of discussion, I remember there was an interesting point made over machine-specific code through encryption (tivoisation) used in the context of casino gaming, voter machines, etc. There may be cases in which the GPL3's freedoms are not ideal, and one would want to restrict in what ways it can be used or redistributed by choosing a more restrictive OSS license.
Edited 2007-09-30 00:02
Member since:2007-02-19
Fans: 0
Member since:2005-07-06
Fans: 11
Member since:
2005-07-15
Fans: 2
I had to read only a few sentences to find the first bullshit (sorry for using this term):
"However where the OSI and the FSF diverge is
that the OSI also require that an OSI approved
open source license should not restrict
commercially important freedoms, such as the
ability to distribute open source and non open
source software together, and to not discriminate
against any persons, field of endeavour or
technology products. This last point is a
particular area of differentiation with the FSF,..."
That's just wrong. That's not where FSF and OSI diverge both are against restriction of any possible use and of course against commercially restriction.
I want to remember the the "Open Source Definition" is nothing more than the "Debian Free Software Guide" so there is absolutely no diverge between OSI and FSF based on their definition of "Open Source" and "Free Software".
...GPLv3 contains wording which appears to
preclude use of GPLv3 covered code in specific
technology areas.
No, GPLv3 doesn't preclude any use. You can use GPLv3 code in any technology areas. But you can't deny your users freedom. That's an important difference.
Edited 2007-09-29 22:54
Member since:
2007-08-01
Fans: 0
"Of a more interesting subject of discussion, I remember there was an interesting point made over machine-specific code through encryption (tivoisation) used in the context of casino gaming, voter machines, etc. There may be cases in which the GPL3's freedoms are not ideal, and one would want to restrict in what ways it can be used or redistributed by choosing a more restrictive OSS license."
The prohibition against tivoisation only applies to "consumer devices." Neither of those cases involve a consumer device as defined in Gpl3, and neither of them would invoke the anti-tivoisation language anyway. The GPL3 only provides that the *owner* of the device in question should have the freedom to modify it. Gamblers don't own slot machines, nor do voters own voting machines. Supposed weakness like this one are produced in abundance by people who don't bother to read and understand the license before commenting.
Member since:2006-10-25
Fans: 1
"Supposed weakness like this one are produced in abundance by people who don't bother to read and understand the license before commenting."
Yes, and I read it as a comment. I had not come to that conclusion myself. I apologize for repeating that misconception and I thank you for the clarification.
Member since:2005-12-02
Fans: 0
"The GPL3 only provides that the *owner* of the device in question should have the freedom to modify it. Gamblers don't own slot machines, nor do voters own voting machines."
The owner of slot would be the casino's. As a gambler I would not want them to be able to control the code either, nor to modify it. Same with voting machines where the government owns them. Casino owners should *never* be allowed to modify code in a machine, nor should the government be allowed to modify code in voting machines. GPLv3 specifically provides this right, which would therefore make it illegal to be used in these contexts, due to other laws which do and have existed. In these contexts the "consumers" are the casino's and governments. Consumer refers to whoever owns the equipment. A gambler or a voter are not "consumers" of those devices, just users. As I mentioned already, due to laws that already exist, GPLv3 licensed code is automatically and already shut out from these markets that you mentioned.
Member since:
2007-02-19
Fans: 0
I thought it was the reverse? Dynamic linking is not a derivative work, but static linking is. isn't it?
Member since:
2007-06-22
Fans: 0
I think people who say that bsd license is a true freedom are hypocrites. If they'd cared for freedom they'd release their code in public domain and wouldn't bitch if someone wouldn't give them credit -> see recent Atheros debakle. This was off topic, I admit, but this issue gets raised every time in a gpl discussion. I think gpl was the desicive reason for open source success and also a reason why competitiors actually can stand to develop the same product (linux) together because they know they all will benefit from it. This is also a reason why companies don't contribute back to bsd-licensed pojects, because their competitor can just snatch it away, add a pair of uniqire features and give it off as their own product and gain an unfair advantage. Like the bsd stack in windows. There are a lot of examples for that. I mean apart from "i dont care for my freedom untill a nazi-government knocks on my door" crowd a lot of people actually care about things like open source drivers an free software ( as in speech ). I am not a gpl fanboy, there are clearly other licenses that accomplish the same thing, they are just like a "police" that get's sometimes annoyng, but without it you wont get the MOTIVATION of alot of people who want to see their "free" code in programs that benefit humanity and not used as a tool to enslave it (for example windows monopoly).
Edited 2007-09-30 08:22
Member since:2005-11-15
Fans: 0
Member since:2005-07-24
Fans: 35
"""
GPL = "Freedom FROM"
"""
That's a reasonable observation.
From another angle:
BSD: Barriers to sharing code are do more harm than good.
GPL: Barriers to code sharing are unfortunate, but do more good than harm.
I think that the availability of FOSS licenses which have different strengths and weaknesses, to suit a variety of projects and authors has been key to our success. Even RMS agrees that Xiph's use of permissive licenses for their reference implementations for the ogg formats is preferable to copylefting them.
One should think very carefully when choosing an appropriate license... always keeping in mind that code reuse is *HARD*. Adding artificial barriers comes at a cost. Sometimes that cost might be outweighed by the advantages.
But in the end, the most important thing is that authors license their code in a way that really reflects their expectations as to what the responsibilities of the users of that code should be. If he expects the users of the code to give back, that should be in the license. Claiming to support permissive licenses and then moaning and whining when people don't pay money back to one's own project (as a certain well known leader of a permissively licensed project has done) does not help anyone.
Edited 2007-09-30 20:23
Member since:2005-07-24
Fans: 0
"BSD: Barriers to sharing code are do more harm than good.
GPL: Barriers to code sharing are unfortunate, but do more good than harm.
You got this backwards. The GPL is not a barrier to code sharing, it's a mandater of it. The GPL prevents a distributer from refusing to share the code. A BSD license allows the distributer to refuse to share the code.
Member since:2005-07-24
Fans: 35
"""
"""
Virtually no FOSS project which is not under GPL can use code from a GPL'd project. Even GPLv2 projects are prohibited from using GPLv3 code. That's a huge barrier to code sharing. A one-way barrier, though, for the most part, since great care has been taken by the GPLvX authors to ensure what they term "compatibility" with other FOSS licenses. That means making sure that GPL projects can take from other projects as they please, without the donor projects receiving anything in return.
I'm willing to accept that situation as (possibly) being good for FOSS as a whole. But please do not just ignore the fact that copyleft licenses do erect barriers to code sharing in the FOSS world.
I would file your argument in the "sometimes the benefits outweigh the disadvantages" category.
Edited 2007-10-02 03:30
Member since:2005-08-18
Fans: 15
Member since:
2007-08-21
Fans: 1
Member since:
2006-12-06
Fans: 0
Member since:
2005-12-02
Fans: 0
Member since:2006-12-07
Fans: 1