I made a comment on IRC about this, so I may as well vent here as well:
http://www.gnu.org/licenses/fdl-1.3.html
I'm not a big fan of these new extra permissions clauses, but limiting them by date so that only our special friends can make use of them seems peverse. Both GPLv3 and now GFDLv1.3 have them now, with dates that are back-dated before the publication of the licenses.
Presumably they're back-dated because we don't want other GFDL material to suddenly become CC-BY-SA just by posting it on a wiki, but that - to me - kind of points out the antithetical nature of this modification to the spirit of the original license.
Thoughts?
Cheers,
Alex.
Alex Hudson home@alexhudson.com writes:
I'm not a big fan [...] limiting them by date
There's flexibility and risk in there in that those works can be relicensed to cc-by-sa-3.0 *or* "future copyleft versions of that license".
This gives Wikipedia and Creative Commons some flexibility that's probably very valuable for them. They can tweak the licence if they think it's necessary or if the Wikipedia community demands it.
But it creates some risks to the freedom of GFDL'd wiki'd works, and it's a one-way relicensing that isn't being reciprocated by Creative Commons. With an endless timeline, this would be foolish and even dangerous, but by limiting it to a 12 month period, the risk becomes pretty small.
...I think.
Ciaran O'Riordan wrote:
Alex Hudson home@alexhudson.com writes:
I'm not a big fan [...] limiting them by date
There's flexibility and risk in there in that those works can be relicensed to cc-by-sa-3.0 *or* "future copyleft versions of that license".
Well, presumably not much more risk than future versions of the GFDL at this rate though.
But it creates some risks to the freedom of GFDL'd wiki'd works, and it's a one-way relicensing that isn't being reciprocated by Creative Commons. With an endless timeline, this would be foolish and even dangerous, but by limiting it to a 12 month period, the risk becomes pretty small.
The thing is, that's not really what it is. This clause in the license is pretty much specifically there so that Wikipedia can vote on whether or not to go CC-BY-SA or not (as a corollary, if they vote no, this clause becomes basically a solution in search of a problem).
If the FSF were asked whether or not Debian (for example) could have the Emacs manual under CC-BY-SA rather than the GFDL, I suspect they would tell them to get lost, or words to that effect. But they don't seem to have trouble making that decision for other people who happened to publish their work onto a wiki.
Putting it bluntly: playing games like this will make people distrust the "or later" clauses. If people choose GFDL, it's pretty bloody obvious (to me) that the relaxed CC-BY-SA license isn't what they wanted.
My personal gripe is that I'm stuck with one particular project which is on GPLv2 because the developers didn't trust "or later". If significant amounts of free software stop trusting "or later", we're going to end up with horrendous problems (viz. the GPL compatibility matrix).
FSF have effectively been trusted with the power to relicense large numbers of works. Using that power to solve other people's problems seems a very slippery slope to me.
Cheers,
Alex.
Alex Hudson home@alexhudson.com writes:
playing games like this will make people distrust the "or later" clauses.
Let's assess the risks first:
If people choose GFDL, it's pretty bloody obvious (to me) that the relaxed CC-BY-SA license isn't what they wanted.
This is a non-issue for Wikipedia because each contributor owns their contributions - not the Wikimedia Foundation. There can be no hostile change of licence there.
Is there a project that will be eligible for this clause and for which the copyright isn't held by the contributors?
replying to correct myself:
Ciaran O'Riordan ciaran@fsfe.org writes:
This is a non-issue for Wikipedia because each contributor owns their contributions
No, I seem to be wrong. The licence says:
"The operator of an MMC Site may republish an MMC contained in the site under CC-BY-SA on the same site at any time before August 1, 2009, provided the MMC is eligible for relicensing.
so the licence gives the site operating the choice to change the licence of content which is owned by the users.
Knowing FSF, I assume they got sufficient reassurances from Wikimedia Foundation that they won't abuse this power. So to assess the risks, the question to ask is what other sites (GFDL'd wikis) are eligible for this change of licence?
Forgive me Ciaran, but I think you're pretty much begging the question:
Ciaran O'Riordan wrote:
Knowing FSF, I assume they got sufficient reassurances from Wikimedia Foundation that they won't abuse this power. So to assess the risks, the question to ask is what other sites (GFDL'd wikis) are eligible for this change of licence?
The reassurances from Wikimedia, whatever they were, is by-the-by really. The point is this; if you contribute to a project under the GFDL, are we really saying that CC-BY-SA is the spiritual ancestor of that license?
I personally find that a difficult proposition.
This appears to be a pragmatic approach to solving a license incompatibility. The truly pragmatic approach, though, would be to get rid of one or other of the licenses, so that one was truly the ancestor of the other, and reduce fragmentation. That's not happening, though: the incompatibility will remain, but a small window will open up when people can choose a license other than that set by the author. In that respect, it's a temporary band-aid over the problem.
If it was a permanent solution, I might have fewer problems with it. But what message does this send out to authors? "To solve proliferation problems, we'll relicense your works when we see fit?". This GFDL->CC-BY-SA process could be all sweetness and light, and cause nothing but good. That's not really the point I raised.
Putting it concisely: on what basis is CC-BY-SA a spiritual successor to the GFDL, other than the technical fact that it has been grand-childed in?
Cheers,
Alex.
On Mon, Nov 03, 2008 at 10:58:28PM +0000, Alex Hudson wrote:
Putting it concisely: on what basis is CC-BY-SA a spiritual successor to the GFDL, other than the technical fact that it has been grand-childed in?
Perhaps the real question is, is what way is it NOT significantly a successor?
Noah Slater wrote:
On Mon, Nov 03, 2008 at 10:58:28PM +0000, Alex Hudson wrote:
Putting it concisely: on what basis is CC-BY-SA a spiritual successor to the GFDL, other than the technical fact that it has been grand-childed in?
Perhaps the real question is, is what way is it NOT significantly a successor?
Perhaps I was making an assumption that it was relatively obvious to readers when it isn't, so I'll give you the main example I was thinking of.
This GFDL->CC-BY-SA conversion clause is time limited to less than one year, so by that point the people who want works under the CC license have to get out of Dodge. That doesn't mean that the original work stops being available, though, and it might not be the original author who actually converted it: after all, the rule is just that it was a GFDL'd doc without invariant sections published to a wiki (to simplify).
After that point in 2009, you end up in the crazy [to me] situation that an author who has chosen a copylefted license (the FDL) might have had their license changed to one which is incompatible with the one they chose, without notice to them, and they can't use the modifications that other people make to their documentation under that new license.
The whole point of copyleft - to me, anyway - is that other people have to publish the changes they make to your work in a manner that you can re-use. The fact that a work could be effectively split into two separate communities who cannot re-use each other's work seems to me pretty antithetical to the spirit of the original license.
Cheers,
Alex.
Ciaran O'Riordan wrote:
If people choose GFDL, it's pretty bloody obvious (to me) that the relaxed CC-BY-SA license isn't what they wanted.
This is a non-issue for Wikipedia because each contributor owns their contributions - not the Wikimedia Foundation. There can be no hostile change of licence there.
What does copyright ownership have to do with the point I raised?
Are you seriously saying that someone can't now arbitrarily relicense Wikipedia content as CC-BY-SA?
Cheers,
Alex.
2008/11/3 Alex Hudson home@alexhudson.com:
The thing is, that's not really what it is. This clause in the license is pretty much specifically there so that Wikipedia can vote on whether or not to go CC-BY-SA or not (as a corollary, if they vote no, this clause becomes basically a solution in search of a problem).
No-one understands the GFDL.
Not even the FSF understands the GFDL.
If you doubt this sweeping statement, try emailing licensing@fsf.org with any query about the GFDL. Any query at all.
You will get back, about three months later, "we suggest you read the text and ask your attorney."
YOU WROTE THE THING, YOU USELESS ...
- d.
On Mon, Nov 03, 2008 at 11:04:30PM +0000, David Gerard wrote:
No-one understands the GFDL.
Not even the FSF understands the GFDL.
All this talk, and no action. If the FSF is so culpable, why doesn't the FSFE step up and do something about the licensing situation?
Noah Slater nslater@bytesexual.org writes:
All this talk, and no action. If the FSF is so culpable, why doesn't the FSFE step up and do something about the licensing situation?
As a level-headed organisation, the first somethings that FSFE will do about this situation are discuss it and analyse it.
I see the potential problems, but FSF are pretty cautious, so I guess they've seen the potential problems too and have reasons why they're not so worrying or why they were worth the risk.
BTW, there's a GFDL 1.3 FAQ: http://www.gnu.org/licenses/fdl-1.3-faq.html (But it's mostly clarifications, not answers to the questions posed here.)
Or is there something specific that you suggest FSFE does about licencing?
For GPLv3, we worked on making the process accessible: http://fsfeurope.org/projects/gplv3/
For licence usage and enforcement, we've built a network and docs: http://fsfeurope.org/projects/ftf/ http://fsfe.org/en/fellows/ciaran/ciaran_s_free_software_notes/status_of_fsf...
"David Gerard" dgerard@gmail.com writes:
If you doubt this sweeping statement, try emailing licensing@fsf.org with any query about the GFDL. Any query at all.
That's partly a text of resources, not a test of knowledge, and further:
YOU WROTE THE THING, YOU USELESS ...
Yes, but they didn't write copyright law. Not in their country, not in your country, and not in my country.
Can we discuss the changes in GFDL 1.3?
More discussion can be found at: http://lists.wikimedia.org/pipermail/foundation-l/2008-November/thread.html#... http://news.slashdot.org/article.pl?sid=08/11/04/0150230 http://lwn.net/Articles/305736/ http://technocrat.net/d/2008/11/3/53277
Alex Hudson home@alexhudson.com wrote:
<http://www.gnu.org/licenses/fdl-1.3.html>
[...]
Thoughts?
I think you're bang on. Sadly, FSF is giving further reasons not to trust it on copyleft and manuals. Free software needs free software documentation, but the FSF seems content to redefine "free" for documentation, even to the point of introducing loopholes to help the (generally non-free-software) Creative Commons licences.
This is quite an attitude change to past statments by RMS like "I cannot endorse Creative Commons as a whole, because some of its licenses are unacceptable. It would be self-delusion to try to endorse just some of the Creative Commons licenses, because people lump them together; they will misconstrue any endorsement of some as a blanket endorsement of all. I therefore find myself constrained to reject Creative Commons entirely." http://web.archive.org/web/20061105211448/http://www.linuxp2p.com/forums/vie...
This U-turn also seems to be reflected in the GNU license list now.
If the FSF were asked whether or not Debian (for example) could have the Emacs manual under CC-BY-SA rather than the GFDL, I suspect they would tell them to get lost, or words to that effect. [...]
Well, the words that RMS actually used when a debian developer asked for manuals under the GPL were "the issue is not significant". http://lists.debian.org/debian-legal/2003/09/msg00391.html
FSF have effectively been trusted with the power to relicense large numbers of works. Using that power to solve other people's problems seems a very slippery slope to me.
It's amazing how much influence one can get by dubious relicensing to the FDL for a few years, isn't it? Meanwhile, ordinary free software developers can't even get bugs in the license drafting website to be fixed and stay fixed...
FSF seems increasingly broken. Is it time for a developer-led organisation to fork its licences, so we can use "or later" again?
Regards,
On Tue, Nov 04, 2008 at 01:04:53AM +0000, MJ Ray wrote:
Free software needs free software documentation, but the FSF seems content to redefine "free" for documentation
MJ Ray, are you arguing that "freedom" means the same in every context? Are you arguing that my freedoms under the United Kingdom's social contract should be identical to my software freedoms? I feel that you are being bewitched by the word "freedom" in this manner. Just because the word remains the same across contexts does not mean that its implications should remain the same.
Thus, I think it's fine that the FSF defines documentation freedom differently than it does software freedom, and in fact freedom of personal expression.
cf. The GNU Verbatim License
even to the point of introducing loopholes to help the (generally non-free-software) Creative Commons licences.
This is, I feel, deliberately misleading.
The Creative Commons is an organisation that provides canned licenses for general use, some of which are free and some of which are not. That the Creative Commons provides non-free licenses has absolutely nothing to do with the topic under discussion.
Your mud slinging is counter-productive.
This is quite an attitude change to past statments by RMS like "I cannot endorse Creative Commons as a whole, because some of its licenses are unacceptable. It would be self-delusion to try to endorse just some of the Creative Commons licenses, because people lump them together; they will misconstrue any endorsement of some as a blanket endorsement of all. I therefore find myself constrained to reject Creative Commons entirely."
I feel you are purposefully misrepresenting the FSF on this matter.
Just because the FSF has stated that it cannot endorse the Creative Commons as a whole, which is an entirely understandable position, it does not mean that they cannot see the value of the free culture licenses that they provide.
I'm not sure if this is a strawman or a deliberate attempt at misdirection.
Either way, I fail to see how it is at all relevant to the topic.
If the FSF were asked whether or not Debian (for example) could have the Emacs manual under CC-BY-SA rather than the GFDL, I suspect they would tell them to get lost, or words to that effect. [...]
Well, the words that RMS actually used when a debian developer asked for manuals under the GPL were "the issue is not significant". http://lists.debian.org/debian-legal/2003/09/msg00391.html
Again, I feel you are deliberately misrepresenting Stallman.
He actually said:
I don't think anyone ever did so. In practice, the issue is not significant, since you can distribute the manual along with the software, and make the software access the manual in whichever way you want.
Which is a perfectly reasonable comment.
It's amazing how much influence one can get by dubious relicensing to the FDL for a few years, isn't it? Meanwhile, ordinary free software developers can't even get bugs in the license drafting website to be fixed and stay fixed...
Irrelevant.
FSF seems increasingly broken. Is it time for a developer-led organisation to fork its licences, so we can use "or later" again?
Well then, why don't you spend less time slinging mud and using current affairs to badmouth the FSF, and more time being productive?
Best,
MJ Ray mjr@phonecoop.coop writes:
This is quite an attitude change to past statments by RMS like "I cannot endorse Creative Commons as a whole, because some of its licenses are unacceptable. It would be self-delusion to try to endorse just some of the Creative Commons licenses, because people lump them together; they will misconstrue any endorsement of some as a blanket endorsement of all. I therefore find myself constrained to reject Creative Commons entirely."
That quote has nothing to do with anything we're discussing here. There are real issues to discuss here, so let's try to stay on topic.
(The situation that RMS was commenting on doesn't even exist anymore. CC separated out the licences that RMS objected to, AFAICT.)
Ciaran O'Riordan ciaran@fsfe.org wrote:
MJ Ray mjr@phonecoop.coop writes:
This is quite an attitude change to past statments by RMS like "[...] I therefore find myself constrained to reject Creative Commons entirely."
That quote has nothing to do with anything we're discussing here. There are real issues to discuss here, so let's try to stay on topic.
Sorry. Did I misunderstand that FDL 1.3 approves conversion to a CC licence? Isn't it approving CC's chequered history and future a little?
(The situation that RMS was commenting on doesn't even exist anymore. CC separated out the licences that RMS objected to, AFAICT.)
I missed that specific change and I didn't find what happened to the SamplingPlus and Developing Countries licences. However, CC remains a broad label which also covers things opposed to free software and free manuals.
Noah Slater nslater@bytesexual.org wrote:
On Tue, Nov 04, 2008 at 01:04:53AM +0000, MJ Ray wrote:
Free software needs free software documentation, but the FSF seems content to redefine "free" for documentation
MJ Ray, are you arguing that "freedom" means the same in every context? Are you arguing that my freedoms under the United Kingdom's social contract should be identical to my software freedoms?
Firstly, no, clearly not. Our freedoms under the various rights Acts are far more wide-ranging. However, I don't see why free software documentation should be less free than free software.
Secondly, please send personal messages off-list.
Thus, I think it's fine that the FSF defines documentation freedom differently than it does software freedom, and in fact freedom of personal expression.
Actually, last I saw, the FSF did not define documentation freedom at all and left us trying to rebuild that pig from the FSF-licence sausages which have been produced. For example, how does having one's past copylefted wiki contributions relicensed to CC form part of a "documentation freedom" concept?
[...]
I feel you are purposefully misrepresenting the FSF on this matter. [...] Again, I feel you are deliberately misrepresenting Stallman.
I'm commenting as I feel. As far as possible, I give links so readers can decide for themselves whether they agree with the feelings that the quoted material stirred in me. If I were deliberately misrepresenting, I'd leave the links out, so reader can't check. (Noah Slater leaves the links out.)
[...]
FSF seems increasingly broken. Is it time for a developer-led organisation to fork its licences, so we can use "or later" again?
Well then, why don't you spend less time slinging mud and using current affairs to badmouth the FSF, and more time being productive?
I'd like to see whether there's wider interest in developer-led licence stewards, so I can decide whether opening my current work to more developers would be worthwhile, or a distracting waste of time.
It's not possible to do productive work directly with the FSF on its licences because the stet-centred process is deliberately inaccessible and incompatible with free software communities. stet may be useful for lawyers and big corporations, but it doesn't collaborate well. Also, *question* FSF's actions (let alone critcise them) and dozens go off the deep at you.
Regards,
On Tue, Nov 04, 2008 at 02:36:00PM +0000, MJ Ray wrote:
That quote has nothing to do with anything we're discussing here. There are real issues to discuss here, so let's try to stay on topic.
Sorry. Did I misunderstand that FDL 1.3 approves conversion to a CC licence? Isn't it approving CC's chequered history and future a little?
No. You are conflating a license with an organisation.
The FSF approves of free software licenses linked to the OSI and other organisations it might otherwise argue with in the general case. I fail to see how this is a contradiction, and I feel that you are trying to confuse people.
(The situation that RMS was commenting on doesn't even exist anymore. CC separated out the licences that RMS objected to, AFAICT.)
I missed that specific change and I didn't find what happened to the SamplingPlus and Developing Countries licences. However, CC remains a broad label which also covers things opposed to free software and free manuals.
Yes, maybe this would be an issue if the FDL was facilitating a switch to ANY of the Creative Commons licenses, which it clearly is not. Why must you insist that these to things are related in anyway?
Noah Slater nslater@bytesexual.org wrote:
On Tue, Nov 04, 2008 at 01:04:53AM +0000, MJ Ray wrote:
Free software needs free software documentation, but the FSF seems content to redefine "free" for documentation
MJ Ray, are you arguing that "freedom" means the same in every context? Are you arguing that my freedoms under the United Kingdom's social contract should be identical to my software freedoms?
Firstly, no, clearly not. Our freedoms under the various rights Acts are far more wide-ranging. However, I don't see why free software documentation should be less free than free software.
It's not "less" free, this is an absurd thing to state.
The word "free" simply means different things in different contexts.
If you can see that "free" society is different from "free" software, which may also be different from "free" speech, why can't you also see that "free" documentation may mean something different too?
Sure, to you and to Debian "free" software and "free" documentation mean the same thing, but to me and the FSF it doesn't. I do not argue with your right to hold that opinion, but I argue with the way that your frame it in your argument.
but the FSF seems content to redefine "free" for documentation
You seem to be implying that your definition of "free" documentation is somehow the original, or the true, version and the FSF has snuck in and redefined it for you without asking, which is an absurd implication. The FSF defined "free" documentation long before Debian or (I presume) you did. It is you who have redefined the FSF's position and are using your redefinition to accuse them of some illusory position change!
Either you're confused or you are trying to confuse.
Secondly, please send personal messages off-list.
It wasn't a personal message, sorry if I offended you.
Thus, I think it's fine that the FSF defines documentation freedom differently than it does software freedom, and in fact freedom of personal expression.
Actually, last I saw, the FSF did not define documentation freedom at all and left us trying to rebuild that pig from the FSF-licence sausages which have been produced.
Are you kidding?
How is producing a license not the ultimate definition of what they consider free?
[...]
I feel you are purposefully misrepresenting the FSF on this matter. [...] Again, I feel you are deliberately misrepresenting Stallman.
I'm commenting as I feel. As far as possible, I give links so readers can decide for themselves whether they agree with the feelings that the quoted material stirred in me. If I were deliberately misrepresenting, I'd leave the links out, so reader can't check. (Noah Slater leaves the links out.)
I leave the links out? What?
It's not possible to do productive work directly with the FSF on its licences because the stet-centred process is DELIBERATELY inaccessible and incompatible with free software communities.
(emphasis mine)
This is *quite* a statement, do you care to back it up?
stet may be useful for lawyers and big corporations, but it doesn't collaborate well.
But writing a license is *fundamentally* different from writing code.
We're talking about producing a legal document. This takes people with decades of legal experience and training. The "patches welcome", "lets do this on a wiki" approach isn't going to work.
Also, *question* FSF's actions (let alone critcise them) and dozens go off the deep at you.
Not at all, I am a strong critic of the FSF when I feel that they have done something I dislike, but I get the impression from some of *your* messages that you like to take any excuse to shift the topic onto FSF bashing.
Best,
I think there are some interesting points in this semi-flame ;)
Noah Slater wrote:
On Tue, Nov 04, 2008 at 02:36:00PM +0000, MJ Ray wrote:
Actually, last I saw, the FSF did not define documentation freedom at all and left us trying to rebuild that pig from the FSF-licence sausages which have been produced.
Are you kidding?
How is producing a license not the ultimate definition of what they consider free?
Well, the GPL isn't a terribly good guide to what the FSF consider to be the definition of "free software", and I don't mean that in a negative way at all: it's like trying to divine the rules for what is considered exceptional art by looking at Constable's "The Hay Wain". It's an example, not an exhaustive guide.
I think MJ is essentially right that there isn't a good treatment of what "free documents" actually are. I dislike the Debian approach which attempts bitstream-blindness, although it makes sense in the context of a digital distribution. I like parts of the GFDL, including the clarity on non-digital distribution, the anti-DRM stuff (modulo comments about whether it properly achieves the aim), etc. But the "Grand Unified Theory" that Eben Moglen has spoken of previously seems to be still out of reach.
We're talking about producing a legal document. This takes people with decades of legal experience and training. The "patches welcome", "lets do this on a wiki" approach isn't going to work.
Actually, I disagree with this on two grounds, one specific and one general.
The general comment is that you don't need to be able to write legal documents to be able to comment on them. With my small amount of legal training I feel thoroughly qualified to comment on the Wikipedia clause, and it doesn't appear that I have misunderstood anything. Maybe I can't write a correction that would be worthwhile, but that doesn't invalidate my position.
The specific comment is that the Wikipedia clause could not have been offered to general examination beforehand because of the fear that various manuals would suddenly become CC'able when the FSF didn't want that to happen. I'm slightly astonished that reasoning holds, because the assumption is that Wikipedia contributors will be happy with a change that other authors would be unhappy with.
Cheers,
Alex.
2008/11/4 Alex Hudson home@alexhudson.com:
I think MJ is essentially right that there isn't a good treatment of what "free documents" actually are. I dislike the Debian approach which attempts bitstream-blindness, although it makes sense in the context of a digital distribution. I like parts of the GFDL, including the clarity on non-digital distribution, the anti-DRM stuff (modulo comments about whether it properly achieves the aim), etc. But the "Grand Unified Theory" that Eben Moglen has spoken of previously seems to be still out of reach.
FWIW, Wikimedia uses the http://freedomdefined.org definition (which happens to have been written by Erik Moeller from Wikimedia).
http://commons.wikimedia.org/wiki/Commons:File_types speaks of patent-encumbered formats not being free enough for WMF purposes.
- d.
Noah Slater nslater@bytesexual.org wrote:
On Tue, Nov 04, 2008 at 02:36:00PM +0000, MJ Ray wrote: [...]
I missed that specific change and I didn't find what happened to the SamplingPlus and Developing Countries licences. However, CC remains a broad label which also covers things opposed to free software and free manuals.
Yes, maybe this would be an issue if the FDL was facilitating a switch to ANY of the Creative Commons licenses, which it clearly is not. Why must you insist that these to things are related in anyway?
As previously quoted, RMS also related all CC licences. Why insist it's unreasonable to associate them?
[...]
If you can see that "free" society is different from "free" software, which may also be different from "free" speech, why can't you also see that "free" documentation may mean something different too?
Because when the documentation is stored as software, its "free"ness can also be evaluated as software as well as documentation. They're not chalk and cheese like society and software.
Sure, to you and to Debian "free" software and "free" documentation mean the same thing, [...]
No, it isn't and nothing depends on that mistake.
but the FSF seems content to redefine "free" for documentation
You seem to be implying that your definition of "free" documentation is somehow the original, or the true, version and the FSF has snuck in and redefined it for you without asking, which is an absurd implication. The FSF defined "free" documentation long before Debian or (I presume) you did.
Cool. Where? TTBOMK, none of those groups have explicitly defined free documentation.
Side-stepping for a moment:-
"David Gerard" dgerard@gmail.com wrote:
FWIW, Wikimedia uses the http://freedomdefined.org definition (which happens to have been written by Erik Moeller from Wikimedia).
http://commons.wikimedia.org/wiki/Commons:File_types speaks of patent-encumbered formats not being free enough for WMF purposes.
A problem with freedomdefined.org is that it also uses the "essential freedoms" devious words, but it seems better than the FSF concept of free documentation as I understand it.
Patent-encumbered formats are a practical threat to freedom.
Back to Noah Slater:
It wasn't a personal message, sorry if I offended you.
Well, this message also seems to be very personal (you, you, you) and I feel that it's in an utterly inappropriate register for discussion.
Actually, last I saw, the FSF did not define documentation freedom at all and left us trying to rebuild that pig from the FSF-licence sausages which have been produced.
Are you kidding?
No.
How is producing a license not the ultimate definition of what they consider free?
A license is a particular instance. The definition would be a general case. (cf Differential Equations terminology)
[...] It's not possible to do productive work directly with the FSF on its licences because the stet-centred process is DELIBERATELY inaccessible and incompatible with free software communities.
(emphasis mine)
This is *quite* a statement, do you care to back it up?
I have good records showing that I have been requesting pretty basic accessibility bugfixes to stet for over two years now. Bugs have regressed as fast as they've been fixed. I understand there are general policy decisions by FSF and SFC (but I don't think either organisation publish their decisions, do they?) that stet gets no development funding besides a 2008 Google-sponsored student (whose work isn't yet published), so I call this situation *deliberate*.
stet may be useful for lawyers and big corporations, but it doesn't collaborate well.
But writing a license is *fundamentally* different from writing code.
Who's saying otherwise?
We're talking about producing a legal document. This takes people with decades of legal experience and training. The "patches welcome", "lets do this on a wiki" approach isn't going to work.
We're also talking about producing a legal document to help a community. That takes community involvement, accessibility and so on, building on the wealth of good knowledge about these subjects. (By the way, in case anyone isn't aware, I have some experience contributing to legal documents as a local councillor now.) A star chamber approach, where we don't even know who's in the chamber until late in the day, also isn't going to work.
Regards,
On Tue, Nov 04, 2008 at 06:17:09PM +0000, MJ Ray wrote:
It wasn't a personal message, sorry if I offended you.
Well, this message also seems to be very personal (you, you, you) and I feel that it's in an utterly inappropriate register for discussion.
It was personal in so much as it largely addressed things that you were doing.
It was public in so much as I was criticising your conduct, which I find to be inappropriate and not conducive to a productive debate.
Best,
The conclusion became more interesting than the main reply, so I'll put that first...
So, to conclude this subthread, some writers here think that the FDL-1.3's general approval of CC-By-SA doesn't lend FSF endorsement to CC in general, and that the new FDL-1.3 "escape route" clause is no worry - but are they the mainstream, or just a vociferous minority?
I think there are two interesting cases arising from FDL-1.3:-
1. any operator could now convert FDL'd manuals for software that are held in wikis or other MMCs to CC-By-SA - I've not found any major GNU software affected (Gnash's manual wiki is FDL 1.2 without an "or later", for example) but it looks like the German-language OpenOffice.org docs could be converted (the US-language ones are LGPL already), as could Azureus manuals;
2. it narrowly fails to convert some of the free software manuals which had been harmed by mis-application of the FDL (usually including the Front Cover Text "a GNU manual" when it's not a GNU manual, which I think would make publishing it some type of fraud) followed by a licensor becoming uncontactable.
Did anyone have an inkling that this was coming and kept pumping FDL'd manuals into a wiki just in case? ;-)
[ENDS]
Now that reply I mentioned...
Noah Slater nslater@bytesexual.org wrote:
It was public in so much as I was criticising your conduct, which I find to be inappropriate and not conducive to a productive debate.
What I find "inappropriate and not conducive to a productive debate" is making surprising claims such as
[...] The FSF defined "free" documentation long before Debian or (I presume) you did.
(Noah Slater, 2008-11-04 15:01 +000)
and then totally cutting out the request for a link to that definition, in a similar fashion to how a source link gets cut when accusing me of misrepresenting a quote. Cute editing. Pick and choose which questions to answer, why not... but I expect if I tried that, I'd be hauled over hot hail.
2008/11/5 MJ Ray mjr@phonecoop.coop:
Did anyone have an inkling that this was coming and kept pumping FDL'd manuals into a wiki just in case? ;-)
Uh, it's only been on the cards (GFDL and CC-by-sa compatibility in some regard) for, what, two years now. If you're surprised, you weren't paying attention, and it's hard to see that as a fault with everyone else.
- d.
David Gerard wrote:
2008/11/5 MJ Ray mjr@phonecoop.coop:
Did anyone have an inkling that this was coming and kept pumping FDL'd manuals into a wiki just in case? ;-)
Uh, it's only been on the cards (GFDL and CC-by-sa compatibility in some regard) for, what, two years now. If you're surprised, you weren't paying attention, and it's hard to see that as a fault with everyone else.
There's a rather large difference between "compatibility" and "relicensing without an author's permission", though.
Cheers,
Alex.
* Alex Hudson wrote, On 05/11/08 11:37:
David Gerard wrote:
2008/11/5 MJ Ray mjr@phonecoop.coop:
Did anyone have an inkling that this was coming and kept pumping FDL'd manuals into a wiki just in case? ;-)
Uh, it's only been on the cards (GFDL and CC-by-sa compatibility in some regard) for, what, two years now. If you're surprised, you weren't paying attention, and it's hard to see that as a fault with everyone else.
There's a rather large difference between "compatibility" and "relicensing without an author's permission", though.
It's helped me make up my mind about the "or later" clause though... I always thought the main issue would be trust of future as yet unknown FSF leaders.
I wonder that wikipedia wasn't just mentioned outright instead of a vaguer sounding clause that has an additional unknown quantity of leaks
Sam
"Sam Liddicott" sam@liddicott.com writes:
I wonder that wikipedia wasn't just mentioned outright instead of a vaguer sounding clause that has an additional unknown quantity of leaks
I wondered this too. Maybe the answer is that it wouldn't be legally sound to do it that way. Or maybe it's that FSF is acknowledging that the Wikipedia Foundation is just one publisher of Wikipedia.
Many organisations publish modified versions of Wikipedia, so it's useful for them to be able to be able to move their version to cc-by-sa instead of having to drop their version, take a fresh copy of Wikipedia after the (possible) change of licence, and redo their changes. Just a guess.
Ciaran O'Riordan wrote:
Many organisations publish modified versions of Wikipedia, so it's useful for them to be able to be able to move their version to cc-by-sa instead of having to drop their version, take a fresh copy of Wikipedia after the (possible) change of licence, and redo their changes. Just a guess.
So my post-2009 scenario is going to rear its head with forks of Wikipedia that are going to be potentially incompatible - people still on GFDL at that point would be creating modifications which can't be merged back.
Going through the Wikimedia discussion, the whole idea of a mostly dual-licensed but somewhat CC-BY-SA-only Wikipedia sounded pretty bad. Having forks would be worse, though.
I have to say, I'm really struggling with this. The only "bright side" I see to this is that the potential damage is mostly limited in scope and time...
Cheers,
Alex.
Alex Hudson home@alexhudson.com writes:
So my post-2009 scenario is going to rear its head with forks of Wikipedia that are going to be potentially incompatible - people still on GFDL at that point would be creating modifications which can't be merged back.
Yes, but isn't this inevitable no matter how the change of licence is done?
In which case it's a general problem, not a problem created by the FSF/CC/WP solution.
Ciaran O'Riordan wrote:
Alex Hudson home@alexhudson.com writes:
So my post-2009 scenario is going to rear its head with forks of Wikipedia that are going to be potentially incompatible - people still on GFDL at that point would be creating modifications which can't be merged back.
Yes, but isn't this inevitable no matter how the change of licence is done?
No, it's only a problem because the ability to relicense is time-limited.
In previous drafts I saw, there was an idea of being able to relicense to a "GNU Wiki License" which would have presumably been CC-BY-SA compatible without needing to relicense to that license - I was imagining a license akin to the GFDL in the way the LGPL is akin to the GPL.
If that option had been open-ended, it would have meant that forks could have been resolved to each other. With the current license, any forks unresolved before August 2009 are unresolvable again.
Cheers,
Alex.
Alex Hudson home@alexhudson.com writes:
Ciaran O'Riordan wrote:
Yes, but isn't this inevitable no matter how the change of licence is done?
No, [...]
In previous drafts I saw, there was an idea of being able to relicense to a "GNU Wiki License" which would have presumably been CC-BY-SA compatible without needing to relicense to that license[...]
If that option had been open-ended[...]
It's possible that after trying, they decided that licence compatibility was impossible.
I'm also not sure if licence compatibility can be open-ended. If the hypothetical GNU Wiki License was compatible with cc-by-sa-3.0, then the window of compatibility would close when cc-by-sa was changed in an incompatible way. So there's an end either way, and dates are probably a cleaner end because they're fixed and they don't create the inconvenience that decisions to update cc-by-sa in good-but-incompatible ways now also invoke this end of compatibility.
Ciaran O'Riordan wrote:
Alex Hudson home@alexhudson.com writes:
In previous drafts I saw, there was an idea of being able to relicense to a "GNU Wiki License" which would have presumably been CC-BY-SA compatible without needing to relicense to that license[...]
It's possible that after trying, they decided that licence compatibility was impossible.
That seems pretty speculative. To quote Yoda, it's do or do not - there is no "try", just the willingness to make the changes. It seems to me that the only stumbling block is the removal of incompatible restrictions in the FDL.
I'm also not sure if licence compatibility can be open-ended. If the hypothetical GNU Wiki License was compatible with cc-by-sa-3.0, then the window of compatibility would close when cc-by-sa was changed in an incompatible way.
Even in that scenario, it could be fixed with an updated license invoked via the "or later.." clause, again assuming that the willingness was there (& the change on the other side was something we were happy with). That's no different a situation to (e.g.) GPLv3 and the Apache v2 licenses.
I admire your willingness to attempt to hypothesise about reasons this could be a good thing or the only solution, but it still boils down to the fact that "or later.." is being used to relicense works to CC-BY-SA without the original author's permission. "Or later..." is supposed to be used to update licenses to reflect the same spirit but respond to modern needs. Introducing the possibility of an irreconcilable fork doesn't seem to me to be in the same spirit.
Cheers,
Alex.
On Wed, 2008-11-05 at 14:07 +0000, Alex Hudson wrote:
Ciaran O'Riordan wrote:
Alex Hudson home@alexhudson.com writes:
In previous drafts I saw, there was an idea of being able to relicense to a "GNU Wiki License" which would have presumably been CC-BY-SA compatible without needing to relicense to that license[...]
It's possible that after trying, they decided that licence compatibility was impossible.
That seems pretty speculative. To quote Yoda, it's do or do not - there is no "try", just the willingness to make the changes. It seems to me that the only stumbling block is the removal of incompatible restrictions in the FDL.
I'm also not sure if licence compatibility can be open-ended. If the hypothetical GNU Wiki License was compatible with cc-by-sa-3.0, then the window of compatibility would close when cc-by-sa was changed in an incompatible way.
Even in that scenario, it could be fixed with an updated license invoked via the "or later.." clause, again assuming that the willingness was there (& the change on the other side was something we were happy with). That's no different a situation to (e.g.) GPLv3 and the Apache v2 licenses.
I admire your willingness to attempt to hypothesise about reasons this could be a good thing or the only solution, but it still boils down to the fact that "or later.." is being used to relicense works to CC-BY-SA without the original author's permission. "Or later..." is supposed to be used to update licenses to reflect the same spirit but respond to modern needs. Introducing the possibility of an irreconcilable fork doesn't seem to me to be in the same spirit.
Or maybe that's what most authors thought was the right direction.
The GFDL has always been a controversial license and it was clear very soon after Wikimedia adopted it that it's language was not right for that content. Most people agree that a CC-BY-SA would have been a more appropriate license, so maybe the update *does* reflect what most authors thought was the right direction. It's clearly impossible to please everyone, you will always find someone that claim it's not right, and FSF lost his mind, etc..., but if you look at things in perspective maybe you will find out that the common feeling is that this is indeed the right move, and will let people that mistakenly adopted the GFDL not fully understanding its complexities fix the licensing and move toward a license that is no less 'free'.
Btw, the FAQ also explains why the dates are chosen that way, it seem a pretty good explanation.
Simo.
Simo,
simo wrote:
The GFDL has always been a controversial license and it was clear very soon after Wikimedia adopted it that it's language was not right for that content. Most people agree that a CC-BY-SA would have been a more appropriate license, so maybe the update *does* reflect what most authors thought was the right direction.
Well, that's more speculation - I'm quite happy to believe most / all authors think it is the right thing, but who knows?
It's pretty much beside my point, though. It's not so much whether this is right or wrong, but whether the *mechanism* is right or wrong. It sets a precedent that the FSF feel ok using the "or later" clause to re-license other people's work without their permission, and I feel that's a dangerous precedent.
I also worry about it making Wikipedia's problem worse with the forking issues, but that's really Wikipedia's problem again. And, to be frank, this has been their problem all along: they chose the wrong license, and now it's being "fixed" with this hack. I appreciate that the FSF is trying to help them; I just don't think this is the right way to do it: Wikimedia should be cleaning up their own mess.
Cheers,
Alex.
* Alex Hudson wrote, On 05/11/08 14:46:
Simo,
simo wrote:
The GFDL has always been a controversial license and it was clear very soon after Wikimedia adopted it that it's language was not right for that content. Most people agree that a CC-BY-SA would have been a more appropriate license, so maybe the update *does* reflect what most authors thought was the right direction.
Well, that's more speculation - I'm quite happy to believe most / all authors think it is the right thing, but who knows?
It's pretty much beside my point, though. It's not so much whether this is right or wrong, but whether the *mechanism* is right or wrong. It sets a precedent that the FSF feel ok using the "or later" clause to re-license other people's work without their permission, and I feel that's a dangerous precedent.
I also worry about it making Wikipedia's problem worse with the forking issues, but that's really Wikipedia's problem again. And, to be frank, this has been their problem all along: they chose the wrong license, and now it's being "fixed" with this hack. I appreciate that the FSF is trying to help them; I just don't think this is the right way to do it: Wikimedia should be cleaning up their own mess.
It's called "The tyranny of good intentions"
Sam
Ciaran O'Riordan ciaran@fsfe.org wrote:
Many organisations publish modified versions of Wikipedia, so it's useful for them to be able to be able to move their version to cc-by-sa instead of having to drop their version, take a fresh copy of Wikipedia after the (possible) change of licence, and redo their changes. Just a guess.
Note also that the clause says "operator" and not "Publisher". Publisher is defined in the early part of the FDL, so I think operator is in the classic IT sense of computer operator (because "user" would be confusing).
So even if Wikipedia doesn't relicense as a whole, most of Wikipedia probably will become cc-by-sa soon.
On Wed, Nov 05, 2008 at 12:52:26AM +0000, MJ Ray wrote:
Noah Slater nslater@bytesexual.org wrote:
It was public in so much as I was criticising your conduct, which I find to be inappropriate and not conducive to a productive debate.
Yes, I feel you used this as a platform to unrelatedly bash the FSF.
What I find "inappropriate and not conducive to a productive debate" is making surprising claims such as
[...] The FSF defined "free" documentation long before Debian or (I presume) you did.
(Noah Slater, 2008-11-04 15:01 +000)
and then totally cutting out the request for a link to that definition
Yes, I was specifically talking about the FDL it's self!
un a similar fashion to how a source link gets cut when accusing me of misrepresenting a quote. Cute editing.
I only leave in enough needed for context, it's not about cute editing.
Pick and choose which questions to answer, why not... but I expect if I tried that, I'd be hauled over hot hail.
I only the answer the questions I feel it is worth answering.
MJ Ray mjr@phonecoop.coop writes:
Isn't it approving CC's chequered history and future a little?
It's endorsing CC.
Endorsing CC isn't a big problem nowadays because they got rid of the licences which didn't allow non-commercial sharing. They now have a set of "baseline" rights that all their licences must give: http://wiki.creativecommons.org/Baseline_Rights
Saying that it endorses CC's history is going too far, and it's not clear what that would mean. Would it be having bad licences in the past? (booo) Would it be removing the bad licences from their repertoire? (yaay)
BTW, while searching for that link, I also found Lessig's comments: http://lessig.org/blog/2008/11/enormously_important_news_from.html
(lessig.org isn't working for me, here's a mirror) http://www.strategist.org.uk/politics/enormously-important-news-from-the-fre...
And a blog entry on the front of creativecommons.org: http://creativecommons.org/weblog/entry/10443
And CC's "statement of intent" earlier this year regarding the SA licences: http://wiki.creativecommons.org/CC_Attribution-ShareAlike_Intent (long, didn't read)