Re: A possible weakened rephrase of clause 5d [was: Re: GPL v3 Draft 3- text and comments]
On Thu, 12 Apr 2007 15:15:38 +0200 Lasse Reichstein Nielsen wrote:
On Thu, 12 Apr 2007 00:46:06 +0200, Francesco Poli
[EMAIL PROTECTED] wrote:
OK, this is my attempt to rephrase clause 5d in a form that is weak
enough to be less harmful than clause 2c of GPLv2:
begin proposed text
d) If the Program has interactive user interfaces which display
legal notices, this feature must be preserved in each
interactive interface that is also present in the work. In
this subsection, an interactive interface is said to display
legal notices if it includes a convenient feature that
displays an appropriate copyright notice, and tells the user
that there is no warranty for the work (unless you provide a
warranty), that licensees may convey the work under this
License, and how to view a copy of this License.
~ end proposed text ~
What does each interactive interface mean? I.e., what delimiters an
interface?
The term is not defined in a precise manner, but I think that it's clear
enough.
Does a program with more than one window have more than one interface?
Not in my understanding.
AFAICS, a program with two interfaces could be something that may be
started in such a way to offer a GUI, but also in a different way to
offer an interactive textual command interpreter, for instance...
What if it is a web application with several clients? That each open
different
windows? (This is not specified in the GPLv3 draft either, but I don't
think
it is at all obvious.)
As long as the web application presents the same kind of window to each
client, it has only one interface, AFAICT.
Anyway, consider this case:
Program A has a graphical user interface with an about feature
showing legal
notices. It also has some very clever algorithms in its belly.
Program B wants to use the algorithms, but cares not for the
interface of program A.
Minor nitpick: program B wants to reuse the algorithm *implementations*
(algorithms are not copyrightable, only their implementations are).
But anyway, let's go on...
Program B has an interactive textual user interface.
(Can the *feature* be preserved when going from graphical to
textual?)
Maybe, but my proposed clause is *not* intended to mandate this, as no
interactive interface of program A is kept in program B.
Maybe it's clearer if the clause says:
this feature must be preserved in each interactive interface that is
present both in the Program and in the work
Program B extracts the algorithms and uses them, but dumps the
interface. Should Program B preserve the legal notices? Obviously
yes.
As stated above, my proposed clause is *not* intended to mandate the
preservation of the feature that displays legal notices, when the
affected interactive interfaces are dropped.
Of course program B must have appropriate copyright notices and stuff
required by the rest of the GNU GPL. But, with my proposed clause,
program B would not be compelled to implement a feature to display legal
notices in the interactive interfaces that are not taken from program
A.
Now the people at Program B Software first extract the clever
algorithms and distributes them, alone, under the original license,
as allowed by
the GPL.
Then they use *this* distribution in program B. The program they use
has
no interactive
user interface at all, and no feature displaying legal notices, so
program B isn't
required to have it either.
That is the same exact result that we would get in your previous example
(assuming that my proposed clause works as intended...).
This problem is inherent to linking a requirement relating to an
entire program
to only parts of that program (here: the user interface). Anybody can
remove those
parts and distribute the rest, thereby removing the requirement.
Your version does alleviate this problem. Clause 5d of GPLv3draft3, as
you quoted it,
does. It requires that legal notices be inserted if deriving form a
program with no user interface (but not from one with a user interface
but without legal notices). It's not necessarily the same notices as
the original program, though,
just the default GPL notices.
Wait, let me understand: are you criticizing my proposed clause because
you feel that it's not restrictive enough?
If this is the case, please note that it's *intended* to be *more*
permissive than clause 5d of GPLv3draft3! After all, I said that my
proposed clause is a *weakened rephrase* of clause 5d!
Indeed, I hope that my proposed clause is more permissive than clause 2c
of GPLv2...
--
http://frx.netsons.org/doc/nanodocs/etch_workstation_install.html
Need to read a Debian etch installation walk-through?
. Francesco Poli .
GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4
pgpALrSDsv57j.pgp
A possible weakened rephrase of clause 5d [was: Re: GPL v3 Draft 3- text and comments]
On Mon, 2 Apr 2007 20:50:27 +0200 Francesco Poli wrote: On Mon, 02 Apr 2007 12:26:42 +0100 Gervase Markham wrote: Francesco Poli wrote: Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-( [...] I would like to see clause 5d dropped entirely. I agree that it's not very good. Given that persuading the FSF to drop the clause entirely at this late stage is unlikely, can we come up with a form of wording to suggest which at least makes it no worse than GPLv2? Well, I think that clause 2c of GPLv2 is already suboptimal (even though not so bad to become non-free) and should have been weakened in GPLv3, rather than strengthened, as it currently is... Anyway, we *could* perhaps try to rephrase it in a weaker form, assuming that the FSF won't drop it, but could consider weakening it, instead... I don't know whether that will actually be the case, but maybe it's worth trying. It won't be easy, though: the simplest solution remains dropping it entirely, IMO. I will try and find the time to draft something and come back later (or at least, I hope...). OK, this is my attempt to rephrase clause 5d in a form that is weak enough to be less harmful than clause 2c of GPLv2: begin proposed text d) If the Program has interactive user interfaces which display legal notices, this feature must be preserved in each interactive interface that is also present in the work. In this subsection, an interactive interface is said to display legal notices if it includes a convenient feature that displays an appropriate copyright notice, and tells the user that there is no warranty for the work (unless you provide a warranty), that licensees may convey the work under this License, and how to view a copy of this License. ~ end proposed text ~ Constructive critics and suggestions are welcome. For reference, here's clause 5d of GPLv3draft3: |d) If the work has interactive user interfaces, each must |include a convenient feature that displays an appropriate |copyright notice, and tells the user that there is no warranty for |the work (unless you provide a warranty), that licensees may |convey the work under this License, and how to view a copy of this |License. Specifically, if the interface presents a list of user |commands or options, such as a menu, a command to display this |information must be prominent in the list; otherwise, the |work must display this information at startup. However, if the |Program has interactive interfaces that do not comply with this |subsection, your work need not make them comply. and clause 2c of GPLv2: |c) If the modified program normally reads commands interactively |when run, you must cause it, when started running for such |interactive use in the most ordinary way, to print or display an |announcement including an appropriate copyright notice and a |notice that there is no warranty (or else, saying that you provide |a warranty) and that users may redistribute the program under |these conditions, and telling the user how to view a copy of this |License. (Exception: if the Program itself is interactive but |does not normally print such an announcement, your work based on |the Program is not required to print an announcement.) -- http://frx.netsons.org/doc/nanodocs/etch_workstation_install.html Need to read a Debian etch installation walk-through? . Francesco Poli . GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpfKEO33bhsN.pgp Description: PGP signature
Re: GPL v3 Draft 3- text and comments
* Gervase Markham [EMAIL PROTECTED] [070404 01:09]: Calling Affero code proprietary is a pretty big stretch. Yes, there's a clause in there which is a restriction on modification - so it's not entirely free. But you still have to release the source to modifications, source follows the binary - all that GPL goodness, because the Affero license is based on the GPL. While proprietary might not be the proper term using the dictionary meaning of proprietary, it's often used in the meaning more evil than I want to deal with. And in that meaning it fits. Forcing users to surrender their right to run for their right to modify, is nothing I'd consider less free, it's not free at all. And, from a practical point of view, there's hardly any code under the Affero. Proprietary software companies are not going to relicense under the Affero in order to link with GPLed code - because the Affero doesn't let them keep their code secret. So, imposing the condition that everyone is allowed to make the code non-free is OK, because it would be a form of non-free that currently no proprietary software companies has any use for? I don't choose my licenses to destroy proprietary software companies, but to keep my code free. Some of your other points were good, but this one is really not going to be a problem in practice. It's an obvious and large problem in theory. Free software is to an large part idealism. Idealism and big problems in the theory don't fit well together. Hochachtungsvoll, Bernhard R. Link -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: GPL v3 Draft 3- text and comments
On Wed, 04 Apr 2007 00:09:30 +0100 Gervase Markham wrote: Francesco Poli wrote: Well, *when* I want a copyleft, I want one that *actually works*... Exemptions for specific incompatible licenses should be left out of the license text (so that who wants them can add them as additional permissions). *When* I choose the GNU GPL, I want to prevent my code from being linked with proprietary code (including AfferoGPL'd code). I'm simplifying things to a great extent here, but I think what I mean is clear enough... Not-quite-DFSG-free != proprietary. On the contrary: Not-quite-DFSG-free == non-free, even though close to the freeness boundary == proprietary, even though close to the freeness boundary By definition, whatever is not free, is proprietary. It may be closer to or farther from the freeness boundary, but it remains proprietary. Likewise DFSG-free works may be closer to or farther from the freeness boundary (on the right side, this time!), but they still are DFSG-free. Calling Affero code proprietary is a pretty big stretch. IMHO, it's just calling things with their name... [...] Some of your other points were good, but this one is really not going to be a problem in practice. The problem is that (if this clause is not dropped) GPLv3'd code will be linkable to non-free-restriction-encumbered code. That's not in the spirit of the GNU GPL v2. -- http://frx.netsons.org/doc/nanodocs/etch_workstation_install.html Need to read a Debian etch installation walk-through? . Francesco Poli . GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpNJCSw4AOdP.pgp Description: PGP signature
Re: GPL v3 Draft 3- text and comments
Francesco Poli wrote: Not-quite-DFSG-free == non-free, even though close to the freeness boundary == proprietary, even though close to the freeness boundary By definition, whatever is not free, is proprietary. I was using proprietary in what I thought was its fairly common meaning, i.e. closed source, controlled by only one company. I have no intention of getting into a fight about whether the Affero additional restriction is acceptable or free or whatever. The FSF thinks it's free; other people disagree. Their reasons are credible. I don't like it. But my point is that you are acting as if this exception turns all GPLed code into LGPLed code - i.e. Microsoft can come along and link it into Windows, or whatever. But that's obviously not true. The only non-GPLed code your GPLed code can be linked with is code that also follows the GPL exactly _except_ that it has a single additional restriction on modification to a small part of it. This may not be a good thing, but it's not even on the same planet as some of the scenarios the phrase being able to link with proprietary code could cover. And considering the small amount of code actually covered by the Affero GPL (and that there's very little evidence that version 2 of the Affero GPL will cause it to suddenly surge in popularity) then it's also very unlikely that code you write will end up in this situation. Lastly, the FSF is keeping their promises. If you can think of a better way for them to do so (and this way is already a whole load better than their last attempt), then suggest it. So I'd suggest you concentrate your efforts on the other points you made in your analysis, which were good and reasonable. In order to facilitate this, I'm not going to contribute further to this discussion, because its very continuance is counter-productive to its point. The problem is that (if this clause is not dropped) GPLv3'd code will be linkable to non-free-restriction-encumbered code. That's not in the spirit of the GNU GPL v2. True. And Debian can easily refuse to distribute applications so linked. Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: GPL v3 Draft 3- text and comments
In message [EMAIL PROTECTED], Gervase Markham [EMAIL PROTECTED] writes Francesco Poli wrote: Not-quite-DFSG-free == non-free, even though close to the freeness boundary == proprietary, even though close to the freeness boundary By definition, whatever is not free, is proprietary. I was using proprietary in what I thought was its fairly common meaning, i.e. closed source, controlled by only one company. Actually, I wish people would use the word proprietary in its proper meaning, ie the property of someone. Linux is proprietary, because it's copyrighted and owned by people. Oh - and you know where the fairly common meaning came from? Yes. Our friends at Microsoft - Windows is open, Unix is proprietary. Cheers, Wol -- Anthony W. Youngman - [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: GPL v3 Draft 3- text and comments
On Wed, 04 Apr 2007 18:40:12 +0100 Gervase Markham wrote: Francesco Poli wrote: Not-quite-DFSG-free == non-free, even though close to the freeness boundary == proprietary, even though close to the freeness boundary By definition, whatever is not free, is proprietary. I was using proprietary in what I thought was its fairly common meaning, i.e. closed source, controlled by only one company. I instead used the above-stated definition: now I think we clarified our terminology... I have no intention of getting into a fight about whether the Affero additional restriction is acceptable or free or whatever. The FSF thinks it's free; other people disagree. Their reasons are credible. I don't like it. But my point is that you are acting as if this exception turns all GPLed code into LGPLed code - i.e. Microsoft can come along and link it into Windows, or whatever. But that's obviously not true. I'm not acting like that. I'm saying that the clause under consideration permits linking with code encumbered by more restrictive terms than the ones of the GNU GPL v3 (this is true for sure, otherwise there would be no reason for special-casing the AfferoGPL v2...). This is a flaw for a license that claims to be a copyleft in the spirit of the GNU GPL v2. I'm simply pointing out this flaw. The only non-GPLed code your GPLed code can be linked with is code that also follows the GPL exactly _except_ that it has a single additional restriction on modification to a small part of it. Wait, AFAIK, there are no publicly disclosed drafts of the AfferoGPL v2 yet: hence we cannot know what it will look like. It could be better than how you depict it, but it could even be worse. We simply do not know. The only thing we know for sure is that it will be more restrictive than the GNU GPL v3, as explained above. This may not be a good thing, but it's not even on the same planet as some of the scenarios the phrase being able to link with proprietary code could cover. It's on the same planet (Propworld), just closer to the spaceport where ships that go to the other planet (DFSGworld) take off! ;-) And considering the small amount of code actually covered by the Affero GPL (and that there's very little evidence that version 2 of the Affero GPL will cause it to suddenly surge in popularity) then it's also very unlikely that code you write will end up in this situation. Do not underestimate the power of FSF endorsement: see how popular and widespread the GFDL is, despite its relatively young age and debian-legal's efforts to explain its freeness issues... :-( Lastly, the FSF is keeping their promises. They are keeping recent promises, by forgetting less recent ones. Back in 1991 the FSF wrote GNU GPL v2 which, in section 9, stated: | The Free Software Foundation may publish revised and/or new versions | of the General Public License from time to time. Such new versions | will be similar in spirit to the present version, but may differ in | detail to address new problems or concerns. One of the key principles of the GNU GPL v2 copyleft mechanism is the prohibition to impose further restrictions. A license that allows adding further restrictions is *not* similar in spirit to the GNU GPL v2, IMO. If you can think of a better way for them to do so (and this way is already a whole load better than their last attempt), then suggest it. My suggestion for the FSF is: first, keep promises made to the *whole* free software community inside one of the most important documents ever published by the FSF (namely the GNU GPL v2), *then*, if at all possible, keep promises made to the *few* AfferoGPL enthusiasts. Please note that I don't actually know how many people love the AfferoGPL, I just assumed you're right in saying that there are not many of them... So I'd suggest you concentrate your efforts on the other points you made in your analysis, which were good and reasonable. In order to facilitate this, I'm not going to contribute further to this discussion, because its very continuance is counter-productive to its point. OK, let's stop here, then. -- http://frx.netsons.org/doc/nanodocs/etch_workstation_install.html Need to read a Debian etch installation walk-through? . Francesco Poli . GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpEDA5pvidqF.pgp Description: PGP signature
Re: GPL v3 Draft 3- text and comments
Francesco Poli wrote: On Mon, 02 Apr 2007 12:26:42 +0100 Gervase Markham wrote: I can't see any judge with a decent grasp of English or the notion of a legal notice or author attribution permitting the attachment of the GNU Manifesto to a work under this clause. Can you give a concrete example of a problematic situation you see? I cannot depict a specific scenario off the top of my head, but my alarm bell rang as soon as I saw the word preservation coupled with undefined (and hence vague) terms as reasonable legal notice and author attribution. Undefined in the license != vague. There are lots of English words the license uses which it does not explicitly define, and yet we seem to manage to understand it pretty well. An author attribution is text which tells you the name of an author. A reasonable legal notice is any notice of relevance to and on the topic of the legal situation surrounding the product. I really can't see any GFDL-like insert GNU Manifesto here problems with this. Since the clause does not seem to be designed as sufficiently narrow to avoid posing nasty problems in the future, I assumed the worst case scenario and concluded that the clause will bite. That was my line of reasoning. How would you rephrase it? BTW, does this section make GPLv3 compatible with the license of OpenSSL? I don't know: I didn't check, as it was not my primary concern. It was a question for the group :-) This clause is a permission to link; therefore, as I read it, the GPLv3 copyleft weakens to an LGPL-style copyleft in the case of linking with the Affero GPL. Each bit of code remains under its own license. Yes, and I dislike it: it sounds as (and probably actually is...) an endorsement of the AfferoGPL v2 by the FSF. Yes, it is. If you never use the Affero GPL, is it really a big deal? They made a promise ages ago, and now are looking for the least painful way to keep it. Having a special exception everyone else can ignore is a far better solution than the previous section-7-based attempt. P.S.: Please do not reply to me, Cc:ing the list, as I didn't asked you to do so. Sorry. It wasn't intentional. Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: GPL v3 Draft 3- text and comments
On Mon, 2 Apr 2007 21:50:12 -0400 Joe Smith wrote: [...] I think most courts do not rule on uncontested fact. This clause is probably intended to prevent EvilCorp(TM) from claiming that the work falls into that class. The other party is unlikely to contest that, claiming the work does fall into that class, as that could only hurt said other party. I'm not sure I follow your reasoning, sorry. Anyway, as I stated, what concerns me is that declaring a fact as true in a license does not make it magically true. It could instead prevent the adoption of the license for some works, or, at any rate, become problematic in some scenarios... [...] I think this stems from source code not requireing a patent license. So if the source code is available, the patent can be bypassed by having the consumer download and compile the code themselves. Of course all of this can only protrect the downstream consumer if the compiled binaries are not being passed around. Hence, with this kind of protection from patents we lose the permission to distribute binaries! It does not look as a good enough protection, then... -- http://frx.netsons.org/doc/nanodocs/etch_workstation_install.html Need to read a Debian etch installation walk-through? . Francesco Poli . GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpiLicZpM9Wc.pgp Description: PGP signature
Re: GPL v3 Draft 3- text and comments
Francesco Poli [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED] On Mon, 2 Apr 2007 21:50:12 -0400 Joe Smith wrote: [...] I think this stems from source code not requireing a patent license. So if the source code is available, the patent can be bypassed by having the consumer download and compile the code themselves. Of course all of this can only protrect the downstream consumer if the compiled binaries are not being passed around. Hence, with this kind of protection from patents we lose the permission to distribute binaries! It does not look as a good enough protection, then... I agree. It does protect the freedom of the end user, but without more effort on the part of the licensor, things can be problematic. I'm not sure about commerical entities compiling from source code and using the application. I suspect that sort of use may still need a patent license. Thus we have effetive discrimination against businesses. (That discrimination is not part of the licence, but is part of the source-code only software patent workaround.) Thus ideally the GPL v3 would not allow public availability of source code as an option, but require further protections. However that could be a problem. There has historicly been a fair amount of GPLv2 covered code that was distributed source-only because of patent issues. On the other hand, most of the time most of the time that happened the party did not have an actual patent license, so that clause would not apply to them. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: GPL v3 Draft 3- text and comments
On Tue, 03 Apr 2007 14:17:42 +0100 Gervase Markham wrote: Francesco Poli wrote: [...] I cannot depict a specific scenario off the top of my head, but my alarm bell rang as soon as I saw the word preservation coupled with undefined (and hence vague) terms as reasonable legal notice and author attribution. Undefined in the license != vague. You're right, but, in the present case, I felt that those terms could be intepreted in a broader sense than intended. There are lots of English words the license uses which it does not explicitly define, and yet we seem to manage to understand it pretty well. An author attribution is text which tells you the name of an author. A reasonable legal notice is any notice of relevance to and on the topic of the legal situation surrounding the product. I really can't see any GFDL-like insert GNU Manifesto here problems with this. I'm afraid that some extra care should be taken to avoid such problems... Since the clause does not seem to be designed as sufficiently narrow to avoid posing nasty problems in the future, I assumed the worst case scenario and concluded that the clause will bite. That was my line of reasoning. How would you rephrase it? That's a hard one... I will try and find the time to draft something (h... I'm saying so too often lately!). [...] This clause is a permission to link; therefore, as I read it, the GPLv3 copyleft weakens to an LGPL-style copyleft in the case of linking with the Affero GPL. Each bit of code remains under its own license. Yes, and I dislike it: it sounds as (and probably actually is...) an endorsement of the AfferoGPL v2 by the FSF. Yes, it is. If you never use the Affero GPL, is it really a big deal? Well, *when* I want a copyleft, I want one that *actually works*... Exemptions for specific incompatible licenses should be left out of the license text (so that who wants them can add them as additional permissions). *When* I choose the GNU GPL, I want to prevent my code from being linked with proprietary code (including AfferoGPL'd code). I'm simplifying things to a great extent here, but I think what I mean is clear enough... -- http://frx.netsons.org/doc/nanodocs/etch_workstation_install.html Need to read a Debian etch installation walk-through? . Francesco Poli . GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpH4Yk4zQWSm.pgp Description: PGP signature
Re: GPL v3 Draft 3- text and comments
Francesco Poli wrote: Well, *when* I want a copyleft, I want one that *actually works*... Exemptions for specific incompatible licenses should be left out of the license text (so that who wants them can add them as additional permissions). *When* I choose the GNU GPL, I want to prevent my code from being linked with proprietary code (including AfferoGPL'd code). I'm simplifying things to a great extent here, but I think what I mean is clear enough... Not-quite-DFSG-free != proprietary. Calling Affero code proprietary is a pretty big stretch. Yes, there's a clause in there which is a restriction on modification - so it's not entirely free. But you still have to release the source to modifications, source follows the binary - all that GPL goodness, because the Affero license is based on the GPL. And, from a practical point of view, there's hardly any code under the Affero. Proprietary software companies are not going to relicense under the Affero in order to link with GPLed code - because the Affero doesn't let them keep their code secret. Some of your other points were good, but this one is really not going to be a problem in practice. Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: GPL v3 Draft 3- text and comments
Francesco Poli wrote: Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-( It's an inconvenience and border-line with respect to freeness. Actually this clause restricts how I can modify what an interactive program does when run. It mandates a feature that I *must* implement in *any* interactive interface of my modified work. It's very close to place an unacceptable restriction on modification. What is more awkward is that it seems that when a non-interactive work is modified so that it becomes an interactive work, the modifier is *compelled* to implement these features in *any* newly created interactive interface... I would like to see clause 5d dropped entirely. I agree that it's not very good. Given that persuading the FSF to drop the clause entirely at this late stage is unlikely, can we come up with a form of wording to suggest which at least makes it no worse than GPLv2? I would be happy to see all these permissions to add restrictions entirely dropped from Section 7. === not a Freeness issue, but a great loss, since, if this mechanism is kept in the final GPLv3 text, GPL-compatibility will no longer be a DFSG-compliance guarantee... :-( Can you give an example of a DFSG-non-compliant term that could be introduced under section 7? b. requiring preservation of specified reasonable legal notices or author attributions in source or object code forms of material added by you to a covered work; or Kills copyleft: are these the cousins of GFDL's Invariant Sections? What exactly is a reasonable legal notice? What exactly is an author attribution? It seems that these terms are not defined anywhere in the license. I'm concerned that they could be interpreted in a broad sense and allow people to take a GPLv3'd work and add some sort of invariant long text that nobody will ever be able to remove or modify... I can't see any judge with a decent grasp of English or the notion of a legal notice or author attribution permitting the attachment of the GNU Manifesto to a work under this clause. Can you give a concrete example of a problematic situation you see? BTW, does this section make GPLv3 compatible with the license of OpenSSL? 13. Use with the Affero General Public License. Kills copyleft: compatibility with a yet unknown license This section introduces a form of compatibility with a license that is yet unreleased and thus possibly non-free: the Affero General Public License, version 2. The AfferoGPL v1 is, in my opinion, a non-free license, due to its clause 2(d). I won't restate all the reasons for my conclusions (more details in http://gplv3.fsf.org/comments/rt/readsay.html?filename=gplv3-draft-3id=1663). As a consequence, I have few hopes that the forthcoming version 2 of the AfferoGPL will be a free license. Being compatible with an unknown (and thus possibly non-free) license destroys the copyleft mechanism of the GPLv3. Destroys is a bit strong. This clause is a permission to link; therefore, as I read it, the GPLv3 copyleft weakens to an LGPL-style copyleft in the case of linking with the Affero GPL. Each bit of code remains under its own license. Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: GPL v3 Draft 3- text and comments
On Mon, 02 Apr 2007 12:26:42 +0100 Gervase Markham wrote: Francesco Poli wrote: Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-( [...] I would like to see clause 5d dropped entirely. I agree that it's not very good. Given that persuading the FSF to drop the clause entirely at this late stage is unlikely, can we come up with a form of wording to suggest which at least makes it no worse than GPLv2? Well, I think that clause 2c of GPLv2 is already suboptimal (even though not so bad to become non-free) and should have been weakened in GPLv3, rather than strengthened, as it currently is... Anyway, we *could* perhaps try to rephrase it in a weaker form, assuming that the FSF won't drop it, but could consider weakening it, instead... I don't know whether that will actually be the case, but maybe it's worth trying. It won't be easy, though: the simplest solution remains dropping it entirely, IMO. I will try and find the time to draft something and come back later (or at least, I hope...). I would be happy to see all these permissions to add restrictions entirely dropped from Section 7. === not a Freeness issue, but a great loss, since, if this mechanism is kept in the final GPLv3 text, GPL-compatibility will no longer be a DFSG-compliance guarantee... :-( Can you give an example of a DFSG-non-compliant term that could be introduced under section 7? What concerns me is an invariant text that could be introduced by exploiting clause 7b: see below for more details. b. requiring preservation of specified reasonable legal notices or author attributions in source or object code forms of material added by you to a covered work; or Kills copyleft: are these the cousins of GFDL's Invariant Sections? What exactly is a reasonable legal notice? What exactly is an author attribution? It seems that these terms are not defined anywhere in the license. I'm concerned that they could be interpreted in a broad sense and allow people to take a GPLv3'd work and add some sort of invariant long text that nobody will ever be able to remove or modify... I can't see any judge with a decent grasp of English or the notion of a legal notice or author attribution permitting the attachment of the GNU Manifesto to a work under this clause. Can you give a concrete example of a problematic situation you see? I cannot depict a specific scenario off the top of my head, but my alarm bell rang as soon as I saw the word preservation coupled with undefined (and hence vague) terms as reasonable legal notice and author attribution. Since the clause does not seem to be designed as sufficiently narrow to avoid posing nasty problems in the future, I assumed the worst case scenario and concluded that the clause will bite. That was my line of reasoning. BTW, does this section make GPLv3 compatible with the license of OpenSSL? I don't know: I didn't check, as it was not my primary concern. 13. Use with the Affero General Public License. Kills copyleft: compatibility with a yet unknown license This section introduces a form of compatibility with a license that is yet unreleased and thus possibly non-free: the Affero General Public License, version 2. The AfferoGPL v1 is, in my opinion, a non-free license, due to its clause 2(d). I won't restate all the reasons for my conclusions (more details in http://gplv3.fsf.org/comments/rt/readsay.html?filename=gplv3-draft-3id=1663). As a consequence, I have few hopes that the forthcoming version 2 of the AfferoGPL will be a free license. Being compatible with an unknown (and thus possibly non-free) license destroys the copyleft mechanism of the GPLv3. Destroys is a bit strong. I admit that weakens would have been a more appropriate term... This clause is a permission to link; therefore, as I read it, the GPLv3 copyleft weakens to an LGPL-style copyleft in the case of linking with the Affero GPL. Each bit of code remains under its own license. Yes, and I dislike it: it sounds as (and probably actually is...) an endorsement of the AfferoGPL v2 by the FSF. I don't want to open the can of worms of the FSF legal theory of linking, hence, let's assume for the sake of the argument that it's a legally sound theory. If this is the case, I don't want a general exemption for a license that will probably be non-free (and will for sure be GPLv3-incompatible, only artificially linkable due to this section and a corresponding one in its text). P.S.: Please do not reply to me, Cc:ing the list, as I didn't asked you to do so. I am a debian-legal subscriber and would rather avoid receiving the same message twice. Reply to the list only (as long as you want to send a public response). See http://www.debian.org/MailingLists/#codeofconduct for more details. Thanks. --
Re: GPL v3 Draft 3- text and comments
The following is intended to be a compression of your comments down into the most important points (generally, the areas you are concerned about), to aid further discussion. As well as some responses to your comments. (I had to manually fix the quoting, so apologies if I mess it up somewhere). Francesco Poli wrote in message news:[EMAIL PROTECTED] [...] 3. No Denying Users' Rights through Technical Measures. No covered work shall be deemed part of an effective technological measure under any applicable law fulfilling obligations under article 11 of the WIPO copyright treaty adopted on 20 December 1996, or similar laws prohibiting or restricting circumvention of such measures. Problematic: possibly untrue This clause is improved (being now denationalized), but still problematic. It could be seen as an untrue statement in some cases. How can the licensor say that the covered work won't be judged as part of an effective technological measure under a given law? That is for the courts to decide. In some scenarios, GnuPG may actually be considered part of an effective technological measure and could be deemed so by a judge... I think most courts do not rule on uncontested fact. This clause is probably intended to prevent EvilCorp(TM) from claiming that the work falls into that class. The other party is unlikely to contest that, claiming the work does fall into that class, as that could only hurt said other party. When you convey a covered work, you waive any legal power to forbid circumvention of technical measures to the extent such circumvention is effected by exercising rights under this License with respect to the covered work, Bad: possibly overreaching This clause is clearer than in the previous draft, but still troublesome, as it seems to be overreaching. For instance, it could be interpreted as covering legal powers to forbid computer crimes such as unauthorized intrusion into computer systems. E.g.: suppose that the covered work is a vulnerability scanner, or password cracker, or anyway a tool that could be used (among other things) to break into other people's computers. Using that tool in this manner is exercising a right under this License and is a circumvention of appropriate technical measures set to protect a computer system or network from unauthorized access. Gaining unauthorized access to a protected computer system or network is forbidden by law in several jurisdictions; do I waive such a legal protection, when I convey the covered work? I suggest dropping the waiver entirely, thus leaving the following disclaimer only. === waiving legal rights can be seen as a fee: this clause could fail DFSG#1 and you disclaim any intention to limit operation or modification of the work as a means of enforcing, against the work's users, your or third parties' legal rights to forbid circumvention of technical measures. Agree with your assesment, assuming the disclaming of intention could let a defentent invoke estoppel or other similar. Presumably that clause is intended to prevent the obvious workaround of moving the anti-copyprotection-circumvention law outside the copyright law. Overall, I find this to be one of the parts of the licence that is very unclear if approched without knowing it is about DCMA-style anti-circumvention laws. If one was not aware of that problem, one may well be quite confused while tying to figure out the purpose of that section [...] d) If the work has interactive user interfaces, each must include a convenient feature that displays an appropriate copyright notice, and tells the user that there is no warranty for the work (unless you provide a warranty), that licensees may convey the work under this License, and how to view a copy of this License. Specifically, if the interface presents a list of user commands or options, such as a menu, a command to display this information must be prominent in the list; otherwise, the work must display this information at startup. However, if the Program has interactive interfaces that do not comply with this subsection, your work need not make them comply. Bad: too restrictive Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-( It's an inconvenience and border-line with respect to freeness. Actually this clause restricts how I can modify what an interactive program does when run. It mandates a feature that I *must* implement in *any* interactive interface of my modified work. It's very close to place an unacceptable restriction on modification. What is more awkward is that it seems that when a non-interactive work is modified so that it becomes an interactive work, the modifier is *compelled* to implement these features in *any* newly created interactive interface... I would like to see clause 5d dropped entirely. === very close to
Re: GPL v3 Draft 3- text and comments
On Wed, 28 Mar 2007 16:07:34 -0400 Joe Smith wrote: [...] For the record: IANAL, IANADD. My comments on the new draft follows. I will send them to the FSF public consultation system RSN (since they are accepting comments for only 60 days, starting on 28 March). IANAL and IANADD either. [...] - -- GNU GENERAL PUBLIC LICENSE Discussion Draft 3 of Version 3, 28 March 2007 THIS IS A DRAFT, NOT A PUBLISHED VERSION OF THE GNU GENERAL PUBLIC LICENSE. [...] TERMS AND CONDITIONS 0. Definitions. [...] The Program refers to any copyrightable work licensed under this License. Style: Program or Work? Since the GPL is designed to be applicable to any work of authorship (not only computer programs), I would suggest using a more neutral term than the Program. Something like the Work would avoid misleading many many people into thinking that the GPL can only be applied to computer programs. [...] To modify a work means to copy from or adapt all or part of the work in a fashion requiring copyright permission, other than the making of a verbatim copy. The resulting work is called a modified version of the earlier work or a work based on the earlier work. A covered work means either the unmodified Program or a work based on the Program. Good: clear definitions The definitions of modify, modified version, work based on another work, and covered work are slightly changed with respect to GPLv3draft2, but they are still clear. It's good that the definition of modified version exploits applicable copyright law without trying to rewrite it. This ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law. [...] To propagate a work means to do (or cause others to do) anything with it that requires permission under applicable copyright law, except executing it on a computer or making modifications that you do not share. Propagation includes copying, distribution (with or without modification), making available to the public, and in some countries other activities as well. To convey a work means any kind of propagation that enables other parties to make or receive copies, excluding sublicensing. Mere interaction with a user through a computer network, with no transfer of a copy, is not conveying. Good: clear definitions The definitions of propagate and convey seem fairly clear. Again, their linking to copyright law ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law. [...] 1. Source Code. Good: clear and appropriate This section seems OK to me. The source code for a work means the preferred form of the work for making modifications to it. Object code means any non-source form of a work. Good: please keep these definitions I like the definitions of source code and Object code as they are. Good to see that they were kept unaltered. [...] 2. Basic Permissions. All rights granted under this License are granted for the term of copyright on the Program, and are irrevocable provided the stated conditions are met. This License explicitly affirms your unlimited permission to run the unmodified Program. The output from running a covered work is covered by this License only if the output, given its content, constitutes a covered work. This License acknowledges your rights of fair use or other equivalent, as provided by copyright law. Good: mentioning of fair use is an improvement Good to see that fair use and similar rights are explicitly mentioned. This is an improvement over GPLv2. Propagation of covered works that you do not convey, and making modified versions of the Program that you do not convey, are permitted without conditions, so long as your license otherwise remains in force. Conveying is permitted under the conditions stated below. Sublicensing is not allowed; section 10 makes it unnecessary. Good: overreaching patent retaliation clause has been dropped I'm glad that the patent retaliation clause has been dropped: it was overreaching IMHO. 3. No Denying Users' Rights through Technical Measures. No covered work shall be deemed part of an effective technological measure under any applicable law fulfilling obligations under article 11 of the WIPO copyright treaty adopted on 20 December 1996, or similar laws prohibiting or restricting circumvention of such measures. Problematic: possibly untrue This clause is improved (being now denationalized), but still problematic. It could be seen as an untrue statement in some cases. How can the licensor say that the covered work won't be judged as part of an effective technological measure under a given law? That is for the courts to decide. In some scenarios, GnuPG may actually be
