GPL v3 and Ruby License

Ruby
1

I went to MIT for the Free Software Foundation Associate members
meeting last weekend. I asked Stallman the question of whether the
FSF planned to contact various individuals who were using dual
licenses and encourage them to participate in commenting on GPLv3.

He sort of yelled at me a lot. Big surprise. You might be able to
find the recordings of this online sometime soon.

I talked to some other FSF members, including Executive Director Peter
Brown, and their suggestion was to encourage community members to come
use the comment system, and that they’d consider more formal outreach
programs for the second draft.

So, here I am making a suggestion that maybe is based on a bit of
paranoia. We’re already having enough of a hard time explaining the
disjunctive license of Ruby… it’ll be a lot harder to explain if
some people start using GPLv3 and others stick with GPLv2 when the
final draft rolls around.

Should the ruby community get active in commenting on GPLv3 in hopes
of making it possible to switch the ruby license to use it (alongside
the current, more permissive terms of course), or should we be
focusing on making a Ruby license that stands on it’s own, or should
we just jump this hurdle down the line when we get to it?

My main cause for concern is that I really rely on the GPL part of the
license, since the terms Matz wrote don’t seem to be an established
legal document. I will probably want to use GPLv3 alongside Matz’s
terms for my software when it comes out, but I don’t want people to be
any more confused than they have to about what the “License of Ruby”
means.

So… now I’ve gone and confused myself, but basically… I’m just
concerned about hybrid Ruby licenses using both GPLv2 and GPLv3
skulking around out there in the somewhat near future, and am hoping
that we think about it before it sneaks up on us.

2

On 4/4/06, Gregory B. [email protected] wrote:

I went to MIT for the Free Software Foundation Associate members
meeting last weekend. I asked Stallman the question of whether the
FSF planned to contact various individuals who were using dual
licenses and encourage them to participate in commenting on GPLv3.

He sort of yelled at me a lot. Big surprise. You might be able to
find the recordings of this online sometime soon.

You might be surprised, but I’m not. Stallman is the FSF’s worst enemy
at this point.

I talked to some other FSF members, including Executive Director Peter
Brown, and their suggestion was to encourage community members to come
use the comment system, and that they’d consider more formal outreach
programs for the second draft.

So, here I am making a suggestion that maybe is based on a bit of
paranoia. We’re already having enough of a hard time explaining the
disjunctive license of Ruby… it’ll be a lot harder to explain if
some people start using GPLv3 and others stick with GPLv2 when the
final draft rolls around.

Should the ruby community get active in commenting on GPLv3 in hopes
of making it possible to switch the ruby license to use it (alongside
the current, more permissive terms of course), or should we be
focusing on making a Ruby license that stands on it’s own, or should
we just jump this hurdle down the line when we get to it?

My personal recommendation would be to change Ruby such that it is
locked to the GNU GPL v2. I personally will be moving to change those
pieces of software that I have written and released under the GNU GPL so
that they are locked to v2. I see no benefit in participating in the GPL
v3 system because Stallman isn’t really interested in working with
community; he has already made his decisions and the only question is
how to word it so that it is made legal pretty much everywhere.

I do not like most of what I have seen of the v3 draft and see it as one
of the worst possible directions that could be taken.

I would personally prefer to see Ruby relicensed under something like
the MIT licence because I don’t think that there’s any merit to the
disjunctive licence support for the GNU GPL.

-austin

3

On 4/5/06, Austin Z. [email protected] wrote:

at this point.
That was sarcasm. Though, I don’t necessarily agree he is the FSF
worst enemy, i did lobby heavily for more ‘practical’ community
outreach at this past meeting. Stallman has a function to serve as a
cultural icon, and much less as a pragmatic leader. I am not sure if
you have met him or been to any FSF meetings, but that is really
apparent in person.

Should the ruby community get active in commenting on GPLv3 in hopes
of making it possible to switch the ruby license to use it (alongside
the current, more permissive terms of course), or should we be
focusing on making a Ruby license that stands on it’s own, or should
we just jump this hurdle down the line when we get to it?

My personal recommendation would be to change Ruby such that it is
locked to the GNU GPL v2. I personally will be moving to change those
pieces of software that I have written and released under the GNU GPL so
that they are locked to v2.

I was thinking about and getting paranoid about the “or any later
version” clause you see in most GPLed softwares terms. This was
proposed as a ‘fix’ by the FSF, but I see it as a back door.

I see no benefit in participating in the GPL
v3 system because Stallman isn’t really interested in working with
community; he has already made his decisions and the only question is
how to word it so that it is made legal pretty much everywhere.

Yes, I complained that it seems like a lot happens behind closed doors
at the FSF and then they pop out with their decision and ask for
critique after the fact. After my discussion with Peter B., he
mentioned that they certainly had room to make some changes in the
process for the second draft, whether they do or not is yet to be
seen. Still, I think if a few of us knock at their door and let them
know what we’re thinking, they might consider it a bit more seriously.

I do not like most of what I have seen of the v3 draft and see it as one
of the worst possible directions that could be taken.

What are your specific gripes? I was honestly ambivalent to slightly
annoyed about the DRM clauses until I heard the rationale at their
meeting, and I was strongly convinced that the new DRM protection is A
Good Thing.

Also, they are working on ensuring compatibility with the Apache
license, which sounds good too. The patent threat is supposedly going
to be softened too.

So what bad parts am I missing?

I would personally prefer to see Ruby relicensed under something like
the MIT licence because I don’t think that there’s any merit to the
disjunctive licence support for the GNU GPL.

That’s a broad stroke. Some people would like to retain copyleft.
Placing it under the MIT license would prevent that. Which of course,
some would call a good thing, but it would make people like me sad.
Still, I do think that Ruby might benefit from a single license, maybe
with a weak copyleft such as that of the LGPL (something in that
spirit)

4

On Wed, 2006-04-05 at 23:50 +0900, Austin Z. wrote:

I would personally prefer to see Ruby relicensed under something like
the MIT licence because I don’t think that there’s any merit to the
disjunctive licence support for the GNU GPL.

Matz has said that he intends to release ruby 2 using a standard license
(maybe a BSD-like license).

http://redhanded.hobix.com/cult/matzTalkedAtDevelopersSummit2005.html

It would be a great move IMHO.

Regards,
Andre

5

On 4/5/06, Gregory B. [email protected] wrote:

On 4/5/06, Austin Z. [email protected] wrote:

On 4/4/06, Gregory B. [email protected] wrote:
[…]

version" clause you see in most GPLed softwares terms. This was
proposed as a ‘fix’ by the FSF, but I see it as a back door.

Yeah. It is, really. The only saving grace is that it’s a back door both
ways: if the FSF were to implement something that no one else wants, no
one would have to adopt it.

know what we’re thinking, they might consider it a bit more seriously.

I do not like most of what I have seen of the v3 draft and see it as
one of the worst possible directions that could be taken.
What are your specific gripes? I was honestly ambivalent to slightly
annoyed about the DRM clauses until I heard the rationale at their
meeting, and I was strongly convinced that the new DRM protection is A
Good Thing.

I completely disagree, and no amount of rationalization will change my
mind on this. If they make it an optional section (e.g., GPLv3 with
DRM clause) then I can see it happening. But they are once again trying
to make moral choices for me, and expecting me to agree with them.

I frankly don’t have an issue with the idea of DRM; Stallman obviously
does. (I have issues with the implementations of DRM, and with the
overreaching powers claimed through DRM, and with the nonsensical
application of DRM to many things, but not with the idea of DRM.)

Also, they are working on ensuring compatibility with the Apache
license, which sounds good too. The patent threat is supposedly going
to be softened too.

Mmm. The patent threat is one of the few things that I like, inasmuch as
it will be limited to retaliation (you try to do patent shit against me
or anyone who uses this software, you lose the right to do anything with
this softare, too).

So what bad parts am I missing?

Mostly the DRM. Both the DRM section and the patent section, however,
make the GPLv3 into something more akin to a EULA than a source licence,
which is what the GPLv2 is (although it’s often mispresented to Windows
users as a EULA, which is not the case).

I would personally prefer to see Ruby relicensed under something
like the MIT licence because I don’t think that there’s any merit to
the disjunctive licence support for the GNU GPL.
That’s a broad stroke. Some people would like to retain copyleft.

Why? Seriously. I am reasonably certain that Matz would not accept any
change submitted solely under the GNU GPL licence, and that sort of nips
the purpose of copyleft in the bud.

At that, the disjunctive licence exists because (a) not everyone is
comfortable with the conditions expressed by Ruby’s non-GPL terms
(either because they don’t feel that it is a licence or that it clearly
states the limits of liability, etc.) or (b) it isn’t clear that they
are GPL-compatible.

So, switching Ruby to an MIT-style licence will solve both (a) and (b).
There are some people who, IMO foolishly[1], think that copyleft
protects
better than non-copyleft, but an MIT-style licence will ensure that a
given version of the Ruby source is still available and licensable.

Placing it under the MIT license would prevent that. Which of course,
some would call a good thing, but it would make people like me sad.
Still, I do think that Ruby might benefit from a single license, maybe
with a weak copyleft such as that of the LGPL (something in that
spirit).

Anything but the LGPL. Over time, I have come to believe that it is the
worst possible licence, worse even than the GPL itself (because the GPL
is at least semi-honest about how it’s forcing your source open). The
LGPL could cause problems for extension writers who do not wish to
release their source or object code.

-austin
[1] Sorry, Greg, but I really do think that copyleft is a crock, and I
think that Ruby doesn’t benefit from the GPL in any way. I think
that because there are licence zealots out there, Ruby would
benefit from GPL compatibility, but should not be tainted in any way
by the GPL if at all possible.

6

Matz has said that he intends to release ruby 2 using a standard
license
(maybe a BSD-like license).

http://redhanded.hobix.com/cult/matzTalkedAtDe>
velopersSummit2005.html

It would be a great move IMHO.

I agree, that sounds great. Thanks for the link!

+1 on BSD… I like.

Tom

7

On Apr 5, 2006, at 10:50 AM, Andre N. wrote:

http://redhanded.hobix.com/cult/matzTalkedAtDevelopersSummit2005.html

It would be a great move IMHO.

I agree, that sounds great. Thanks for the link!

James Edward G. II

8

Relative to the v2 vs v3 discussion, you might have a look at the latest
Linux Journal (May 2006). On page 48, Doc Searls extracts (with
commentary) portions of the Linux-Kernel Mailing List between January
25, 2006 and February 2, 2006 where Linus Torvalds weighs in against his
use of anything other than v2 for the Linux kernel.

I guess it depends on what you want to allow to happen in the future.
Having a ‘v2 or later’ clause in a contract does seem to open the barn
door to lots of possibilities.

Bob G

9

On Apr 5, 2006, at 5:50 PM, Andre N. wrote:

Matz has said that he intends to release ruby 2 using a standard
license
(maybe a BSD-like license).

http://redhanded.hobix.com/cult/matzTalkedAtDevelopersSummit2005.html

It would be a great move IMHO.

He says Rite may be released under a BSD -like license. However,
YARV is currently released under ruby’s license.

Asenchi said on 11 Feb 2005 at 14:01

Out of curiosity, is it possible to change a license 13 years into
a project?

matz said on 11 Feb 2005 at 16:48

I?m not sure. I?m afraid that it requires agreement from every non
trivial code contributers.

– Daniel

10

On 4/5/06, Bob G. [email protected] wrote:

Relative to the v2 vs v3 discussion, you might have a look at the latest
Linux Journal (May 2006). On page 48, Doc Searls extracts (with
commentary) portions of the Linux-Kernel Mailing List between January
25, 2006 and February 2, 2006 where Linus Torvalds weighs in against his
use of anything other than v2 for the Linux kernel.

yes, this was discussed at the FSF meeting.

I guess it depends on what you want to allow to happen in the future.
Having a ‘v2 or later’ clause in a contract does seem to open the barn
door to lots of possibilities.

Though I seem to be well known for being the one guy who supports the
GPL on this list, I’m also just asking pretty much exactly that.

Do we spend the time now removing any ‘v2 or laters’ anywhere they are
found? Or do we just expect that v3 will be used in the future? Or
do we avoid the GPL entirely, as others have suggested?

11

I really didn’t want to start a license flame war. I’ve responded
below, but please keep in mind that my personal beliefs and my
pragmatic beliefs are somewhat disjointed, and that for the
community’s sake, I’d always side with my pragmatic feelings.

Meaning… when GPLv3 comes out, do we.

a) Resist it by removing any “any later version” clauses from our
license agreements which use the license of ruby

b) Embrace it and re-release code using GPLv3 and Matz’s terms, but
make sure the FSF hears the voice of our community beforehand.

c) Can it entirely and favor an established license such as the BSD
license.

d) Draft up our own license, possibly with cartoon foxes.

On 4/5/06, Austin Z. [email protected] wrote:

one would have to adopt it.
but the tricky thing is, someone can redistribute your code under
GPLv3 even if you didn’t want it to be. So as long as the “any later
version” thing is still in your agreement, those releases are all game
for ‘upgrade’.

That’s scary to me. If I want to re-release my code under a later
version, I’ll do so.
At the same time, I imagine it’s something that’s meant to protect the
code unless there is a major flaw and redistribution needs to be done
on say, an abandoned project.

Who knows.

I was honestly ambivalent to slightly
annoyed about the DRM clauses until I heard the rationale at their
meeting, and I was strongly convinced that the new DRM protection is A
Good Thing.

I completely disagree, and no amount of rationalization will change my
mind on this.

Does any amount of rationalization change your mind on ANYTHING austin?
:wink:

Seriously though, the DRM stuff… i don’t know. Maybe they had good
propaganda, maybe they had a good point. I’m sold. People can always
use GPLv2 to avoid it, if they need or want to. There is a strong
vendetta against it in the FSF, and right or wrong, they seem pretty
damn consistent on their views.

I would personally prefer to see Ruby relicensed under something
like the MIT licence because I don’t think that there’s any merit to
the disjunctive licence support for the GNU GPL.
That’s a broad stroke. Some people would like to retain copyleft.

Why? Seriously. I am reasonably certain that Matz would not accept any
change submitted solely under the GNU GPL licence, and that sort of nips
the purpose of copyleft in the bud.

I am not so much concerned about the language itself. I’m concerned
about the fact that a lot of libraries use the License of Ruby.
(Including all of mine).

I don’t accept solely GPLed code myself for changes to Ruport, so I
guess I am a hypocrite to some extent. However… I am writing
reporting applications using Ruport that are under the GPL only, so I
like knowing that it forms a nice copylefted bundle when I send it out
to clients.

I could use some sort of proprietary license if I wanted, seeing as
ruby’s terms would permit this, but I tend to go the other way.

Of course, most of this has been for small stuff that no one would
give a damn about anyway, so maybe i’m just being fussy.

At that, the disjunctive licence exists because (a) not everyone is
comfortable with the conditions expressed by Ruby’s non-GPL terms
(either because they don’t feel that it is a licence or that it clearly
states the limits of liability, etc.) or (b) it isn’t clear that they
are GPL-compatible.

So, switching Ruby to an MIT-style licence will solve both (a) and (b).

I would not complain loudly if I saw an MIT or BSD license used for
Ruby. I would just quietly sigh, as it would mean I could no longer
use ‘the license of ruby’ for my code, and would probably go with
GPLv3 alongside something else. Which goes back to sucking.

This would probably make most people happy, of course. If it really
happens, cool.

If not, my original question comes back. Do we lock down to GPLv2, or
get ready for GPLv3, or what?

There are some people who, IMO foolishly[1], think that copyleft protects
better than non-copyleft, but an MIT-style licence will ensure that a
given version of the Ruby source is still available and licensable.

I do not think copyleft protects better than non-copyleft. I think
it’s stronger and it’s in line with a personal ethical standpoint of
mine. However, as you can see, I do not use the GPL on it’s own for
my ruby projects, mainly because I want to respect those who do not
like the idea.

Stronger is not necessarily better, of course. Stronger also can be
seen as ‘more restrictive’ by some. However, I do believe that
copyleft is not coercion because you cannot coerce someone through
licensing terms.

I am mostly in favor of weak copyleft, in most pragmatic applications
of licenses. Protect your source, ensure it will remain free
software, ensure the license can’t be buried under a proprietary
license, etc. As far as linking and integrating with non-free
software… i’d like to be an idealist and say it’s evil… but of
course, I wouldn’t have a job if that were the case :wink:

release their source or object code.
I was not suggesting using the LGPL. I hate it just as much as
everyone.
(including the FSF).

I was suggesting a sort of weak-copyleft, which does not restrict
non-free linking, but protects the freedom of the source itself. I am
not sure what that means exactly.

Maybe that you cannot modify the code and then release it lumped under
a new license. That would be enough to make me happy.

12

On Thu, Apr 06, 2006 at 02:03:34AM +0900, Gregory B. wrote:
} I really didn’t want to start a license flame war. I’ve responded
} below, but please keep in mind that my personal beliefs and my
} pragmatic beliefs are somewhat disjointed, and that for the
} community’s sake, I’d always side with my pragmatic feelings.
}
} Meaning… when GPLv3 comes out, do we.
}
} a) Resist it by removing any “any later version” clauses from our
} license agreements which use the license of ruby
}
} b) Embrace it and re-release code using GPLv3 and Matz’s terms, but
} make sure the FSF hears the voice of our community beforehand.
}
} c) Can it entirely and favor an established license such as the BSD
license.
}
} d) Draft up our own license, possibly with cartoon foxes.

You forgot one:

e) Ignore it, continue to release under v2 with the clause that allows
others to choose to license it under v3, and not include any non-v2
patches/code.

–Greg

13

Hi,

On Thu, Apr 06, 2006 at 12:15:18AM +0900, Gregory B. wrote:

[…]
What are your specific gripes? I was honestly ambivalent to slightly
annoyed about the DRM clauses until I heard the rationale at their
meeting, and I was strongly convinced that the new DRM protection is A
Good Thing.

Pointers, or care to explain?
14

On 4/5/06, Gregory B. [email protected] wrote:

b) Embrace it and re-release code using GPLv3 and Matz’s terms, but
make sure the FSF hears the voice of our community beforehand.

c) Can it entirely and favor an established license such as the BSD
license.

d) Draft up our own license, possibly with cartoon foxes.

(a) or (c). The GPLv3, at least as it exists so far, and there’s no
indication in Stallman’s past that he’ll actually listen to anyone, is a
non-starter. (d) is problematic.

but the tricky thing is, someone can redistribute your code under
GPLv3 even if you didn’t want it to be. So as long as the “any later
version” thing is still in your agreement, those releases are all game
for ‘upgrade’.

Yes, sort of. They can’t change the wording on the release to be GPLv3
or any later if my release is GPLv2. At least, that’s how I understand
it. It’s more what restrictions you as a user choose to be stuck under.

I was honestly ambivalent to slightly annoyed about the DRM clauses
until I heard the rationale at their meeting, and I was strongly
convinced that the new DRM protection is A Good Thing.
I completely disagree, and no amount of rationalization will change
my mind on this.
Does any amount of rationalization change your mind on ANYTHING
austin? :wink:

I know this is mostly a joke, but yes. I am, however, convinced that the
DRM clause is a bad thing and has no business in any licence I’ll touch.

Seriously though, the DRM stuff… i don’t know. Maybe they had good
propaganda, maybe they had a good point. I’m sold. People can always
use GPLv2 to avoid it, if they need or want to. There is a strong
vendetta against it in the FSF, and right or wrong, they seem pretty
damn consistent on their views.

Consistent? Yes. Morally wrong on this one? I think so. They’ve
overstepped the bounds of the FSF with it. I don’t want a damned EULA.

I would personally prefer to see Ruby relicensed under something
like the MIT licence because I don’t think that there’s any merit
to the disjunctive licence support for the GNU GPL.
That’s a broad stroke. Some people would like to retain copyleft.
Why? Seriously. I am reasonably certain that Matz would not accept
any change submitted solely under the GNU GPL licence, and that sort
of nips the purpose of copyleft in the bud.
I am not so much concerned about the language itself. I’m concerned
about the fact that a lot of libraries use the License of Ruby.
(Including all of mine).

I don’t accept solely GPLed code myself for changes to Ruport, so I
guess I am a hypocrite to some extent. However… I am writing
reporting applications using Ruport that are under the GPL only, so I
like knowing that it forms a nice copylefted bundle when I send it out
to clients.

I could use some sort of proprietary license if I wanted, seeing as
ruby’s terms would permit this, but I tend to go the other way.

Of course, most of this has been for small stuff that no one would
give a damn about anyway, so maybe i’m just being fussy.

Maybe.

Ruby. I would just quietly sigh, as it would mean I could no longer
use ‘the license of ruby’ for my code, and would probably go with
GPLv3 alongside something else. Which goes back to sucking.

If you changed the wording of that, you could:

“Licensed under the same terms as Ruby”.

That is, as long as you’re okay without the GPL. Which I am.

[…]

There are some people who, IMO foolishly[1], think that copyleft
protects better than non-copyleft, but an MIT-style licence will
ensure that a given version of the Ruby source is still available and
licensable.
I do not think copyleft protects better than non-copyleft. I think
it’s stronger and it’s in line with a personal ethical standpoint of
mine. However, as you can see, I do not use the GPL on it’s own for
my ruby projects, mainly because I want to respect those who do not
like the idea.

Right. I’m not really trying to indict you here; we’ve had this
discussion and you’ve shown yourself to be far more pragmatic than the
people to whom I refer.

Stronger is not necessarily better, of course. Stronger also can be
seen as ‘more restrictive’ by some. However, I do believe that
copyleft is not coercion because you cannot coerce someone through
licensing terms.

That’s a common misconception. One can coerce through licensing terms.
If the choice is “my way or the highway”, where the highway is extremely
difficult, it’s coercion. A lot of Libertarians I know would disagree me
on this point, but they’re also of the opinion that coercion can only
come from physical force. The verb “coerce”, however, is defined as: “To
cause to do through pressure or necessity, by physical, moral or
intellectual means”. A licence can be coercion.

I am mostly in favor of weak copyleft, in most pragmatic applications
of licenses. Protect your source, ensure it will remain free
software, ensure the license can’t be buried under a proprietary
license, etc. As far as linking and integrating with non-free
software… i’d like to be an idealist and say it’s evil… but of
course, I wouldn’t have a job if that were the case :wink:

Something like the MPL. The problem is, with the way that the GPL is
written, the MPL is incompatible with it. This will not be changing
under GPLv3.

I was not suggesting using the LGPL. I hate it just as much as
everyone. (including the FSF).

Mmmm. I used to recommend the LGPL. But I have also carefully read the
subtleties that Stallman put it in there that makes it a worse (more
restrictive) licence than the GPL itself.

I was suggesting a sort of weak-copyleft, which does not restrict
non-free linking, but protects the freedom of the source itself. I am
not sure what that means exactly.

It means you prefer the MPL, by and large.

-austin

15

On 4/5/06, Esteban Manchado Velázquez [email protected] wrote:

Hi,

On Thu, Apr 06, 2006 at 12:15:18AM +0900, Gregory B. wrote:

[…]
What are your specific gripes? I was honestly ambivalent to slightly
annoyed about the DRM clauses until I heard the rationale at their
meeting, and I was strongly convinced that the new DRM protection is A
Good Thing.

Pointers, or care to explain?

Check out the fsf website for loads more.

Basically, I see the issue of things like Tivo being something real
and scary. Lock your code away and don’t give users the key. That’s
not free anymore.

16

On 4/5/06, Gregory B. [email protected] wrote:

Activities - FSFE

Check out the fsf website for loads more.

Basically, I see the issue of things like Tivo being something real
and scary. Lock your code away and don’t give users the key. That’s
not free anymore.

It’s not the code that’s locked away; it’s the data. I see no reason
that TiVo should have to release their DRM code, even if I disagree
with their DRM and even if I disagree with the DMCA (which I do).
The Linux variant they use? Sure. But the DRM code? No.

Some of it comes down to the idea of IP ownership, which Stallman
clearly opposes. But I don’t. I think that the rights of ownership are
different with IP vs physical property (of necessity), but I think
that there are still important and viable rights that may even include
restricting who is allowed to use the IP involved. (I do disagree
with software and business method patents, even though I will probably
end up working on stuff that leads to a patent.)

But I also believe in the idea of binary drivers while supporting the
development of open source drivers.

-austin

17

On 4/5/06, Austin Z. [email protected] wrote:

On 4/5/06, Gregory B. [email protected] wrote:

d) Draft up our own license, possibly with cartoon foxes.

(a) or (c). The GPLv3, at least as it exists so far, and there’s no
indication in Stallman’s past that he’ll actually listen to anyone, is a
non-starter. (d) is problematic.

(d) was mostly a joke. Except for the part about the cartoon foxes.

I actually prefer a, unless somehow we could get a large community
effort on b.

though ‘doing nothing at all’ sounds like a bad idea.

the e) option proposed of keeping everything as is but not accepting
anything that is v3 or later is the most compromising, though it keeps
the confusion factor high.

I want the confusion factor lower, copylefted or not. :wink:

On 4/5/06, Austin Z. [email protected] wrote:

On 4/5/06, Gregory B. [email protected] wrote:

On 4/5/06, Austin Z. [email protected] wrote:

but the tricky thing is, someone can redistribute your code under
GPLv3 even if you didn’t want it to be. So as long as the “any later
version” thing is still in your agreement, those releases are all game
for ‘upgrade’.

Yes, sort of. They can’t change the wording on the release to be GPLv3
or any later if my release is GPLv2. At least, that’s how I understand
it. It’s more what restrictions you as a user choose to be stuck under.

I interpret it differently. I see them being able to release it under
GPLv3, and then being able to slap on v3 or later, v3 only, or
whatever they want. If I am wrong, then I am more comfortable with
this, if I am right, then I am scared.

Maybe I should submit this question to rms or someone at the FSF?

Seriously though, the DRM stuff… i don’t know. Maybe they had good
propaganda, maybe they had a good point. I’m sold. People can always
use GPLv2 to avoid it, if they need or want to. There is a strong
vendetta against it in the FSF, and right or wrong, they seem pretty
damn consistent on their views.

Consistent? Yes. Morally wrong on this one? I think so. They’ve
overstepped the bounds of the FSF with it. I don’t want a damned EULA.

The DRM stuff is an explicit moral standpoint. (for better or for
worse). But anyone who thinks the GPL isn’t more about software ethics
and politics than it is about distribution and modification rights
probably hasn’t read the license.

So… I understand your standpoint, you don’t want someone elses
ethics governing your code. I respect that. However, for those who
believe in it, it does provide some strong language to help protect
their beliefs.

The question is, in this community, what serves us best? It would
seem as if we prefer the simple, unencumbered route, we’re business
friendly in general, and the trend goes towards BSD/MIT. I’d honestly
rather see the license become BSD or MIT and have an easier
explaination for most libraries then explain this quagmire.

Of course, my code will probably always be some weird mish-mosh
including the GPL explicitly some how, for personal reasons that I
never want to be GPL incompatible.

Of course, I know that both BSD and MIT are GPL compatible. But what
if, some day, they weren’t. There goes my paranoia again. But
that’s my rationale for my personal decisions which hopefully explains
a bit to others who haven’t seen you and I discuss this before.

But if Ruby was under BSD or MIT, nothing would stop me from offering
a dual-license with BSD and the GPL or whatever I wanted. So I guess,
it wouldn’t hurt.

If you changed the wording of that, you could:

“Licensed under the same terms as Ruby”.

That is, as long as you’re okay without the GPL. Which I am.

[…]

Hmm… I could say:

Licensed disjunctively under the terms of Ruby and the GPL version foo.

Right now, that would be redundant, but if Ruby was under BSD/MIT or
any GPL compatible license, the effect would still be the same for my
code.

So I could live with that. What I can’t live with is that matz’s
terms don’t meet the Free Software Definition without the GPL.
BSD/MIT/your free software license here, would.

I do not think copyleft protects better than non-copyleft. I think
it’s stronger and it’s in line with a personal ethical standpoint of
mine. However, as you can see, I do not use the GPL on it’s own for
my ruby projects, mainly because I want to respect those who do not
like the idea.

Right. I’m not really trying to indict you here; we’ve had this
discussion and you’ve shown yourself to be far more pragmatic than the
people to whom I refer.

I am trying to find grounds along the fringe of the FSF’s ideals in
which I can say, “I strongly support this philosophy, but I’m not
going to go to fisticuffs with those with slightly different
viewpoints”.

It’s hard. :slight_smile: I want people to be able to avoid the ‘extreme’ stuff
if they want (such as strong copyleft), and embrace it if they choose
to, too. Saying all software must be copylefted or no software should
be copylefted is equally exclusive.

So, my decisions tend towards favoring choice. Right now in Ruby, we
have it, it’s just confusing as hell. As you and others have
mentioned, maybe a single permissive GPL compatible free software
license would be optimal.

cause to do through pressure or necessity, by physical, moral or
intellectual means". A licence can be coercion.

I’m an anarcho-capitalist. I guess that puts me in the dissenting
group on things like that.
I believe that strong contracts that are freely entered into can be
upheld and can be used to support various political and ethical goals.

This is easy to criticize, so I’ll take a shot at myself.

“You mean, if you misread the GPL and in it somewhere there is a
clause that all of your software are belong to santa clause, you’d
honor it?”

Begrudgingly, yes. It’s the individuals responsibility to understand
contracts they enter. If the ball has rolled in the GPLs court for a
long time because people didn’t know what they were getting into… oh
well.

One of the biggest problems in Free Software is not the FSF and their
views, and not those such as yourself who understand and oppose
various on various issues, but those who just slap the GPL on things
because Linux is GPLed.

That has caused a lot of murky water, and needs to be dealt with
somehow.
But that’s a story for a different group, I suppose.

I am mostly in favor of weak copyleft, in most pragmatic applications
of licenses. Protect your source, ensure it will remain free
software, ensure the license can’t be buried under a proprietary
license, etc. As far as linking and integrating with non-free
software… i’d like to be an idealist and say it’s evil… but of
course, I wouldn’t have a job if that were the case :wink:

Something like the MPL. The problem is, with the way that the GPL is
written, the MPL is incompatible with it. This will not be changing
under GPLv3.

I imagine I must be the FSF’s “most loving critic”. I do love these
guys and their ideas, but I don’t love the vagueness and rhetoric they
are prone to:

‘’‘This is a free software license which is not a strong copyleft;
unlike the X11 license, it has some complex restrictions that make it
incompatible with the GNU GPL. That is, a module covered by the GPL
and a module covered by the MPL cannot legally be linked together. We
urge you not to use the MPL for this reason.’‘’

Austin, can you explain what the hell they are talking about, if you
know? Otherwise, I’ll go digging around with them and find out.

I was not suggesting using the LGPL. I hate it just as much as
everyone. (including the FSF).

Mmmm. I used to recommend the LGPL. But I have also carefully read the
subtleties that Stallman put it in there that makes it a worse (more
restrictive) licence than the GPL itself.

I just don’t recommend it because the FSF retracted it’s suggestion to
use it. I mean… if the people who make a license say… oh, shit…
don’t use this, it doesn’t do much for a vote of confidence.

What are the specific clauses you find offensive though, out of
curiosity?

I was suggesting a sort of weak-copyleft, which does not restrict
non-free linking, but protects the freedom of the source itself. I am
not sure what that means exactly.

It means you prefer the MPL, by and large.

I’m interested to know what makes it incompatible.

But I might consider it if I ever deviate from the License of Ruby.
I might go to MPL 1.1 + GPLv3 for my code in the future.

18

On 4/5/06, Gregory B. [email protected] wrote:

On 4/5/06, Austin Z. [email protected] wrote:
[…]

necessity, by physical, moral or intellectual means". A licence can
be coercion.
I’m an anarcho-capitalist. I guess that puts me in the dissenting
group on things like that. I believe that strong contracts that are
freely entered into can be upheld and can be used to support various
political and ethical goals.

[…]

In theory, this stance is great. In practice, do you read every
licence you come across? Click-acceptance is legal these days, you know,
so there might be a licence clause that requires you to give your
firstborn the name “Myron Fiddlebrick”, and you might have agreed to
such without having read it.

I’m one of those odd folks who does read all of the fine print (on
contracts that I sign; I generally ignore the licence details on
software because it isn’t clear how enforceable it is and I prefer the
stated intent of the licence, e.g., “like a book”, etc.), and I’ll
challenge sections that I don’t like. If it matters enough to me, I
won’t do business with the person who puts out that sort of licence or
contracts.

Coercion can happen in many forms, and libreadline is a perfect example
of coercion. (Of course, I would argue that there are times when such
coercion is important and useful, such as when NeXT implemented an
Objective-C frontend for GCC and was forced to release its source.)

I don’t really have an issue with a copyleft licence: if you want to
restrict how your code can be given out (e.g., like the CC ShareAlike
licence provision), then that’s your choice. The problem I have with the
GNU GPL is the deception that goes on around it and the “moral”
boundaries it sets up around its restrictions. I also think, as I’ve
said, that the GPLv3 goes beyond what is reasonable for an open source
licence.

guys and their ideas, but I don’t love the vagueness and rhetoric they
are prone to:

Yeah. That’s Stallman through and through. I was one of the ~300 or so
folks heavily involved in the discussion list about the MPL back in the
day, and I got into some arguments with Eben Moglen about the whole
thing when the FSF was trying to get Netscape to solely use the GNU
GPL. (I consider the day that the Mozilla Foundation relicensed
everything as MPL and GNU GPL as a net loss for quality open source.)

The problem is entirely in the GNU GPL’s wording preventing further
restrictions on the code. Here’s a gedankenexperiment:

  • Create a copyleft licence patterned after the GNU GPL (make it the
    GNU GPL without the propaganda). Call it the ASCL (Austin’s Strong
    Copyleft Licence). Add one additional term that enforces your right
    to be known as the author of a work (this is not the advertising
    clause from the original BSD, mind you). The ASCL is incompatible
    with the GNU GPL.

The GNU GPL requires that your licence have a subset of restrictions
that the GNU GPL provides. It is not even clear if the GNU GPL would be
compatible the ASCL even if the additional term were removed. However,
even one additional restriction, no matter how reasonable, renders
your licence wholly incompatible with the GNU GPL.

It’s nonsense, and it could be fixed pretty easily.

-austin

19

I think you agree that the ‘v2 or later’ is not particularly a good
thing, although you say ‘GPL’ without specifying a version.

The problem (as I see it) is that commercial interests can bring to bear
huge legal (and PR) resources on a legal document that is not crystal
clear from the outset.

‘v2 or later’ does not seem to be very crystal clear.

Bob G

20

On 4/5/06, Austin Z. [email protected] wrote:

that TiVo should have to release their DRM code, even if I disagree
with their DRM and even if I disagree with the DMCA (which I do).
The Linux variant they use? Sure. But the DRM code? No.

If you want to modify the code, it comes back with a different
checksum, thus is useless on your Tivo.

I could claim that the teenage mutant ninja turtles live in my
toaster, but I mean, if I can’t see them or play with them in any
meaningful way, what’s the point of even telling you that?

Okay, so that was a terrible analogy.

The GPLv3 allows for userland DRM. It does not allow for external DRM.

Some of it comes down to the idea of IP ownership, which Stallman
clearly opposes. But I don’t. I think that the rights of ownership are
different with IP vs physical property (of necessity), but I think
that there are still important and viable rights that may even include
restricting who is allowed to use the IP involved. (I do disagree
with software and business method patents, even though I will probably
end up working on stuff that leads to a patent.)

Intellectual property is a confusing term. So are you talking about
copyright, trade secrets, patents, or what?

To put things in perspective, I barely believe in physical property.
And yes, this does come down to software patents, copyright, trade
secrets, trademarks, etc… but DRM with respect to GPLed code is
mostly about saying, if you need some key to make this work, you need
to give it to us.

You can’t give us a locked safe and promise it has a billion dollars in
it.

But I also believe in the idea of binary drivers while supporting the
development of open source drivers.

Yes. I use the ATI drivers on linux. I also use ndiswrapper to use
windows drivers for my wireless. Still haven’t found a good (and
convenient) free alternative. But if the drivers were free in the
first place…

But that’s nothing to do with DRM.