legal status of progmodes/sym-comp.el

legal status of progmodes/sym-comp.el

[Glenn Morris]

It's can of worms time again...

Dave Love's latest version of python.el requires his sym-comp library.
This is now installed as lisp/progmodes/sym-comp.el.
It is "Copyright (C) 2004  Dave Love".

As far as I know, he has expressed no interest in contributing this to
Emacs, so I don't see that we can just take it and make it copyright FSF.

I mean, I have a blanket past and future Emacs copyright assignment,
but that doesn't mean you can take any elisp code I happen to put on a
website and make it copyright FSF - only those things I explicitly
contribute.

(We seem to be interpreting the lawyer's statement:

   http://lists.gnu.org/archive/html/emacs-devel/2007-05/msg00466.html
   From:    Richard Stallman
   Subject:    We can use python.el
   Date:       Sat, 12 May 2007 12:47:54 -0400

   Our lawyer says we can use python.el, including the more recent changes.

to mean that any changes Dave makes to python.el on his website can be
incorporated into Emacs for the rest of time, even if he doesn't
explicitly say so. That seems odd to me as well.)

Re: legal status of progmodes/sym-comp.el

[Chong Yidong]

Glenn Morris <rgm@gnu.org> writes:

> It's can of worms time again...
>
> Dave Love's latest version of python.el requires his sym-comp library.
> This is now installed as lisp/progmodes/sym-comp.el.
> It is "Copyright (C) 2004  Dave Love".
>
> As far as I know, he has expressed no interest in contributing this to
> Emacs, so I don't see that we can just take it and make it copyright FSF.
>
> I mean, I have a blanket past and future Emacs copyright assignment,
> but that doesn't mean you can take any elisp code I happen to put on a
> website and make it copyright FSF - only those things I explicitly
> contribute.
>
> (We seem to be interpreting the lawyer's statement:
>
>    http://lists.gnu.org/archive/html/emacs-devel/2007-05/msg00466.html
>    From:    Richard Stallman
>    Subject:    We can use python.el
>    Date:       Sat, 12 May 2007 12:47:54 -0400
>
>    Our lawyer says we can use python.el, including the more recent changes.
>
> to mean that any changes Dave makes to python.el on his website can be
> incorporated into Emacs for the rest of time, even if he doesn't
> explicitly say so. That seems odd to me as well.)

Thanks for the catch.  I'm forwarding this to Dave Love as well.

Re: legal status of progmodes/sym-comp.el

[Dave Love]

Chong Yidong <cyd@stupidchicken.com> writes:

>> Dave Love's latest version of python.el requires his sym-comp library.
>> This is now installed as lisp/progmodes/sym-comp.el.
>> It is "Copyright (C) 2004  Dave Love".

That's obviously an error -- since it contains code from lisp.el --
probably due to Emacs regressions breaking my context-dependent
settings, which I didn't spot in that case.  I'll fix the notice when I
get a chance.

>> As far as I know, he has expressed no interest in contributing this to
>> Emacs, so I don't see that we can just take it and make it copyright
>> FSF.

Obviously it was meant to be in Emacs.  It would have been, with other
abstractions, if I'd got round to it when I was a maintainer.  However
it was rejected previously and expurgated from python.el.  There's no
point in using it just for Python.

>> I mean, I have a blanket past and future Emacs copyright assignment,
>> but that doesn't mean you can take any elisp code I happen to put on a
>> website and make it copyright FSF - only those things I explicitly
>> contribute.

Indeed, in general, but that was a mistake.  You can use the sym-comp
code if it's now acceptable for some reason.

It would be better if contributions weren't rejected and then either
used or reimplemented by someone else years later, or broken randomly,
or just left to linger in CVS for years without even accepting bug
reports/fixes from people who wrote them.

Re: legal status of progmodes/sym-comp.el

[Chong Yidong]

Dave Love <fx@gnu.org> writes:

>>> Dave Love's latest version of python.el requires his sym-comp library.
>>> This is now installed as lisp/progmodes/sym-comp.el.
>>> It is "Copyright (C) 2004  Dave Love".
>
> That's obviously an error -- since it contains code from lisp.el --
> probably due to Emacs regressions breaking my context-dependent
> settings, which I didn't spot in that case.  I'll fix the notice when I
> get a chance.
>
> Obviously it was meant to be in Emacs.  It would have been, with other
> abstractions, if I'd got round to it when I was a maintainer.  However
> it was rejected previously and expurgated from python.el.  There's no
> point in using it just for Python.
>
>>> I mean, I have a blanket past and future Emacs copyright assignment,
>>> but that doesn't mean you can take any elisp code I happen to put on a
>>> website and make it copyright FSF - only those things I explicitly
>>> contribute.
>
> Indeed, in general, but that was a mistake.  You can use the sym-comp
> code if it's now acceptable for some reason.

Okay.  Thanks for the clarification!

Re: legal status of progmodes/sym-comp.el

[Dave Love]

Chong Yidong <cyd@stupidchicken.com> writes:

>> Indeed, in general, but that was a mistake.  You can use the sym-comp
>> code if it's now acceptable for some reason.
>
> Okay.  Thanks for the clarification!

I should have said that one reason you can't -- or, at least, couldn't
-- just take even assigned code into Emacs is that it may not be covered
by employer disclaimer, if that still makes any difference.  (I don't
understand the lawyer's comments on the issues that worried me.)  That
code is old enough not to be a problem, though.

However, there's no point unless it's merged back into lisp.el.  If you
read the commentary, the whole point is to provide one of the missing
abstractions over major mode features.

Re: legal status of progmodes/sym-comp.el

[Glenn Morris]

Dave Love wrote:

> It would be better if contributions weren't rejected and then either
> used or reimplemented by someone else years later, or broken randomly,
> or just left to linger in CVS for years without even accepting bug
> reports/fixes from people who wrote them.

rms actively wants your python.el fixes. I know of no reason why you
shouldn't just be maintaining python.el in CVS yourself, unless you
don't want to.

http://lists.gnu.org/archive/html/emacs-devel/2008-02/msg00841.html

    From:   Richard Stallman
    Subject:        python.el fixes for Emacs 22
    Date:           Mon, 11 Feb 2008 08:34:49 -0500

    I asked Dave to send a diff and a change log, as a good
    contributor should, but he did not respond. I think that the
    changes are important enough that we should do the extra work to
    install them anyway.

Re: legal status of progmodes/sym-comp.el

[Dave Love]

Glenn Morris <rgm@gnu.org> writes:

> rms actively wants your python.el fixes.

Perhaps you misunderstand.  I haven't made fixes to the version in Emacs
(which I haven't tried to understand), just maintained the original
code.  It at least doesn't have the bad changes, such as apparently
breaking inferior execution in a fundamental way.  I merely fixed it
recently to account for the last regressions I could find from Emacs 22,
and synched it with the one I use in Emacs 21.

Using the differences between the versions is equivalent to copying
mine.

> I know of no reason why you
> shouldn't just be maintaining python.el in CVS yourself, unless you
> don't want to.

For a start, my CVS access was chopped some time ago.

Re: legal status of progmodes/sym-comp.el

[Stefan Monnier]

> For a start, my CVS access was chopped some time ago.

This is simply not the case.

http://savannah.gnu.org/project/memberlist.php?group=emacs shows
unambiguously that you're considered as a member of the project, and
that gives you access rights to the CVS repository.

I guess your CVS access is simply *not working* and you just need to
figure out why,


        Stefan