[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: license check for tqsllib

[ Replying to both at once ]

On Wed, Oct 22, 2003 at 05:54:24AM +0100, Andrew Suffield wrote:
> On Tue, Oct 21, 2003 at 02:07:06PM -0500, Branden Robinson wrote:
> > On Mon, Oct 20, 2003 at 10:16:09AM -0600, Joel Baker wrote:
> > > I don't *like* it (to the point that I'm willing to spend a large
> > > number of my own hours, working to convince people to switch for a
> > > 3-clause variety), but there is enough software in the world that
> > > falls under it, and enough room to manuever in terms of whether it's
> > > even enforceable (or more than GPL-incompatible in the most technical
> > > of senses, and compatible in RMS's declaration of intent) that I'd
> > > really hate to see us drop everything that uses it, much as I'd hate
> > > to see us drop everything that could potentially have a patent on it
> > > - for much the same reason.
> >
> > How about, once the 4.1.5 vote is done, we try to amend the DFSG to say
> > that no future software using the advertising clause will be permitted
> > into Debian?

Could be problematic; thing A is OK, but thing B is not OK, when they have
the exact same license terms, based solely on copyright? Which copyright do
we use, the first or last occurance?

Discouraging it but permitting it (as 4 does for others), on the other
hand, seems to be the current status quo, and is (IMO) perfectly
reasonable, even for enshrining in the text of the DFSG.

> How about we find any software that still has the advertising clause and
> aim to chuck it out of the archive? It can't be that difficult by now.

Ah, thank you, Mr. Suffield, for the typical extremism. I reply, as I did
in the first message: "How about we chunk out any software that still has
potential patent problems?"

It has been opinioned that the advertising clause is probably unenforceable
in most jurisdictions; it is not actually intended to be incompatible from
the GPL's point of view; and while it is horrific and ugly, we seem to have
a fairly high success rate at getting it removed when we ask upstreams to
consider doing so.

I've put about 40 or 50 hours (actual ones, not billable ones) into
contacting copyright holders in the NetBSD source tree, asking them to
consider relicensing, and preparing patches to fix the license texts, as
well as dealing with the gatekeepers of their CVS repository.

To date, only 1 author has refused (Theo de Raadt, who claims it isn't
necessary, since the code is available under a 2-clause BSD license via
OpenBSD's source).

If you want to go on a hunt through the archive and convince upstreams to
change it, be my guest; however, we can't even decide to yank the GFDL
stuff out of the archive, yet, when it's clear that relicensing is not
likely to happen for most/all GNU-upstream documents, and the license has
many known, practical problems.

Why on earth does it seem reasonable to yank 4-clause BSD stuff when we
have an extremely high success rate at converting it, and the bulk of the
work (inspecting licenses) is the same, with a fairly small incremental
(contacting upstream)?

Of course, much like patents, if anyone actually enforces the clause in a
way that seems likely to bring us problems, yank the stuff out; no argument
there. Can you document any case of someone using the advertising clause in
relation to any software currently in main or contrib? (Since the 'chuck
it' would probably translate into 'move it into non-free').
Joel Baker <fenton@debian.org>                                        ,''`.
Debian GNU NetBSD/i386 porter                                        : :' :
                                                                     `. `'

Attachment: pgp78LcZITV5b.pgp
Description: PGP signature

Reply to: