Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 01 Dec 2000 19:22:17 -0700
From:      Brett Glass <brett@lariat.org>
To:        Terry Lambert <tlambert@primenet.com>, rsidd@physics.iisc.ernet.in (Rahul Siddharthan)
Cc:        tlambert@primenet.com (Terry Lambert), freebsd-chat@FreeBSD.ORG
Subject:   Re: Here is what IBM thinks about using FreeBSD on their newer
Message-ID:  <4.3.2.7.2.20001201190548.044fa3b0@localhost>
In-Reply-To: <200012011219.FAA23538@usr01.primenet.com>
References:  <20001201114509.B61418@lpt.ens.fr>

next in thread | previous in thread | raw e-mail | index | archive | help
At 05:19 AM 12/1/2000, Terry Lambert wrote:

>Actually, the ASP scenario was exactly how I'd interpreted
>Brett's phrase "performance for profit".  I just don't think
>the model for doing that is going to be successful.

The phrase isn't mine. It's a standard part of music copyright
law, and Perens was the first to suggest that GPL activists
attempt to apply it to software.

>I'll agree that Brett ratholed into an adjacent topic, though.

I don't think it was "ratholing" at all. IBM's behavior is
important, and as you have already mentioned, it is influenced
by concerns about the GPL.

>His concern is to get people who would not otherwise use the
>GPL, to use the GPL.

And then to "trap" companies with it. Stallman specifically
advocates this in his essay "Why Software Should Not Have
Owners" -- an essay which, by the way, he hastily toned down
a few days after I pointed out his original exhortation to
programmers to sabotage their organizations.

>The main thrust of his point is scripting languages, but it
>appears to me to be "the camel's nose", since he doesn't
>limit the performance to scripting languages.  He presumes
>that the payback of having access to modified (he seems to
>assume tha this equals "improved", rather than "trade dress")
>is enough to pay back the original company for releasing the
>code that is not currently being released, under the GPL.
>
>He may have a point on scripts.  Scripts are generally in
>the category "throw away code" (the same place I choose to
>put "fetchmail"), and so cost relatively little to create.
>If the creation cost is very low, then the amount one needs
>to benefit from the code in order to amortize developement
>costs is also very low, and so it could be that the value
>they get back would easily exceed the value they lose by
>releasing the code.

The problem is that some programs written in "scripting
languages" are not throw-aways at all. Many companies who
add value to arcane systems by providing them with user-
friendly GUIs (of which Whistle was one!) do it via these
languages. I've worked for companies whose entire base of
IP (and expertise!) was embodied in Perl scripts.

>If he ties in performance in a general sense, though, he will
>poison-pill the code: code that elects the license (even if
>the clause is at the authors discretion) will prevent legal
>use of GPL'ed code that must be "performed" in binary, by
>linking against OS libraries, since the requirement becomes
>providing all necessary code, as source, that is needed to
>repeat the performance.

It could be worse than that. Since FreeBSD is compiled by GCC,
he could set things up so that that any binary of FreeBSD that's
compiled by a GPL3ed version of GCC is considered to be a
"performance" of a GPLed program, with all of the effects that
this might entail. He has a strong incentive to do this, since 
it would throw alternatives to Linux -- including BeOS! -- into 
turmoil. I personally believe that Stallman, Perens, and
others WILL attempt to close this noose; it's just a matter of 
WHEN they'll do it. 

>His ideology may eventually win (IMO, to the detriment of us
>all), but I don't think that he is going to be able to force
>the issue this way; it is more likely he will slit his own
>throat with the attempt.  Of course, this was always a
>danger of the "or later version of the license".

What's more, all GNU utilities will immediately go to the new
license. This poses a big danger for FreeBSD, because it
tracks these utilities. Unless it freezes these utilities
at the last version that was released under the previous
license, and reimplements everything fast, it will immediately
be "infected" by the new requirements in a detrimental way.

>I also see it as being problematic for things like Linux,
>which unlike the FSF tools, accept contributions without
>having to have the rights granted to a single legal
>entity.  The problem with that has always been that any
>author could claim version differences for their code
>contributed to the project.  Having a trap-door clause
>that lets any author do the same with a performance
>clause will, I predict, open a can of worms that could
>kill the GPL for good.

In theory, yes. But in practice, the FSF will own so much
code that's vital to projects such as FreeBSD that, protest
as they will, they'll be between a rock and a hard place. I
believe that by adopting any GPLed code at all, FreeBSD is
painting itself into a corner. YMMV, of course.

--Brett



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-chat" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4.3.2.7.2.20001201190548.044fa3b0>