Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 25 Mar 2002 03:08:40 -0800
From:      Terry Lambert <tlambert2@mindspring.com>
To:        Hans Reiser <reiser@namesys.com>
Cc:        hiten@uk.FreeBSD.org, Chris Mason <mason@suse.com>, Josh MacDonald <jmacd@CS.Berkeley.EDU>, Parity Error <bootup@mail.ru>, freebsd-fs@FreeBSD.ORG, reiserfs-dev@namesys.com
Subject:   Re: [reiserfs-dev] Re: metadata update durability ordering/soft updates
Message-ID:  <3C9F0538.B3D00EE2@mindspring.com>
References:  <20020317225759.82774.qmail@web21109.mail.yahoo.com> <3C95ACBA.4040108@namesys.com> <3C95B838.F8ECE39A@mindspring.com> <3C95C8C3.7080803@namesys.com> <3C966CDF.25A7A379@mindspring.com> <3C9E1D6E.3080604@namesys.com> <3C9E6BEC.B2EB8D86@mindspring.com> <3C9F030A.7040205@namesys.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Hans Reiser wrote:
> So, if a vendor (not being hypothetical here) bases a cluster filesystem
> on reiserfs, is reiserfs central to the resulting work?
> 
> I think so.  I also think that linking or not linking is not determinative.

It depends on what you mean by "bases".

If you mean they take the ReiserFS source code and modify it
to be "ReiserFS + clustering code that works only with Reiserfs
becuse it shares code with it", then the answer is yes.

If you mean they take a bunch of systems using ReiserFS as their
filesystem, and cluster the machines using clustering software
that can work with any FS, and it's only coincidental to the
deployment that "the systems which are members of the cluseter
use ReiserFS", then the answer is "no".


If the answer is "yes", then the code is a derivative work.  Now
do they have to give out their source code?

The answer is "not unless they copy and distribute the code".
They only have to give out their source code (or an offer of it,
valid for 3 years), if they sell, give away, or otherwise
distribute the binary software for their clustering software.

If they merely sell the use of the cluster (e.g. "Rent web hosting
from ``Joe's web hosting'', on our nifty cluster"), then the answer
is "no", because they did not "distribute or publish" the code;
thus they are not regulated by section 2(b) of the GPL.

Now say IBM buys "Joe's web hosting".  Then they are distributing
the code to IBM.  The easy way to deal with this is to make IBM
relink the clustering code after the purchase, and not transfer
binary copies of the combined program.  Thus the code is, again,
not distributed, and the non-GPL'ed code linked with it doesn't
fall under the GPL.

Clause 2(b) clearly limits application of the license to programs
which you copy and distribute.  Similarly, 3(a), 3(b), and 3(c)
also require copying and distribution.

Section 3(c) is the most interesting.  If you are given GPL'ed
source code to a kernel component, and binary code that can be
relinked with any coponent, including the GPL'ed kernel component,
when you as an end user do the linking, then the resulting binary
can not be redistributed, since you, as the user, do not have the
source code to the binary portion, and therefore can not comply
with section 3 by making available or offering to make available,
or repeating a non-existant offer to make available, the source
code to the binary portion of the program.

Thus you can comply with the letter of the license, but not
its spirit.

There are at least 6 "direct" loopholes in the license, and
another 4 "indirect" loopholes, which I'm aware of to date.
The indirect loopholes are questionable, in that a court might
not hold their application to be legal (intentional entry into
a contract with the intent of non-performance).

The license isn't very good at what it was supposedly designed
to do...

Actually, the Cygnus eCOS license is a much better instrumentality
of "The GNU Manifesto" than either the GPL or the LGPL.  If you
are trying to support "The GNU Manifesto" through your choice of
license, you are much, much better off using the eCOS license.


-- Terry

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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3C9F0538.B3D00EE2>