Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 14 Dec 2001 12:28:37 +1030
From:      Greg Lehey <grog@FreeBSD.org>
To:        hiten@uk.FreeBSD.org, Terry Lambert <tlambert2@mindspring.com>, "Brandon D. Valentine" <bandix@looksharp.net>, "Gary W. Swearingen" <swear@blarg.net>, Hiten Pandya <hitmaster2k@yahoo.com>
Cc:        chat@FreeBSD.org, phk@FreeBSD.org, Poul-Henning Kamp <phk@critter.freebsd.dk>
Subject:   IBM's intentions with JFS (was: IBM suing (was: RMS Suing was [SUGGESTION] - JFS for FreeBSD))
Message-ID:  <20011214122837.O3448@monorchid.lemis.com>
In-Reply-To: <20011213044944.L56723-100000@turtle.looksharp.net> <k6y9k6aa4n.9k6@localhost.localdomain> <3C193048.4FBDB302@mindspring.com> <quhequdhaf.equ@localhost.localdomain> <3C187D1F.24D8E4D2@mindspring.com> <20011213093555.76629.qmail@web21107.mail.yahoo.com> <1id71idej9.71i@localhost.localdomain> <20011213051012.Y56723-100000@turtle.looksharp.net> <3C1875D6.5DE4F996@mindspring.com> <3C186381.6AB07090@yahoo.com>
References:  <3C186EA5.4EA87656@mindspring.com> <20011213093555.76629.qmail@web21107.mail.yahoo.com> <3C186381.6AB07090@yahoo.com> <3C1875D6.5DE4F996@mindspring.com> <1id71idej9.71i@localhost.localdomain> <3C1875D6.5DE4F996@mindspring.com> <20011213051012.Y56723-100000@turtle.looksharp.net> <3C186381.6AB07090@yahoo.com> <3C1875D6.5DE4F996@mindspring.com> <3C186381.6AB07090@yahoo.com>

next in thread | previous in thread | raw e-mail | index | archive | help
There seem to be a lot of misconceptions going on about IBM's
intentions with regard to the GPL release of JFS2.  In particular, I
think that Terry is wide of the mark, despite having worked with IBM.
I don't speak officially for IBM (I'm not paid enough :-), but I can
confidently state that:

1.  IBM has released JFS2 as open source.  You can do whatever the
    license allows with it, including unrestricted commercial use.

2.  IBM chose the GPL rather than the BSD license because the BSD
    license would allow competitors to take the code and use it as the
    basis of one of *their* commercial, closed-source solutions.  IBM
    does not want this.

3.  There are no other restrictions on its use.

4.  It is possible to port JFS2 to FreeBSD without violating any
    license, either via loophole or otherwise.

    In this connection, I would like to point out an issue with the
    LGPL which I personally think is sailing close to the wind: I
    can't see any real license distinction between linking GPL'd code
    into the kernel and loading the module either during the boot or
    afterwards.  In the first case, the code is statically linked, in
    the latter it's dynamically linked.  In each case, the result is
    the same.  That's not important, though, because RMS does think
    there's a difference (or he's prepared to pretend there's one, and
    he's prepared to sanction use in this form).

I'll address some of the points which have been raised:

On Thursday, 13 December 2001 at  8:14:57 +0000, Hiten Pandya wrote:
> The real way to solve this problem is:
>
> IBM has release JFS code under the 'General Public License'. What
> we have to do is ask this question internally to IBM, which Greg
> might do that for us, if possible; and then we see what the JFS
> Team from IBM's views are about this.

I've done this.  As I expected, they're all in favour of it.

> example: A project called UFS2, which would be our target for
> improving current UFS code.

You'll be chasing the experts on this one.  I'd personally prefer to
see a JFS port.

> If we get positive results by September 2002, that JFS code has been
> ported in its entirety without affecting FreeBSD or the Licensing
> terms in any manner of way, than we can possibly, merge it (under
> src/gnu or something) to FreeBSD.

It's nice to plan your time, but why specifically September 2002?

> In a nutshell: We should ask IBM JFS (Core) Team, and see what they
> say; We Start the Project || UFS2.

I think the real issue is getting enough manpower to do the work.  As
I've said before, IBM will help, not hinder.

On Thursday, 13 December 2001 at  1:33:10 -0800, Terry Lambert wrote:
>
> There are several ways to do this with a JFS port; the last five
> skate by on technicalities, while the first simply disregards the
> license entirely:
>
> 4)	Build boot code that can at least read JFS, and load the
> 	real JFS as a kernel module.
>
> 	o	Since you have to deal with all the lookup and
> 		log version selection issues, this is more than
> 		half way to a reimplementation without the GPL.

No, I don't believe this to be the case.  I'm currently looking at
reading JFS1 file systems, and for read-only access I can completely
ignore the journal (assuming the file system is clean).

> 		This is legally risky, if Greg is right, and the
> 		intent was to prevent commercial use of the code,
> 		since you defeat their intent.

No, you misunderstand.  See above: the intention was to prevent
competitors incorporating IBM code in their proprietary products.  You
can't prevent commercial use of GPL software.

On Thursday, 13 December 2001 at  6:06:06 -0500, Brandon D. Valentine wrote:
> On Thu, 13 Dec 2001, Terry Lambert wrote:
>
>> As I said before: feel free to write the code.  Just don't expect
>> people who, philosophically, want the FreeBSD code to be usable
>> (and used!) as a source of reference implementations to participate
>> in this process.
>
> That fact that this thread had to even exist is unforunate.  We won't go
> into the GPL is evil aspect, since it's been pretty well covered
> elsewhere on the web by various people.  I would mention that for all of
> the reasons mentioned herein porting JFS is, as should be obvious by
> now, a very risky project straddling ugly licensing issues.

I don't see any ugly licensing issues.

> I also don't know what makes JFS a better candidate for porting than
> any of the other GPL'd filesystems.  Those who have significant
> involvement with the Linux community will note that the Linux port
> of JFS doesn't get nearly as much publicity or high profile use as
> the Linux port of XFS.

Correct.  One of the intentions of the JFS project is to get in the
Linux source tree.  They believe that their approach, which has made
fewer contentious changes to the kernel than XFS has.  This doesn't
have to worry the FreeBSD project, of course.

On Thursday, 13 December 2001 at 12:57:14 -0800, Gary W. Swearingen wrote:
> Terry Lambert <tlambert2@mindspring.com> writes:
>
>> 5)	Use an alternate boot partition of a different FS type
>> 	(e.g. FAT32 or FFS), and load the JFS module from there.
>
>> 	o	This is harder than option #3, since it requires
>> 		minimally that union mounts work, since mounts
>> 		over mount points before the kernel is booted
>> 		can't work, and therefore for your / to be JFS,
>> 		it will have to be mounted over top of the boot /,
>> 		which would be non-JFS.
>
> Maybe the JFS4FBSD developer could replace the boot loader with one
> which works rather like LILO and gets the kernel, the JFS module,
> and whatever else it needs from a pre-prepared non-FS image which
> the loader can read without FS support.  Even harder to argue that
> this is mere common storage and not a linked program.

The loader might have to be GPL'd.

On Thursday, 13 December 2001 at 11:57:44 -0800, Gary W. Swearingen wrote:
> Terry Lambert <tlambert2@mindspring.com> writes:
>
>> someone create FS images for you to test with, all without
>> looking at a line of code that would render you "contaminated".
>
> I wonder about that after having seen it said several times.
>
> I'm quite sure (without explict evidence) that if this was taken to
> court that an accused infringer would be determined to be contaminated
> by any open source code, upon the presumption that he had looked at the
> code, there being no practical means to discover the truth and the ease
> of hiding it.

"Contamination" was a term thrown around during the USL wars of the
early 90s.  I even have a USENIX pin somewhere saying "mentally
contaminated".  Cases like this have been tried in court in Europe:
people who worked for one company were required to sign contracts
which prohibited them from ever working for any of their competitors,
effectively requiring them to discard their expertise and take up some
other line of work.  The contracts were found to be in breach of law.
I don't know if US law would rule any differently.

Greg
--
See complete headers for address and phone numbers

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?20011214122837.O3448>