Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 9 Jan 1999 11:38:43 -0500 (EST)
From:      "Matthew N. Dodd" <winter@jurai.net>
To:        "Robert V. Baron" <rvb@cs.cmu.edu>
Cc:        cvs-committers@FreeBSD.ORG
Subject:   Re: Coda license term changes ... GPL
Message-ID:  <Pine.BSF.4.02.9901091137040.8926-100000@sasami.jurai.net>
In-Reply-To: <yzs7lux1lyc.fsf@sicily.odyssey.cs.cmu.edu>

next in thread | previous in thread | raw e-mail | index | archive | help

Wow, yet another 'Open Source' project switching to a more restrictive
license.

Why don't they ask the current Debian-core what they really think about
the GPL?

On 8 Jan 1999, Robert V. Baron wrote:
> The first piece of mail below was posted to coda-announce earlier in
> the week.  The netbsd folk (on developers) have been having mixed
> feeling over it.  A clarification was posted to (developers) which is
> also included below.  I was wondering how the FreeBSD folk felt.
> Note: a number of people have taken offense over the OSS usage in
> the first post.  This is clarified in the second post.
> 
> ================================================================
> To: coda-announce@TELEMANN.coda.cs.cmu.edu,
>         Coda Discussion <linux-coda@TELEMANN.coda.cs.cmu.edu>
> 
> Happy 1999!
> 
> I have just uploaded the first 5.0 pre release.  
> 
> There are many differences with 4.6:
> - an entirely new directory system which removes a really nasty bug that
> plagued the servers for years.  
> 
> - many improvements to the networking code, in particular, much more
> stable computations of bandwidth and retransmission times.
> 
> - Coda is now GPL'd - primarily because we want to indicate that we are
> really an OSS project.
> 
> This is a good moment for maintainers of Debian, Sparc Linux and other
> platforms to jump forward.  
> 
> We will smooth out some remaining bugs during the next few weeks.  We will
> come out with a Windows 95 client next week and an NT server soon, as well
> as with the NetBSD/FreeBSD releases.  Windows 95 is hard to build right
> now (wait until next week) or read the coda-HOWTO, for BSD you should be
> ok.
> 
> A couple of other points:
> 
> - a new Coda HOWTO was written.  Have a look at:
> 	http://www.coda.cs.cmu.edu/coda-howto.html
>   source in 
>         ftp://ftp.coda.cs.cmu.edu/pub/coda/doc/coda-howto.sgml
> 
> - We have a bugs list (Jitterbug): please read them and fix them!
> 
> - I have made a list of some projects that are falling between the cracks
> here.
>        	http://www.coda.cs.cmu.edu/todo.html
> 
> 
> Have fun, let us know what is good (and bad). 
> 
> - Peter -
> =================================================================
> Let me try clarify the dicussions about Coda and its recent adoption
> of the GPL.
> 
> 1. HISTORY
> 
> First, I have been a BSD follower since 4.2BSD.  I worked on the
> porting of Mach to various platforms.  Perhaps most signficantly I did
> the Mach i386 distribution and supported the Mach 2.5 kernel and later
> Mach 3.0 kernel.  I have most recently worked on kernel support for
> Coda on x86 for NetBSD and FreeBSD, though the original code was
> written by others.
> 
> I am low on the Coda organizational totem pole.  Peter Braam (who made
> the original post) is a senior systems scientist who leads Coda
> development.  He reports to the Coda PI, Mahadev Satyanarayanan (Satya
> for short).  Satya is responsible for making the GPL decision and did
> so only after careful thought.  He has advised many students whose
> theses make up Coda.  Numerous people (~20) have worked on Coda in the
> past.
> 
> When Coda was proposed to be included in the NetBSD & FreeBSD systems
> and allowed in, there was no plan at that time that to GPL Coda.  So I
> was acting in good faith.  Satya made the decision to GPL Coda only in
> December 1998.
> 
> 2. RATIONALE FOR DECISION
> 
> Satya has pondered how Coda will survive when contract funding for it
> ceases.  He also has considered who, if anyone, should profit from it
> when all the people who contributed to it at the university have
> graduated and gone their separate ways. Next, there is the issue of
> Coda source fragmentation.  Satya believes that GPL is more likely to
> ensure that Coda remains coherent, preserving a single copy of Coda
> with the same API for every user/builder/supporter.  Finally, Satya
> also believes that GPL reassures potential contributors that their
> work will forever remain publicly available, and not become the
> captive of some company that derives sole benefit from it.  These were
> the reasons for his deciding to GPL Coda.
> 
> Note that components of Coda that have independent use (typically
> libraries such as RVM and RPC2) are under the less restrictive LGPL.
> Most importantly, KERNEL CODE IN NetBSD AND FreeBSD IS NOT GPL'ED.
> They preserve the original copyright from CMU under which they were
> integrated into NetBSD and FreeBSD.  We have no plans to change this.
> However, if it will make the NetBSD and FreeBSD communities more
> comfortable, WE ARE WILLING TO TRANSFER OWNERSHIP OF THIS CODE TO
> THEM.
> 
> 3. WHERE TO GO FROM HERE
> 
> Let's let the dicussions proceed for a while and I will wait for an
> answer from core.  I think that GPL is a fact of life.  Our
> compilation tools are GPL'ed, visit /usr/src/gnu.  I don't accept that
> we have a system without the comp tools.  Further, I would find it
> difficult to survive w/o some flavor of emacs which is GPL/ed, etc ...
> I don't think that it should matter that the Coda application code is
> GPL'ed.
> 
> In a past discussion, it was maintained that if Coda was in the kernel
> it should be in the source tree.  It was resolved that it was
> unfeasable to split Coda apart and populate the source tree with the
> directories and still maintain synchronization with Coda development.
> Perry suggested that Coda be maintained in a shadow tree like gnu is.
> This was defered for a while because we were not ready.  Perry's plan
> is sensible and I will follow up on it if Coda remains part of NetBSD
> and FreeBSD.
> 
> Finally, you could just yank Coda out of the kernel and make it an lkm
> only.  This would make life more difficult for me, since lkms are
> currently difficult to debug in NetBSD.  However, I am willing to do
> this if that's the preferred outcome.
> 
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe cvs-all" in the body of the message
> 

-- 
| Matthew N. Dodd  | 78 280Z | 75 164E | 84 245DL | FreeBSD/NetBSD/Sprite/VMS |
| winter@jurai.net |      This Space For Rent     | ix86,sparc,m68k,pmax,vax  |
| http://www.jurai.net/~winter | Are you k-rad elite enough for my webpage?   |


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



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.02.9901091137040.8926-100000>