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

Bug#186116: marked as done (ITP: ParMETIS: distributed graph partitioning and sparse matrix reordering)



Your message dated Sun, 13 Apr 2003 11:24:31 -0400
with message-id <E194jLL-00035H-00@auric.debian.org>
and subject line Bug#186116: fixed in parmetis 3.0-2
has caused the attached Bug report to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what I am
talking about this indicates a serious mail system misconfiguration
somewhere.  Please contact me immediately.)

Debian bug tracking system administrator
(administrator, Debian Bugs database)

--------------------------------------
Received: (at submit) by bugs.debian.org; 24 Mar 2003 16:43:42 +0000
>From hazelsct@mit.edu Mon Mar 24 10:43:41 2003
Return-path: <hazelsct@mit.edu>
Received: from ozymandias.mit.edu [18.53.0.125] 
	by master.debian.org with esmtp (Exim 3.12 1 (Debian))
	id 18xV2x-0003Ex-00; Mon, 24 Mar 2003 10:43:39 -0600
Received: from localhost ([127.0.0.1] helo=mit.edu)
	by ozymandias.mit.edu with esmtp (Exim 3.35 #1 (Debian))
	id 18xV2w-0008Cc-00; Mon, 24 Mar 2003 11:43:38 -0500
Message-ID: <3E7F35B8.2090208@mit.edu>
Date: Mon, 24 Mar 2003 11:43:36 -0500
From: Adam C Powell IV <hazelsct@mit.edu>
Organization: Massachvsetts Institvte of Technology
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.2.1) Gecko/20021220 Debian/1.2.1-3
X-Accept-Language: en
MIME-Version: 1.0
To: Debian Bug Tracking System <submit@bugs.debian.org>
CC: debian-devel@lists.debian.org
Subject: ITP: ParMETIS: distributed graph partitioning and sparse matrix reordering
Content-Type: multipart/mixed;
 boundary="------------090603000609020005070704"
Delivered-To: submit@bugs.debian.org
X-Spam-Status: No, hits=-1.7 required=4.0
	tests=HAS_PACKAGE,NOSPAM_INC,SPAM_PHRASE_02_03,USER_AGENT,
	      USER_AGENT_MOZILLA_UA,X_ACCEPT_LANG
	version=2.44
X-Spam-Level: 

This is a multi-part message in MIME format.
--------------090603000609020005070704
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Package: wnpp
Severity: wishlist

I intend to package ParMETIS, which is a tool for optimal partitioning 
of graphs and meshes and reorder sparse matrices for to minimize fill 
when using direct solvers.  This tool is primarily used for load 
balancing and communication optimization in finite element simulations 
on parallel computers, including Beowulf clusters.  It uses MPI for 
communication, and should work just fine on Debian beowulfs.

Unfortunately, as you can see below, the license is very non-free, so I 
obtained permission from the author (attached) to distribute it with 
Debian.  Also unfortunately, if I link PETSc to ParMETIS, that will 
throw it into contrib; or I can maintain two versions: petsc in main 
with no parmetis dependency, and petsc-contrib in contrib with it.  I 
don't like this, and have encouraged upstream to dual-license or 
otherwise change distribution and use terms in a way which will allow us 
to put it in main and protect his financial interests.  But no bite yet. :-(

Upstream COPYRIGHT file:

Copyright Notice
----------------

The ParMETIS/METIS package is copyrighted by the Regents of the 
University of Minnesota. It can be freely used for educational and 
research purposes by non-profit institutions and US government agencies 
only.  Other organizations are allowed to use ParMETIS/METIS only for 
evaluation purposes, and any further uses will require prior approval. 
The software may not be sold or redistributed without prior approval. 
One may make copies of the software for their use provided that the 
copies, are not sold or distributed, are used under the same terms 
and conditions.

As unestablished research software, this code is provided on an
``as is'' basis without warranty of any kind, either expressed or
implied. The downloading, or executing any part of this software
constitutes an implicit agreement to these terms. These terms and
conditions are subject to change at any time without prior notice.


If there are no objections, I'll upload within a week or so.

[Please CC me on replies to -devel, and of course, do not CC 
submit@bugs.debian.org.]
-- 

-Adam P.

GPG fingerprint: D54D 1AEE B11C CE9B A02B  C5DD 526F 01E8 564E E4B6

Welcome to the best software in the world today cafe! 
<http://lyre.mit.edu/%7Epowell/The_Best_Stuff_In_The_World_Today_Cafe.ogg>


--------------090603000609020005070704
Content-Type: message/rfc822;
 name="message.eml"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="message.eml"

Envelope-to: hazelsct@lyre.mit.edu
Received: from fort-point-station.mit.edu ([18.7.7.76])
	by lyre.mit.edu with esmtp (Exim 3.35 #1 (Debian))
	id 18xTv0-0006OM-00
	for <hazelsct@lyre.mit.edu>; Mon, 24 Mar 2003 10:31:22 -0500
Received: from mhub-c2.tc.umn.edu (mhub-c2.tc.umn.edu [160.94.128.45])
	by fort-point-station.mit.edu (8.12.4/8.9.2) with ESMTP id h2OFVLQt023754
	for <hazelsct@mit.edu>; Mon, 24 Mar 2003 10:31:22 -0500 (EST)
Received: from PetrosII (falcon.cs.umn.edu [160.94.179.138] (may be forged)) by mhub-c2.tc.umn.edu with ESMTP for hazelsct@mit.edu; Mon, 24 Mar 2003 09:31:21 -0600 (CST)
From: "George Karypis" <karypis@cs.umn.edu>
To: "'Adam C Powell IV'" <hazelsct@mit.edu>
Subject: RE: Distributing ParMETIS with Debian
Date: Mon, 24 Mar 2003 09:29:03 -0600
Message-ID: <000f01c2f21a$1a1b8d80$f88010ac@PetrosII>
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_0010_01C2F1E7.CF82A420"
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook, Build 10.0.4024
In-Reply-To: <3E7E0DF8.8090707@mit.edu>
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106
Importance: Normal
X-Umn-Remote-Mta: [N] falcon.cs.umn.edu #+HF+LO

This is a multi-part message in MIME format.

------=_NextPart_000_0010_01C2F1E7.CF82A420
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: 7bit

 
Adam,
 
Sure, go ahead and include ParMetis in Debian. There are a couple
outstanding bugs with the
current release of ParMetis that I will be fixing soon, but the changes
will be minimal.
 
George

-----Original Message-----
From: Adam C Powell IV [mailto:hazelsct@mit.edu] 
Sent: Sunday, March 23, 2003 1:42 PM
To: metis@cs.umn.edu
Subject: Distributing ParMETIS with Debian


Greetings,

I am writing to request your permission to distribute ParMETIS with
Debian GNU/Linux.  Debian (http://www.debian.org/) is an all-volunteer
distribution of the Linux kernel with thousands of accompanying software
packages, similar to but much larger than RedHat, SuSE, Mandrake,
Lindows, etc. (in fact, Lindows is based on Debian's free packages).  I
am one of the ~850 volunteer maintainers of Debian; I maintain the PETSc
<http://www-unix.mcs.anl.gov/petsc/>  and Babel
<http://www.llnl.gov/CASC/components/babel.html>  packages among others.
Because PETSc can link to ParMETIS, I would like to be able to include
it in the Debian distribution.

[I am aware of the work of Eray Ozkural to package METIS for Debian, as
I was his mentor and sponsor when he was doing that work as part of the
process of becoming a Debian maintainer.  Unfortunately, for various
reasons his application to become a maintainer was rejected, and his
packages including METIS removed.  This will thus be a new packaging
effort by an already-approved maintainer.]

Debian packages are grouped according to the "freedom" of software
licenses.  Since the ParMETIS license is not "free" by Debian standards
(http://www.debian.org/intro/free ; basically the same as the Open
Source  <http://www.opensource.org/docs/definition.php> Definition), it
would go ino the "non-free" section of Debian, which would alert
potential users that copying and usage are restricted, so they should
check the copyright file for such terms.  (The other two sections are
"main", consisting entirely of free software, and "contrib", consisting
of free software which depends on non-free software, e.g. when PETSc
linking to ParMETIS is uploaded, its dependency on non-free ParMETIS
will require moving it into contrib.)  In every ParMETIS package (static
libs, shared libs, documentation, etc.), your copyright conditions would
appear as the file /usr/share/doc/parmetis*/copyright , and users would
be bound by those terms just as if they downloaded ParMETIS from your
website.

When/if I receive your permission, I will upload into the Debian archive
the pristine ParMETIS source tarball, together with a gzipped patch
containing the Debian packaging, and a third small file with MD5
checksums of the first two and a list of Debian packages required to
build ParMETIS binaries; this third file will be GPG-signed by me to
guarantee the integrity of all three files.  I would thus become the
"Debian maintainer" of the ParMETIS package, in the same way as I am the
maintainer of packages for PETSc, Babel, Surface Evolver, and installers
for the Compaq Alpha Linux compilers.  With your permission, I would
like to make some small changes, such as building a shared ParMETIS
library, and fixing any bugs I might find.  I will inform you in advance
of any such changes, and you will be able to exercise as much or as
little control over these changes as you like.

The Debian packaging mentioned above consists of a set of scripts and
definition files for the building of binary Debian packages based on the
source code.  Taken together, these files serve a similar purpose to the
spec file for creating RedHat .rpm packages, but with considerably more
flexibility.  The binary Debian packages will then be built by
autobuilder machines for all of the target architectures, which
currently include: IA-32, IA-64, Alpha, Sparc, PowerPC, M68K (yes, there
are still a bunch of Amiga users around :-), MIPS (big- and
little-endian), ARM, S/390 and PA-RISC; support for Hitachi SuperH and
alternate kernels on some of these architectures is in the works.

This packaging will make ParMETIS and packages which depend on it
considerably easier to use, by allowing users to download, say, a
parallel fluid dynamics package named "fluids" by typing "apt-get
install fluids", which would download and install fluids, the ParMETIS
library, and all other support libraries required to run fluids.  Users
(who run Debian) will not need to compile or link anything, this work
will all be done in advance.

If you (and the University of Minnesota) agree to the above, I will
begin working on the Debian packaging for ParMETIS, and when that
packaging is complete, will inform you and upload it into Debian.  This
will place ParMETIS in the "unstable" distribution immediately, and when
it meets several criteria (no build failures on any architecture, no
bugs of severity serious or higher, 10 days since upload into unstable),
it will be moved into the "testing" distribution.  If the package is in
the "testing" distribution when a release is made, it will be included
in that "stable" distribution release.

Independent of your decision regarding Debian, I would like to request
that you (U.Mn. etc.) make the METIS family available under a free/open
source license of some kind.  For example, the Babel SIDL compiler is
released under the GNU Lesser General Public license, which among other
things, eliminates the need for this sort of permission request for
Debian redistribution (though I am in frequent contact with Babel's
authors, and consulted them before my initial upload), and also allows
me to patch it in various ways important to my users.  In your case, you
might be more interested in the GNU General Public License, which would
not allow linking into proprietary software packages.  (As the copyright
holder, you could also license *METIS to companies under separate
conditions.)  As another example, I have simplified linking of the PETSc
libraries with C++ code without the extra disk space and build time
required by the authors' C++ scheme, and patched a problem in PETSc's
timestepping code; both of those patches were sent to the PETSc authors.
Thus open source licensing adds value to PETSc, as it does for such
well-known software as Linux, Apache, Mozilla, GNOME, OpenOffice,
autoconf/automake/libtool, etc., and could do the same for ParMETIS.

Thank you for your time and attention.  I hope that you will approve of
my request to package and distribute ParMETIS with Debian, and that you
will give serious consideration to open source licensing of ParMETIS.

Sincerely,

-- 

Adam Powell	 http://lyre.mit.edu/~powell/
<http://lyre.mit.edu/%7Epowell/> 	
Thomas B. King Assistant Professor of Materials Engineering	
77 Massachusetts Ave. Rm. 4-117	 Phone (617) 452-2086	
Cambridge, MA 02139 USA	 Fax (617) 253-5418	


------=_NextPart_000_0010_01C2F1E7.CF82A420
Content-Type: text/html;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable

[removed to avoid having to read "MSHTML" formatted material]

------=_NextPart_000_0010_01C2F1E7.CF82A420--

--------------090603000609020005070704--


---------------------------------------
Received: (at 186116-close) by bugs.debian.org; 13 Apr 2003 15:30:19 +0000
>From troup@auric.debian.org Sun Apr 13 10:30:19 2003
Return-path: <troup@auric.debian.org>
Received: from auric.debian.org [206.246.226.45] (mail)
	by master.debian.org with esmtp (Exim 3.12 1 (Debian))
	id 194jQw-0004o0-00; Sun, 13 Apr 2003 10:30:18 -0500
Received: from troup by auric.debian.org with local (Exim 3.35 1 (Debian))
	id 194jLL-00035H-00; Sun, 13 Apr 2003 11:24:31 -0400
From: hazelsct@debian.org (Adam C. Powell, IV)
To: 186116-close@bugs.debian.org
X-Katie: lisa $Revision: 1.23 $
Subject: Bug#186116: fixed in parmetis 3.0-2
Message-Id: <E194jLL-00035H-00@auric.debian.org>
Sender: James Troup <troup@auric.debian.org>
Date: Sun, 13 Apr 2003 11:24:31 -0400
Delivered-To: 186116-close@bugs.debian.org

We believe that the bug you reported is fixed in the latest version of
parmetis, which is due to be installed in the Debian FTP archive:

parmetis-dev_3.0-2_i386.deb
  to pool/non-free/p/parmetis/parmetis-dev_3.0-2_i386.deb
parmetis-doc_3.0-2_all.deb
  to pool/non-free/p/parmetis/parmetis-doc_3.0-2_all.deb
parmetis_3.0-2.diff.gz
  to pool/non-free/p/parmetis/parmetis_3.0-2.diff.gz
parmetis_3.0-2.dsc
  to pool/non-free/p/parmetis/parmetis_3.0-2.dsc
parmetis_3.0.orig.tar.gz
  to pool/non-free/p/parmetis/parmetis_3.0.orig.tar.gz



A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed.  If you
have further comments please address them to 186116@bugs.debian.org,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Adam C. Powell, IV <hazelsct@debian.org> (supplier of updated parmetis package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing ftpmaster@debian.org)


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Format: 1.7
Date: Tue,  8 Apr 2003 17:54:53 -0400
Source: parmetis
Binary: parmetis-dev parmetis-doc
Architecture: source i386 all
Version: 3.0-2
Distribution: unstable
Urgency: low
Maintainer: Adam C. Powell, IV <hazelsct@debian.org>
Changed-By: Adam C. Powell, IV <hazelsct@debian.org>
Description: 
 parmetis-dev - Parallel Graph Partitioning and Sparse Matrix Ordering Library
 parmetis-doc - Parallel Graph Partitioning and Sparse Matrix Ordering Lib - Docs
Closes: 186116
Changes: 
 parmetis (3.0-2) unstable; urgency=low
 .
   * New permission from upstream should result in acceptance (closes: #186116).
Files: 
 ca05e1f9308790cf0b84ea0fb59598e7 590 non-free/math extra parmetis_3.0-2.dsc
 0bf146e60d9eb63f70ae388d7b35e6ba 7881804 non-free/math extra parmetis_3.0.orig.tar.gz
 d9bdf9f68780f892e0c2ae7e3ffcde1b 6495 non-free/math extra parmetis_3.0-2.diff.gz
 61da1b2fb968ace1761dc1130eed5daa 7544190 non-free/math extra parmetis-doc_3.0-2_all.deb
 369c3876d5acb3808f730f5a71ed0964 233828 non-free/math extra parmetis-dev_3.0-2_i386.deb

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+k0WxUm8B6FZO5LYRAqJ2AJkBlUXL331nVbgsGu/9vVyYRIh4qwCfV9RY
CRBttXYPWOhoO7xGpPXbDcc=
=BBkJ
-----END PGP SIGNATURE-----



Reply to: