Your message dated Sun, 18 Oct 2009 13:24:21 -0400
with message-id <1255886661.3941.368.ca...@workhorse>
and subject line Re: blacs-mpi/scalapack transition
has caused the Debian Bug report #550648,
regarding RM: blacs-pvm-test, scalapack-lam-dev, scalapack-lam-test,
scalapack1-lam, scalapack-mpich-dev, scalapack-mpich-test, scalapack1-mpich --
ROM, NVIU
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 this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact ow...@bugs.debian.org
immediately.)
--
550648: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=550648
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
--- Begin Message ---
Package: release.debian.org
X-debbugs-cc: muammarelkha...@gmail.com
Please remove the old blacs-mpi and scalapack binary packages from
testing, which have been blocking the transition of those packages for
five months. The new packages use mpi-default-dev instead of building
multiple binaries against the mpich and lam implementations of MPI.
I NMUd the new packages with the consent of the maintainer (copied here)
as shown in bugs 491028 and 491105, and uploaded MUMPS which depends on
them (but currently has an FTBFS bug). I don't believe any reverse
dependencies in unstable still link against the old binaries.
Thanks,
-Adam
--
GPG fingerprint: D54D 1AEE B11C CE9B A02B C5DD 526F 01E8 564E E4B6
Engineering consulting with open source tools
http://www.opennovation.com/
signature.asc
Description: This is a digitally signed message part
--- End Message ---
--- Begin Message ---
On Sat, 2009-10-17 at 20:16 -0430, Muammar El Khatib wrote:
> Hi,
>
> On Sat, Oct 17, 2009 at 4:56 AM, Luk Claes <l...@debian.org> wrote:
> > Adam C Powell IV wrote:
> >> Hello,
> >>
> >> blacs-mpi and scalapack have been in transition for over four months.
> >> AFAICT, it's because of old binary packages in testing linked to mpich
> >> and lam getting in the way. (The new packages use mpi-defaults.)
> >>
> >> Would it be possible to remove those binaries and hint blacs-mpi and
> >> scalapack into testing?
> >
> > There is an uninstallable binary package which prevents them from
> > migrating (reported as bug #549707).
Luk, thanks for pointing out this bug which was recently filed. I had
not seen it, and was unaware of this issue blocking blacs-mpi and
scalapack from entering testing. I'm closing my bug against
release.debian.org .
> I have uploaded a revision of scalapack in mentors.d.n:
> http://mentors.debian.net/debian/pool/main/s/scalapack/
>
> If I understood #549707 correctly scalapack-pvm-test depends on
> scalapack1-pvm, but libscalapack-pvm1 is shipped instead and that's
> the problem. So, what I did was to update the debian/shlibs.local
> file to make scalapack-pvm-test depends on the correct binary. Please,
> correct me if I am wrong with this.
Muammar, there is still a problem with shlibs.local, which is that it
refers to non-existent packages libscalapack-mpich1 and
libscalapack-lam1 . I think you switched the shlib package name from
libscalapack-mpi1 to libscalapack-openmpi1 to avoid a lintian warning.
But the problem is that on architectures where OpenMPI is not available,
the shlib soname is libscalapack-lam.so.1 which seems even more
inappropriate for a package called libscalapack-openmpi1 . That's why I
created the package names libscalapack-mpi1 and libscalapack-mpi-dev .
If in spite of this you would like to keep the package name as
libscalapack-openmpi1 I will respect your wishes as the maintainer. But
shlibs.local needs to be modified so all of the shlib package names are
the same (libscalapack-openmpi1) before the package can be uploaded.
Thanks,
-Adam
--
GPG fingerprint: D54D 1AEE B11C CE9B A02B C5DD 526F 01E8 564E E4B6
Engineering consulting with open source tools
http://www.opennovation.com/
signature.asc
Description: This is a digitally signed message part
--- End Message ---
