[DebianGIS-dev] Bug#521221: [h...@hdfgroup.org: Bug 1534 HDF5 library names]
tags 521221 + wontfix
tags 521221 + upstream
thanks
FYI, if anyone is able to push a different approach to HDF group
this could be fixed. It is a non sense fixing this in debian only
because we would loose source compatibility with upstream.
- Forwarded message from h...@hdfgroup.org -
Date: Fri, 5 Jun 2009 12:38:36 -0500 (CDT)
From: h...@hdfgroup.org
To: fran...@debian.org
Subject: Bug 1534 HDF5 library names
User-Agent: Alpine 2.00 (LRH 1167 2008-08-23)
Hello Mr. Lovergine,
Thank you for your suggestion. We have looked into using
suffix to distinguish different configurations of HDF5
libraries but decided against it due to three concerns:
1. This would require the HDF5 library users to maintain different
versions of autoconfigure and/or Makefiles. This is a
minor concern.
2. A more major one is that HDF5 can have several different
configurations on many big platforms. E.g.,
Due to the big diversity of users in those big platforms,
the System Admin may have built and installed multiple
variations of the same HDF5 release:
a. with or without MPI parallel support;
b. 32bit or 64 bits OS support (e.g., AIX, SunOS, ...);
c. different compilers like Gcc, Intel, or PGI vendors;
d. different versions of the same compilers (e.g., gcc 4.2, 4.3);
e. with or without some HDF5 supported compression features;
In those platforms, the admin usually installs the binaries
of each variation in a separated directory. E.g.,
/usr/apps/hdf5/v1.8.3_gcc4.2_mpich2_107/{bin,lib,include}
/usr/apps/hdf5/v1.8.3_Intel10.1/{bin,lib,include}
...
3. Lastly, a critical concern is that we need to deal with the
HDF5 header files too. The header files of an MPI-Supported
HDF5 built are different in contents from a serial-only HDF5
built. They cannot be installed in the same directory. If
we were to rename the header files with different suffixes,
it would make HDF5 application source code files much more
complex.
Therefore, we have chosen to stay with the same name for the HDF5
library and header files. but rely on other software loading
tools such as module to help the users to pick and choose
their own flavors of HDF5 libraries.
We recognize that this is not the perfect solution for everyone
in all situations but we feel this is a good compromise for our
users.
From Francesco P. Lovergine
...
* MPI flavors use the same names of serial libraries, which prevents installing
and using both serial and MPI flavors of the libraries on the same system.
Other libraries provide a _mpi suffixes for MPI-enabled libraries to allow
that.
=
- End forwarded message -
--
Francesco P. Lovergine
___
Pkg-grass-devel mailing list
Pkg-grass-devel@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/pkg-grass-devel
[DebianGIS-dev] Bug#521221: marked as done (libhdf5-openmpi-1.6.6-0: simultaneous installation of -serial and -mpi packages)
Your message dated Mon, 8 Jun 2009 10:20:30 +0200
with message-id 20090608082030.gb4...@mithrandir
and subject line [h...@hdfgroup.org: Bug 1534 HDF5 library names]
has caused the Debian Bug report #521221,
regarding libhdf5-openmpi-1.6.6-0: simultaneous installation of -serial and
-mpi packages
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.)
--
521221: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=521221
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
---BeginMessage---
Package: libhdf5-openmpi-1.6.6-0
Version: 1.6.6-4
Severity: wishlist
===
Consider simultaneous installation of serial and parallelized (-mpi) packages.
So the needed
library will be choosen by the application. This will give more flexibility as
applications
that need the serial version can abe installed on the same system with
applications that
need the -mpi version. (Similar as has been done by libgpiv3 and libgpiv-mpi3.)
-- System Information:
Debian Release: squeeze/sid
APT prefers unstable
APT policy: (500, 'unstable')
Architecture: i386 (i686)
Kernel: Linux 2.6.26-1-686 (SMP w/2 CPU cores)
Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968)
Shell: /bin/sh linked to /bin/bash
Versions of packages libhdf5-openmpi-1.6.6-0 depends on:
ii libc6 2.9-4 GNU C Library: Shared libraries
ii libopenmpi11.3-2 high performance message passing l
ii zlib1g 1:1.2.3.3.dfsg-13 compression library - runtime
libhdf5-openmpi-1.6.6-0 recommends no packages.
libhdf5-openmpi-1.6.6-0 suggests no packages.
---End Message---
---BeginMessage---
tags 521221 + wontfix
tags 521221 + upstream
thanks
FYI, if anyone is able to push a different approach to HDF group
this could be fixed. It is a non sense fixing this in debian only
because we would loose source compatibility with upstream.
- Forwarded message from h...@hdfgroup.org -
Date: Fri, 5 Jun 2009 12:38:36 -0500 (CDT)
From: h...@hdfgroup.org
To: fran...@debian.org
Subject: Bug 1534 HDF5 library names
User-Agent: Alpine 2.00 (LRH 1167 2008-08-23)
Hello Mr. Lovergine,
Thank you for your suggestion. We have looked into using
suffix to distinguish different configurations of HDF5
libraries but decided against it due to three concerns:
1. This would require the HDF5 library users to maintain different
versions of autoconfigure and/or Makefiles. This is a
minor concern.
2. A more major one is that HDF5 can have several different
configurations on many big platforms. E.g.,
Due to the big diversity of users in those big platforms,
the System Admin may have built and installed multiple
variations of the same HDF5 release:
a. with or without MPI parallel support;
b. 32bit or 64 bits OS support (e.g., AIX, SunOS, ...);
c. different compilers like Gcc, Intel, or PGI vendors;
d. different versions of the same compilers (e.g., gcc 4.2, 4.3);
e. with or without some HDF5 supported compression features;
In those platforms, the admin usually installs the binaries
of each variation in a separated directory. E.g.,
/usr/apps/hdf5/v1.8.3_gcc4.2_mpich2_107/{bin,lib,include}
/usr/apps/hdf5/v1.8.3_Intel10.1/{bin,lib,include}
...
3. Lastly, a critical concern is that we need to deal with the
HDF5 header files too. The header files of an MPI-Supported
HDF5 built are different in contents from a serial-only HDF5
built. They cannot be installed in the same directory. If
we were to rename the header files with different suffixes,
it would make HDF5 application source code files much more
complex.
Therefore, we have chosen to stay with the same name for the HDF5
library and header files. but rely on other software loading
tools such as module to help the users to pick and choose
their own flavors of HDF5 libraries.
We recognize that this is not the perfect solution for everyone
in all situations but we feel this is a good compromise for our
users.
From Francesco P. Lovergine
...
* MPI flavors use the same names of serial libraries, which prevents installing
and using both serial and MPI flavors of the libraries on the same system.
Other libraries provide a _mpi suffixes for MPI-enabled libraries to allow
that.
=
- End forwarded message -
--
Francesco P. Lovergine
---End Message---
___
Pkg-grass-devel mailing list
Pkg-grass-devel@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/pkg-grass-devel
[DebianGIS-dev] r2269 - packages/osm2pgsql/trunk/debian
Author: nd-guest Date: 2009-06-08 21:45:34 + (Mon, 08 Jun 2009) New Revision: 2269 Modified: packages/osm2pgsql/trunk/debian/changelog Log: close #532145 Modified: packages/osm2pgsql/trunk/debian/changelog === --- packages/osm2pgsql/trunk/debian/changelog 2009-06-07 13:40:30 UTC (rev 2268) +++ packages/osm2pgsql/trunk/debian/changelog 2009-06-08 21:45:34 UTC (rev 2269) @@ -1,6 +1,6 @@ osm2pgsql (0.66.20090526-1) unstable; urgency=low - * New upstream svn snapshot. + * New upstream svn snapshot. (Closes: #532145) * Removed 10-missing-includes patch, fixed upstream. * Updated long description, mostly taken from osm2pgsql.spec.in. * Updated copyright year of debian packaging. @@ -10,7 +10,7 @@ * Use dh_prep instead of dh_clean -k to fix a lintian warning. * Updated manpage. - -- Andreas Putzo andr...@putzo.net Tue, 26 May 2009 19:40:51 + + -- Andreas Putzo andr...@putzo.net Mon, 08 Jun 2009 21:44:57 + osm2pgsql (0.52.20080408-3) unstable; urgency=high ___ Pkg-grass-devel mailing list Pkg-grass-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/pkg-grass-devel
[DebianGIS-dev] Processed: Re: Bug#532145: Acknowledgement ([osm2pgsql] osm2pgsql: always forces a password prompt - breaks unassisted map generation)
Processing commands for cont...@bugs.debian.org: tags 532145 + fixed-upstream pending Bug#532145: [osm2pgsql] osm2pgsql: always forces a password prompt - breaks unassisted map generation There were no tags set. Tags added: fixed-upstream, pending thanks Stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) ___ Pkg-grass-devel mailing list Pkg-grass-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/pkg-grass-devel
[DebianGIS-dev] Bug#532145: Bug#532145: Acknowledgement ([osm2pgsql] osm2pgsql: always forces a password prompt - breaks unassisted map generation)
tags 532145 + fixed-upstream pending thanks Hi, On Jun 07 04:08, Eddy Petrișor wrote: Package: osmqpgsql I just looked into the code and it seems that whenever the user parameter was given, the password prompt appeared, evn if not necessary. Thanks for your bug report and the patch. The bug is already fixed upstream the same way as you did. Please expect an updated packages soon. (D)Patch is attached, I tested it on my password-less config on a lenny machine (made a local backport) and it works properly. If used without a password when is actually needed this message appears, (which, imo, could be improved to add a hint, if -W wasn't present - maybe a password is required?), so I'd say the fix makes sense and doesn't break anything that worked before: 0 e...@heidi ~/usr/src/osm/planet/git-planet-rom $ osm2pgsql -d gis -m -U osm planet-rom.osm osm2pgsql SVN version 0.52-20090607 $Rev: 7141 $ Connection to database failed: FATAL: Ident authentication failed for user osm The error message comes from postgres, i guess that's nothing osm2pgsql should deal with :) Andreas signature.asc Description: Digital signature ___ Pkg-grass-devel mailing list Pkg-grass-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/pkg-grass-devel
[DebianGIS-dev] Bug#532145: Bug#532145: Acknowledgement ([osm2pgsql] osm2pgsql: always forces a password prompt - breaks unassisted map generation)
Andreas Putzo a scris: tags 532145 + fixed-upstream pending thanks Hi, On Jun 07 04:08, Eddy Petrișor wrote: Package: osmqpgsql I just looked into the code and it seems that whenever the user parameter was given, the password prompt appeared, evn if not necessary. Thanks for your bug report and the patch. The bug is already fixed upstream the same way as you did. Please expect an updated packages soon. Are you planning to do official backports, too? I made my own patched package, but maybe other people would find them useful. (D)Patch is attached, I tested it on my password-less config on a lenny machine (made a local backport) and it works properly. If used without a password when is actually needed this message appears, (which, imo, could be improved to add a hint, if -W wasn't present - maybe a password is required?), so I'd say the fix makes sense and doesn't break anything that worked before: 0 e...@heidi ~/usr/src/osm/planet/git-planet-rom $ osm2pgsql -d gis -m -U osm planet-rom.osm osm2pgsql SVN version 0.52-20090607 $Rev: 7141 $ Connection to database failed: FATAL: Ident authentication failed for user osm The error message comes from postgres, i guess that's nothing osm2pgsql should deal with :) Well, is, imo, sufficient to indicate the possible cause of the login issue. Better handling is preferable, but I don't this is crucial to improve this, especially if it complicates the code a lot (I haven't looked). -- Regards, EddyP = Imagination is more important than knowledge A.Einstein signature.asc Description: OpenPGP digital signature ___ Pkg-grass-devel mailing list Pkg-grass-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/pkg-grass-devel
