Librarian and Integrators meeting

Thursday 30 Oct 2008, 11:00 → 12:00 Europe/Zurich

TBA

Stefan Roiser (CERN)

The bi-weekly meeting of the LHC librarians and LCG/AA projects

Attendees: Bertrand Bellenot, Emil Obreshkov, Andrea Valassi, Hubert Degaudenzi, Andreas Pfeiffer, Stefan Roiser News === LCG 54h --- The configuration has been released Oct. 21. It is already in use by Atlas LCG 55b --- The next configuration to be produced. The current set of problems with it are - Some Coral libraries contain unnecessary dependencies on cppunit, these should be removed. - For this configuration we had to roll back to VC 7.1 the current issues with this compiler are - The "application" nmake file is broken after the VC9 adaptations - The dynamically generated cmt macro python twodigit returns "2.5" instead of "25" - PyCOOL does not work with gccxml 0.9. This is not necessary as this configuration will be released with gccxml 0.7 - Another issue from the VC9 porting. All macros should be passed to the compiler with "-D" instead of "/D" as the same macros will be used for compilation (that's fine cl does unterstand -D) and for the parsing by gccxml (which only understands -D). - Some minor adjustments to COOL are pending - The nightly builds for VC7.1 are broken, they need to be put into shape again. - This configuration will neither contain gcc 4.3 nor Windows VC9 builds. The ROOT release based on it (5.18.x) is not ready yet for these compilers. LCG 56 --- The next big configuration after 55b. We will try to start releasing a first configuration right after the next ROOT development release, scheduled for Nov. 26. - This configuration will contain both VC9 + VC7.1 builds on windows - The first release to contain gcc 4.3 builds - Apart from some other minor upgrades it will also contain the move to Boost 1.36 which is the first Boost version to officially support gcc 4.3 and VC9 - A question came up for this release, whether the generators were already successfully compiled with the gfortran of gcc 4.3? HepPDT --- We are trying to investigate the usage of HepPDT in LHC. CMS 3.02.03 LHCb not used - to be verified Atlas to be verified Round Table === Root --- Probably a new patch release (5.18.00f) is pending. The release date is not yet known. It will be tried to incorporate it into the 55b configuration. CMS --- - Is starting to port their software to gcc 4.3 Atlas --- - A cmt problem with diamond graph on project level with LD_LIBRARY_PATH was found - a new release of cmt is pending - Planning to release on top 54h next week - i.e. 14.4.0 - Once 55b is available Atlas would like to switch to it ASAP - The nighthly builds against dev1 (55 series) look very good - A problem was found when running slc4 binaries on slc5 machines. When SELinux was disabled Atlas were able to run some jobs. Apart from that another problem with memory was found, maybe a problem with the SLC5 kernel? LHCb --- - Gaudi is ready to be frozen today - to be used in LCG 55. Release to be done next week. - LHCb is using cmt 20080222 - The latest installation of doxygen might be broken? The external references for tag files are always understood as a relative path. Hubert will do more testing ...

11:00 → 11:10 News

11:10 → 11:40 Round table

11:40 → 12:00 Aob