ScienceTools: Jim presented the report. (Richard) Will the dictionary file be checked into CVS? (Jim) Yes, this should be done. We have to decide exactly where to put it. [He has since put it into evtClassDefs and tagged it as v6r6p1]
FSSC Report: (Eric W.) v9r11 has been ingested; will start porting to the usual collection of Linux platforms.
Documentation: No report this week. Chuck is home with the flu.
GlastRelease: (Richard) Background and all-gamma simulations have been run. The skimmer doesn't like the new headers. Tom was able to do the skimming by hand on a sample; since we don't want to wait a day or two for a skimmer fix, he'll be doing the remainder. [Done. Bill has the files.]We're shooting for the end of April to get all the Pass7 reprocessing done. See details in Confluence.
(Philippe) has been looking at Pass 6 done with v17r16. sees 20% more events passing trigger in all-events file than in triggered-only set. Some small discrepancy might be expected, but this seems large. He doesn't yet know whether he's not using the right cuts or if there is an underlying problem.
L1 news (Anders) There is a new variable, TEMBug, in the SVAC tuple. Will be moving to GR v15r47p7.
(Anders) Has anyone tried out v17r<something> on real data? If not, I will. (Heather) has for an early version but encourages Anders to take another look; she recommends v17r16.
Header problem (Heather) has been looking at an intermittent problem extracting versions from headers: show uses doesn't always work. It seems that CMT has lost its way. Bah! Bah! Bah!
More cores (Richard) 400 more cores have been added to our batch allocation.
Python (Heather) Just confirming: we're going to stick with 2.5.1 but make 2.6 available for those who want to try it out, right? [right.] Requests for IPython and a newer version of pyfits have been satisfied in 2.5.1-gl1.
Archiving status (Tom) One archive failed. Wilko will track down and resubmit. Otherwise we're good to go as soon as
Disks (Richard) Will return 16Tbytes to Babar, no longer needed. We're looking into applying quotas to a "new" user disk, 9 Tbytes of space released from xrootd use.
SCons RM (Navid) Email notification is now active for ScienceTools builds on RHEL4.
Neal put in a fix for Mac/LSF and it appears to have worked. RM will be able to make Mac builds soon.
There is a bug in use of swig: output .py file ends up in the wrong place so it can't be found by programs needing to reference it. Should not be difficult to fix.
SCons/Windows and GoGui (Joanne) Dealt with a pre-condition to item 2 (custom tool for swig in project files) on the To-do list. Swig, which works fine on central Linux, was not working on my RHEL4 laptop, most likely because the swig executable has hard-coded path information which is of no use when SLAC afs space isn't accessible. Some changes to externals.scons took care of it.
I have a rough plan for how to get dependency information into solution files. The first, relatively cut-and-dried step — relocating project and solution files to a common install area — is done.
Externals (Emmanuel) rebuilt Fred v0r100 for RHEL4 and made an interface package for expat, regularizing the way Fred finds libraries it needs, but apparently to no avail. He still sees the same error he saw earlier when trying to run Fred from Gleam (it's ok when run stand-alone).
He is also currently looking into SCons builds on Windows.
(Heather) spent some time with Emmanuel's SCons build of GR v15r49. She plans to extend this work to the v17.. series and ultimately to HEAD.
previous | minutes index | next |