Core Minutes 3 March 2015

Core Minutes 3/3/2015

Present: Joe Asercion, Joanne Bogart, Jim Chiang, Richard Dubois, Elizabeth Ferrara, Tom Glanzman, Heather Kelly, Michael Kuss, David Landriu, Leon Rochester

ScienceTools: (Jim) Since our last meeting (Feb. 17) there has been a new ST tag, 10-00-02, basically bug fixes to 10-00-00. In development, he's added two new function classes (PiecewisePowerLaw, MultipleBrokenPowerLaw). The code is currently in LATEST. In progress is work on handling effective area systematics, to be finalized this week. See Science Tools Development Notes for details.
FSSC: (Joe) He has been looking into errors associated with the new versions of ape and cfitsio. He has fixes for tip and evtbin and other packages will probably require a similar fix. Under certain circumstances the new cfitsio throws an exception when it sees a null column, which the older version did not. It's not a fatal error, but the caller has to be prepared to catch the exception. (Jim) Or perhaps we could put something in the null columns.
Reprocessing: (Tom) While transferring FT1 files to FSSC a duplicate event was discovered. This was a "sweep event", occurring at the start of the run. Normally they're empty, but this one triggered somehow (perhaps ghost event). Taking careful inventory, M.E. discovered 85 similar events. We need code to detect and skip such events.
Resources: (Tom) wain25 is (still) being routinely assaulted by users it is meant to serve. We plan to move some servers from xroot to nfs.

(Richard) We're looking into buying new servers to replace glast-win01, glast-win02, etc. We expect to get three new servers.
Pass8 Public Release(Leon) There is a long list of to-do items, many of them non-trivial, which are to be completed before the release aggresively scheduled for the end of this month.

The saga of comparing releases (v9 to v10; 32-bit to 64-bit) continues. Most differences are due to the architecture change.
CAL Asymmetry Calibrations(Leon) Sasha has finally pushed through an idea he had long ago. Instead of doing asymmetry calibrations one by one, he has been able to fit the whole sequence. This is a good idea because statistics for some crystals are quite low.
Core talk:
- Windows: (Heather, Richard) glast-win04, which has been the Windows build machine, is one of the Windows servers soon to be decommissioned. Our erstwhile group of 3 Windows developers has dwindled down to just one, Bill, who has graciously acknowledged that he can do without automatically-generated builds.
- New fsw release: (Heather) has consulted the Online Subversion repository to confirm that no changes were made to ldf for the new fsw release. We can therefore conclude that no changes are required on our side. The entire discussion concerning the new fsw release and how it affects us can be viewed in JIRA EXTLIB-15.
- Gaudi debug (Heather) Michael has requested a (rhel6 64-bit) debug version of Gaudi to aid him in his investigations of memory usage. She's working on it.

minutes index

J. Bogart, Last Modified: