Core Minutes 12/15/2009

• Down-under report: (Richard) visited Melbourne, Adelaide, and Sydney. During the first couple days he ran a tutorial at Monash University. There were about 10 people in attendance. The first day — introduction to the LAT and Likelihood demo — went well. The demo was on 3C454, a booming extragalactic source that makes great training material. The second was less successful, in spite of heroic efforts by Eric to address problems with the pyLikelihood interface on Linux. But overall the tutorial was a success. Richard's impression was that these people might be intereted in gamma ray data after they see something in other wavelengths.

• ScienceTools: (Jim) See this week's brief report in Confluence; nothing much to report other than some pulsar tool development.

  (Jim, Liz, Julie) A brief discussion concerning bracketing IRFs converged on agreement to put the in CALDB.

• FSSC: (Eric) continues to work on merging changes needed to support all FSSC platforms back into SLAC cvs. Typical changes are small, e.g. adding a #include of a header file because gcc 4 requires it.

• Documentation: (Chuck) sends the following:

  Karen, Tom Glanzman, and I spent a fair amount of time last week sorting out the various issues relating to the mandatory Cyber Security training, especially regarding problems being experienced by our overseas collaborations. There seem to be a lot of 'Catch-22' situations. As a result, there were a series of updates to the WB page, but we now think that we've found a path through the thicket. The latest version has been posted to the web, and I now have to create a new tag for the WB, so that the mirror site can be updated and remain consistent with both the WB and the Newsletter (in which we're also running a blurb in an effort to get the word out in time for the January 4th deadline).

  On another front, I believe that I reported having posted the new GRBanalysis Tutorial to the web last week; at least, the CMT build version of it. SCons builds are still failing in that pyLikelihood cannot seem to find the HDU column in the FT1 file, even though gtselect and gtmake complete with no problems reported.

  With respect to the final issue of the NL, most of the contributions have been received pretty much on schedule, and we're about 80% done. So... it looks like we should meet Thursday's publication date.

  Beyond these issues, I'm primarily working on the normal, end-of-year cleanup tasks.

• GR:

  • L1 branch (Heather) TkrUtil is quite old. Leon would like to see it updated, but L1 folk are not eager, in line with their usual conservatism. She has already made a GR tag including the newer version of astro, to avoid possible incompatibilities in f2tutil. That tag is not yet in use. Ditto the tag made a couple weeks ago to include the enhanced (a year ago!) rdbModel which will retry MySQL connections. (Richard) That one is immininent.

  • Pass 7 (Richard) Pass 7 work is in limbo pending resolution of problems which may be due to reprocessing, IRFs or pass 7 itself. See discussion in links off the C & A Agenda page .

  • Heading for Pass 8 (Heather) has upgraded a bunch of packages in HEAD, picking up nearly all the changes Joanne made for SCons builds. There are a few more things to go into HEAD before she makes a GR tag in the v17 series.

    Eric Charles' ACD recon tags will go in the first v18 tag along with other updates from Tracy and Leon. (Leon) Recent work has mostly been Tracy's, responding to problems seen by introducing clustering; e.g., what happens when there is more than one.

• Core talk:

  • u30 clean-up (Heather) has been accomplished! We should be ok for a while.

  • Ditching rhel3 (Heather) SLAC will continue to support rhel3 until about next fall, though rhel3 machines in the batch farm will be converted to rhel5 before that. She would like to end our dependence on it by late January. (Jim) believes ASP will be able to do without it by then.

  • ASP and SCons (Joanne) All problems uncovered so far by Jim have been addressed. She's waiting for more feedback as Jim exercises it. (Jim) It's pretty close to working.

  • GR and SCons (Joanne) Remaining issues she will address are reorganization of the obf external and work on job options. She wrote a python script to extract what we need from a flight software build and write to a more SCons-friendly arrangement. It doesn't yet deal with the messier aspects (duplicate library names; include files we need buried in source directories) but she hopes to complete it this week.

  • Still more SCons (Joanne) Since the SCons RM is now using the SCons checkpoint release for all builds, extra support for Windows is no longer incompatible with other platforms. She made a new tag of SConsFiles, 00-03-00, which includes everything relating to building project files on Windows. Instructions for Windows users in Confluence will be updated accordingly.

  • Externals for GR (Heather) The new Gaudi builds on rhel4 as well as Windows. She will also build it for rhel5 but probably not for rhel3, in hopes that we will cut the cord soon. However changes must be made to our GR code to accommodate it; it's not backwards compatible.

  • Macs (Richard) Currently one of our batch macs is running Tiger, the other is running Leopard. If we don't need Tiger any more, could upgrade that one to Snow Leopard. (John) ScienceTools doesn't build on Snow Leopard. (Eric) The Leopard binary from FSSC runs on Snow Leopard; don't know how well. (John) It isn't the same gcc version.

• Welcome back! to Tom Stephens, who will be taking over some of the SCons support. He'll be able to give us about 10-20 hours a week for now; full-time in February.

previous

minutes index

next

J. Bogart, Last Modified: 01-Jun-2010 15:45:37 -0700