ScienceTools: (Jim) gave the report. (Richard) When is a new tag expected? Navid needs one incorporating some fixes for Mac builds. (Jim) can hurry one along. When this kind of situation arises, Navid — or whoever — need not be shy about asking for a new tag.
FSSC Report: (John, Eric) are finding and addressing lots of niggling differences in behavior on different platforms. Some merely affect test programs; others could cause a crash when running the tools. They're aiming for a new release on Feb. 6th.
Overlay Two areas to focus on currently: 1) understanding what the code in its current state does 2) getting production-scale Pipeline job(s) going. (Leon) Concerning 1), the procedure for selecting overlay events looks for those where periodic trigger bit is set, regardless of occurrence or non-occurrence of other triggers. (Tracy) Is this what we want? (Leon) Once selected, events are read in and binned. There is plenty of opportunity to impose additional cuts if we want. But it would be a good idea to understand what we want beforehand.
(Tom) provided the following concerning 2):
A pipeline task has been assembled, Overlay-maker, to extract periodic trigger events from on-orbit LAT data and write out new files to be used in future simulations. This task reads DIGI files and outputs MERIT and new OVERLAY root files. Five runs have been processed for validation. It is expected that overlay events will ultimately be extracted from ~1700 runs; such a pipeline task should take less than one day to complete. The next step in the processing is to concatenate these run-based files, bin them in McIlwain L and write out new binned files. At that point they are ready for simulation.
(Richard) Once this is all working we'll need 2 billion background events for Pass 7. (Philippe) wants some all-gammas and a sample day. And then whatever it takes to produce new IRFs. (Richard) will have to restrict sample day to merit only; we don't have enough disk space to save recon. [no objections voiced]
GR tags (Heather)
(Heather) There is a new GlastClassify tag, not yet in any release tag. (Tracy) It's needed to run Pass 7, but is a transparent change which should not cause any problems.
More about AnalysisNTuple: (Richard) Leon made mods for Heavy Ion stuff. If we're going to make runs with it, it should be in a real tag. (Heather) will make a tag including it off of the v15r53 branch. (Richard) New variables related to magnetic latitude are also proposed. (Tracy) And delta window time and livetime. (Anders) Eric may be using delta event time in the MC for the ACD coherent noise. Readout livetime is in the data itself (from FSW).
Platforms (Heather) would like to try out RHEL5 builds, but it's awkward doing it by submitting jobs to the couple of RHEL5 machines available in batch. (Richard) There is a RHEL5 noric, probably noric15.
(Heather) is looking into making a build of G4 with 64-bit flag on. Ultimately we may want to do this for other externals as well and make a serious attempt to use 64-bit builds in the Pipeline.
Documentation: (Chuck) sends this summary:
There's a new ASP Data Viewer help page linked in from the 2nd-level, Data Portal 'Help' navbar. I've also added some updates to the Science Analysis splash page, and added the Christmas Ice Cream Social pictures that Richard took to the photo section of the ISOC public website. In the process, I'm bringing Helen Butler up to speed on both the ISOC and the SAS picture galleries, so she can take them over.
Currently, I'm working on this month's newsletter, and researching the history of TKR, RECON, and CAL(orimeter)in order to develop a high-level overview of the software and link in the Doxygen documentation. I'm also doing some research on the Source Catalog. As for routine updates to the workbook, I've recently created a new Change Log and added a link to it from the workbook's rednavbar splash page.
Confluence outage (Heather) It will be taken down from 6-8 PM PST Wednesday night for an upgrade.
ACD mini-review (Heather) Any news? (Richard) Waiting for Bill's return.
RMViewer (Heather) John responded to the plea for guinea pigs by trying out RMViewer 0.5 and also sending comments, thereby earning a virtual beer . (Joanne) was able to use it to download a user release at SLAC and run some of the programs it contained. She was less successful downloading from home, but that could have to do with the poor network connection.
Xerces (Emmanuel) Xerces with patch for Mac (gl1 version) is built. He got it done shortly before Navid left on vacation last week.
CHS (Joanne) All changes needed for building under SCons are ready to go. She would like clearance from one or both of Navid and Bryson before committing, though.
GoGui (Joanne) is working through various issues connected with getting SCons to make acceptable VS 9.0 project files. It's very slow going but the sign is positive.
GR SCons build (Emmanuel) GlastRelease-15-49-00 is all done.
New python version prospects [Email from Heather after the meeting covers all bases nicely:]
At the software meeting today, Toby asked about our prospects of moving to python 2.6. I'm assuming Jim and Navid will have some opinions on this matter. My only concern is trying to keep our eye on the ball as far as SCons goes - but my understanding is that this version of python will allow us to prepare for python 3.
Tom has asked that we at least provide a version for testing python scripts. I would ask that we provide python 2.6 for rhel4 for now. As I recall, tcl/tk 8.4 is readily available on the rhel4 machines, and so we don't have to trouble ourselves with doing a special build using the tcl/tk version from HEADAS.
At this point - I'm just looking for feedback and then we can see what can be done.
previous | minutes index | next |