RSP and 64-bit (Tom G.) tempo2 is an application that adds a pulsar phase column to FT1 files. It is used in the binaries (MQ) RSP task. I have recently run into two problems. First it does not build/run reliably on 64-bit architectures (crashes at end). Second, under some circumstances some fraction of the last events in an FT1 file do not have their pulsar phase column filled properly (even on 32-bit machines). This is being investigated. [thanks to Tom for the above. ed.]
ScienceTools: (Jim) See the Science Tools blog in Confluence.
FSSC: (Eric) is working on the port of v9r16p1 to Snow Leopard. The main hang-up is xerces 2.8 which is now pretty old [current release is 3.1.1. ed.] and does not compile out of the box on Snow Leopard. I have to apply several overlapping (and incompatible) patches from the developers, and it is not going smoothly. I expect to have it worked out this week. I hope...
(John)It sounds like things are very close to an agreement (on the SLAC end) on the filename changes and new header keyword that the FSSC would need to support the dual ingest pipelines. I will be making the changes to our dataserver to support what I think will happen in as flexible a way as I can now, so we can handle the changes as quickly as possible. For our ingest system to be ready, however, Don Horner will need some sample files by email first.
[Many thanks to John and Eric for the above FSSC report. ed.]
Documentation: (Chuck) sends the following:
It's relatively quiet in WB World at the moment. I have been testing GRBanalysis, and it seems to be working okay with the SCons builds, and Nicola assures me that that is the case. However, significant changes have been - and are being - made to GRBanalysis itself and, unfortunately, there are a couple of problems with the released version. Bottom line, I have not been able to successfully complete the analysis following the WB's instructions, and Nicola has asked that I refrain from posting any updates for another week or two.
Pass 7 plans: (Anders) It's slowly moving forward. See the roadmap for details. (Leon) Next step is generating overlay (for both Pass 7 and Pass 8).
Pass 8: (Tracy, Leon) Finishing touches are in progress for adding diagnostic data to overlay. Tracy has made enough progress with global pattern recognition that it's worthwhile to present it at the next meeting (Thursday this week) and perhaps decide whether this approach is worth pursuing further.
Luca Baldini, now joined by Johan, is working on CAL moments analysis. The next topic to be investigated is categorizing backgrounds by looking at periodic triggers.
There will be a workshop in Pisa at the end of May. See the Pass 8 agenda page for more information and a link to to the Workshop page.
GR versions: (Toby) has made a new tag for flux, fixing a bug detected by Riccardo which only affects generation of certain point sources. The code originally worked properly but some other change elsewhere apparently broke it. The fix should not change behavior otherwise. (Heather) will add it to HEAD and also the pass 7 branch.
Windows batch (Heather) There is a new account, fermi-rm, which can submit Windows batch jobs. Previously this could only be done from the glast account (not even from glastrm), and we've been having problems with that on glast06. It turns out the stumbling block for glastrm is that its home directory is on a network drive rather than local disk.
Mac news (Heather) bldmac02 has been converted from leopard to snow leopard.
Gaudi upgrade (Heather) There is progress. She can now generate the required rootmap files on Windows and Linux. GlastSvc is about ready to go. She expects the changes required for certain other packages will be analogous to those already done for GlastSvc, so should go quickly.
Windows and Fred (Heather) It turns out the problem Gloria was having with Fred on Windows 7 was not really due to Windows 7, but to an incomplete ruby installation, missing some extensions. After using the one-click ruby installer, all was well. Workbook documentation has been updated accordingly.
SCons topics (Tom S.) After a couple iterations, he now has a command line installer which works from Europe as well as on this side of the Atlantic.
He changed hoops to build a shared rather than static library. This eliminates build problems caused by declspec declarations on Windows and works perfectly well on Linux also. Next he'll be looking into test program errors in rhel5 64-bit builds.
(Heather) The upgrade to SCons 1.3.0 is done on Windows. Before it can be done for Linux and Mac she needs access to the afs area where it should b installed.
(Joanne) has a slightly-enhanced RMViewer ready to go, but building it leads to a different adventure on each platform. rhel4 and rhel5 are done. Mac is expected to be straightforward unless the move of bldmak02 to Snow Leopard interferes, but Windows will take a bit longer. Before RMViewer can be linked, I'll need to build Qt and MySQL libraries. The ones Navid used for RMViewer development are no longer accessible.
previous | minutes index | <next |