To and from Pisa: Eric Charles has returned from his sojourn. Leon will stay over for about 3 months immediately following the Collaboration Meeting, helping the Pisans get involved with Pass 8 work. But first he has to figure out what to do about the cats.
Heather back soon: She already managed to attend this meeting, but the official return date is Feb. 1. Fortunately Rose is a mellow baby.
ScienceTools: (Jim) See this week's brief report in Confluence
FSSC: (Eric) Between holidays and the AAS meeting not much has happened lately. (Richard) How did the tutorial go? (Julie) heard it was very successful. Participants had taken the time to bone up on their own first. Feedback will help to improve the documentation.
Documentation: (Chuck) is very glad to be back, with both lungs inflated. He expects to start out 2010 concentrating on pointlike and the pipeline.
GR:
Pass 7 (Anders) Philippe's Vela study of a month ago, shows discrepencies, particularly at low energies, which are still not understood.
The Data-Clean class is currently of no use for Pass 7 since one of the cuts uses a value tuned for Pass 6 which is not appropriate for Pass 7.
Pass 8 (Tracy) There is a new GR release (v18r0p0) which will serve as a baseline for Pass 8 development. Expect some updates in the next few days incorporating work on cluster axis. It involves changes to the TDS which, as is often the case, affect several packages. (Leon) There is one problem with the baseline release concerning new — that is, newly incorporated — ACD code affecting Windows only. Eric Charles is looking for a place to debug. The UW terminal server might do but Leon has heard rumors of its imminent demise. (Toby) No, it would be nice to upgrade to a more powerful machine, but there are no plans to pull the plug on the current installation without a replacement.
(Tracy) There will be a
Pass 8 meeting
this next Thursday morning. All interested parties are welcome to attend.
See also this
overview of
the project. There is a new Pass 8 mailing list, p8reconlist.
You may sign up in the
usual place.
Ditching rhel3 Not discussed at today's meeting, but responses to a recent email from Heather indicate that we're on target for end of January/beginning of February.
Python file installation (Toby) SCons dumps all python files into a single install directory, making it difficult to avoid name clashes and not making use of normal python methods (modules) for dealing with this. One could install complete modules in a subdirectory without having to add to PYTHONPATH. (Joanne) was under the impression that the current arrangement had been requested, but it should not be difficult to implement something along the lines Toby describes. We should discuss it offline.
GR and SCons (Joanne) As of last meeting, there were a couple remaining issues on her plate: obf reorganization and job options. obf reorganization is done for Linux. With minor enhancements the scheme (python script plus per-flight-build input file describing libraries and includes to be copied) should work for Windows as well.
Some progress has been made with job options but there is still a way to go and some care will be required to avoid breaking CMT builds as the new arrangement is implemented.
ASP and SCons (Jim) has run into a show-stopper which he is so far unable to resolve. SCons builds of swig-wrapped shareables are not handling derived exception classes properly. The comparable CMT build has no such problems.
previous | minutes index | next |