Core Minutes 5/29/2012

• CHEP: (Richard) Among topics of interest to us:

  • Glen Crawford, whom many of us know from SLD days, is now a director of research in the DOE Office of Science. He gave the keynote address.
  • There was a talk by Dell. Cooling and power are the main bottlenecks as they try to increase number of sockets. (There is no longer a race to increase clock speed or number of cores per socket.)
  • Video-conferencing: CERN is going to use a commercial product; will be free to CERN users. They'll drop EVO support by the end of this year. SeeVogh is supposed to be a commercial drop-in replacement for EVO, but we've seen it's not that simple. See Announcements below concerning a test SeeVogh meeting.
  • Data preservation: DESY put their system tests into virtual machines.
  • ROOT: they're replacing cint with a different interpreter, replacing CMT with cmake. ROOT 6 will not support Windows initially.

• ScienceTools: (Jim)

  • gtltcube Justin uncovered a bug in the binning for gtltcube when phi-dependence is enabled. The bug is in healpix::CosineBinner. There is a workaround in the phibins=0 case that was not included in the code to handle the phibins!=0 case. We will probably be able to use the same workaround to fix this bug.

  • igrg Melissa looked at the C implementation of the WMM (World Magnetic Model) code to see if it could be used instead of the f2c'd igrf fortran code. Unfortunately, the interface does not have access to all of the quantities we need (McIlwain L, among others), so we will proceed with f2c-ing the new IGRF code. This does not look like a difficult task.

  • healbix binnin in gtbin The diffuse group has expressed interest in having healpix-based maps that can be displayed using the Aladin software for analyzing large angular scale diffuse features. Discussion can be found here and here.

    Johann has a student who has implemented the binning classes required to make this available in gtbin and we have contacted the FSSC to see if they are ok with these changes being made to the evtbin package. The FSSC will consider this issue at their meeting today.

• FSSC: (Tom S. and Alex) No news.

• Pass7 reprocessing: (Richard for Tom G.) Done through March 31st, which was the goal. We're on schedule. The remainder of the work (making FITS files; catching up with current data) won't take long.

• Pass7 development: (Leon) No news this week.

• Pass8: (Leon, Tracy) Still dealing with remnants of new Gaudi fall-out. The latest problem had to do with auditor variables, some of which disappeared. Tracy managed to make them reappear, but they weren't filled. He now believes he can take care of that issue. Another recent problem had to do with Overlay and converter protocol, since fixed. Tracy notes the Gaudi conversion has held up Pass8 development by 2 or 3 weeks.

• Core talk:

  • Externals and Windows: news from CHEP (Heather) LHC has abandoned Windows support — in particular for Gaudi. This shouldn't affect us much since we won't be upgrading Gaudi any time soon, perhaps never again.

    In a similar vein, ROOT 6 will not be available initially for Windows. Heather plans to upgrade to 5.34 and stay there until ROOT 6 exists for Windows.

  • DB migration: (Heather) Arash is waiting for a final list of users and grants. Heather had been working on such a list, but it was on the laptop that got fried. She will start anew in Confluence.

  • CMT RM (Heather) The CMT RM on Linux is skipping packages that have recently been changed; just doesn't build them. As a first step in debugging, Michael Kuss has added print statements to relevant scripts.

  • Systests (Heather) u17, the systests disk, was about to fill up. Heather will look into clearing some space.

  • Event display progress (Heather) Johann was able to run WIRED.

  • SCons RM (Tom S.) Joanne reported a problem with the creation of GR user releases on Windows; the most recent user release for vc90 dates from March. Since the installers use the user release zip file as a sign that a particular build exists, none of these builds are available from the installers. Tom will look into it. Other than that, things are quiet. [Tom quickly found and fixed the problem. The latest vc90 LATEST — 5673 — has a user release zip file and this build is accessible from the installers].

  • GR vc90 (Joanne) fixed known issues as of last meeting. In particular, the project files generated for Gaudi programs now build their targets successfully. A new issue has arisen: the mechanism in place for installing headers, xml files, etc., is tied to a library project. But not all packages build a library, notably xmlGeoDbs and Gleam do not. She hopes to have a more general mechanism in place in a day or two, but a simple work-around is already available. Alpha testers are invited to try it out. See instructions in Confluence.

• Announcements

  • SeeVogh Test Meeting will be tomorrow (Wednesday), starting at 11:15, shortly after the SCA meeting.

previous

minutes index

next

J. Bogart, Last Modified: