Present: Joanne Bogart, Toby Burnett, Xin Chen, Heather Kelly, Michael Kuss, Leon Rochester, Tracy Usher, Karl Young
GlastRelease: See Toby's status page and the Gaudi 11 conversion page. To summarize:
All sandbox packages have been merged back into the standard copies. The sandbox itself can be deleted as soon as we're sure we don't need it anymore.
The new container package IExternal is fully operational.
Linux binaries for a couple of the external packages still need to be made available for download by non-SLAC users (Alex?). SLAC users can just set $GLAST_EXT to /afs/slac.stanford.edu/g/glast/ground/externalPackages/rh72_gcc2953
There are still some problems, most likely minor, building the new stuff on each platform.
Tracy (in collaboration with Norm Graf) has made a CMT package for building G4. Here are some instructions. It works with both VC6 and VC7. Need to try it out on Linux also. This makes it possible to see the code easily when debugging and also allows us to build with CLHEP 1.8.0 (rather than 1.7.5 that the pre-built G4 used), consistent with what we're using everywhere else.
Nightly builds, Release Manager: (Karl) Alex has been doing some refurbishing to get everything working with the new GlastRelease/IExternal arrangement. Karl is essentially done getting System tests going; ready to be integrated with Release Manager.
Science Tools: (Toby, Karl) Toby has made a new container package with, so far, just one child package, CALDBinterface. Will the container package force anyone who wants to just use some piece of Science Tools check out the whole thing? No, it is possible to check out individual contained packages. Meanwhile, Karl and Jim Chiang have been working on a design of classes needed for Science Tools. Karl will send it along to Toby for feedback.
ROOT stuff: (Heather) She has been working on an update of root2idl. Unfortunately it is still necessary to use RioGrande, a reduced, I/O-only extract of ROOT, rather than ROOT itself. root2idl and the packages it (and it alone) needs, RioGrande and CINT, perhaps belong in a new container.
Calibration infrastructure: (Joanne) Everything needed for a reasonably comprehensive test of the system (but only for a single trivial type of data, in XML format) is in place. So far the test program gets through the configure and initialize stages "successfully" (without crashing), but the toy algorithm fails to find calibration data. See the web page for more complete status.
previous | minutes index | next |
J. Bogart Last Modified: 04-Aug-2004 15:40:02 -0700