Present: Ursula Berthon, Joanne Bogart, Monica Brigida, James Chiang, Seth Digel, Richard Dubois, Riccardo Giannitrapani, Berrie Giebels, Nicola Giglietto, Navid Golpayegani, Traudl Hansl-Kozanecka, Heather Kelly, Michael Kuss, Julie McEnery, Leon Rochester, Alex Schlessinger
gcc 3.2 move, or not: (Alex, Richard, others) We have been using gcc 2.95.3 in an unsupported configuration (that is, combination of compiler and system libraries is not the default for any extant Redhat version) which is therefore difficult to maintain and distribute. We would be in the same boat with gcc 3.2/ Redhat 7.x.
It would be preferable to change both compilers and OS version together (gcc 3.2 is the native compiler for Redhat 9) but SLAC is not yet providing Redhat 9 machines in its batch farm and we don't know when the switch will be made. For one thing, Babar is not ready to move. Also the Redhat 9 kernel currently in use at SLAC has some problem with nfs. (And although batch farms at SLAC and Lyon continue to use Redhat 7, support for Redhat 7.x is waning at some institutions for individual laptops and desktops.)
Why even contemplate moving to gcc 3.2 under these circumstances? At present, Riccardo has only been able to build Corba on Linux with gcc 3.2. Corba is required to use FRED on Gleam events. In particular, it would be nice to have a FRED display server working for DC1.
Upshot is that we will defer the move for the time being; Riccardo will see about building Corba with gcc 2.95.3.
ACD Ribbons: (Heather) Code changes are needed in several packages to support sensitive ribbons. Most of it is now done; for example, changes required to do ribbon noise. Work on the last of these packages is in progress. In particular, an additional method is needed to compute active distance for ribbons. This is especially messy since ribbons in one orientation (fixed Y) are broken into 5 segments across the top, each of which is at a different Z. Might choose to approximate with a single Z value.
ToT: (Leon) Bill can use ToT information to detect upward-going stopping protons. Work to understand and correct raw ToT values is by no means complete, but significant progress has been made (using muon data and simulation) particularly in correcting for effects due to the projected angle between the trajectory and the long axis of the strip. See these pictures and plots for details. Leon's work to date will be incorporated into DC1. Further corrections can be applied post-Recon, when the n-tuple is generated. For a more complete understanding of effects Leon will be looking at photon and electron data.
FT1, FT2 (Jim) Johann has implemented a computation of FT1 quantities, insofar as they are well-defined, in meritAlg. Jim hopes to get relevant people together soon to get the list finalized. As for FT2, exposureAlg outputs an ascii file with the right quantities. Goodi will be used to do the translation to FITS.
Classification trees (Richard for Bill) Bill reports that, looking at all-gammas generated with v3r3p4, Toby's implementation of his Insightful Miner classification trees behaves like the original.
System tests: (Richard for Matt) Matt has been in the process of rationalizing job options and such. Richard would like to see a full run of System Tests for v3r3p4; we will undoubtedly be making another tag or two soon.
ROOT converter: (Ursula) With help from Heather, Ursula has been able to create a suitable Gaudi skeleton for this project; work on the guts may now commence.
At SLAC: Navid is here for the week He has been working on an optimized Linux build. Some small code changes were needed to get certain test programs to build.
Heather will be here Thursday and Friday of this week (after driving up from San Diego with small fellow traveler).
Our user nfs space has filled up again, a mere week after clean-up. We might need some kind of quota system or a shorter expiration period (currently a year).
At Lyon: (Berrie) He can run about 1/3 as many events in a batch job there as compared to the SLAC farm.
previous | minutes index | next |
J. Bogart Last Modified: 01-Jun-2010 15:47:15 -0700