ROOT 6 planning
Present: Rolf, Marcin, Peter
Axel: news
o 6-02-04 released
o Working on memory usage, in particular case in which creating a TClass entry from ROOT pcm provokes parsing even if it should not
o MCJIT: updted llvm and clang also for ROOT, cling tests pass, ROOTtest runs and underlines an issue, the recursive jitting. Working on it, will ask for help once all tests on our side will not arr
o Realistic to have MCJIT it by the end of may, probably there in March for 6.03/4.
Marcin
o CMT Atlas wrapper is missing from LCGCMT 73, it's 20minutes work, asked yesterday to have it
o Full reco chain working with ROOT6 - 2 releases of Athena one with R5 and one with R6.
o Include path trouble again: there is a full local path somewhere that ROOT could not automatically solve.
o Success rate of the tests is the same of R5 and R6 builds
o Now the RSS burden is 350MB. There is room to go down at least to 175MB. ROOT will work to get *at least* to that level
Bill
o Still waiting for a buid with 6.02.03
o Tried the BigPCH: the procedure was working fine but the RSS was not reduced. Will try with the 6.02.03 and see.
o No idea of the timescale of the Release Validation
o Saga of the python script checking checksums: it's not a stopper, actually the main reason to go to 6.02.03.