Root planning Meeting
Present: The root team, marco, andrea, bill (R), valeri (R), Benedikt, Pere
ACTION:
o Fons: add a PyROOT category for bugs
o Fons: provide a recipe to enable explicitly components (and only those which are selected) to avoid problems like the one of posgre
o Vassil, Marco:
- The TNet issue reported during last meeting was fixed 2 days after the beta release.
- The problem was in the pch generation.
o Valeri:
- Expct to be able to move on with work with the forthcoming nightly with latest 5 release.
o Marco:
- Back to reporting bugs and reproducers.
- Status of LHCb looks better than before: still quite many bugs in the tests.
- Nice reproducer for a funny PyRoot behaviour which is behind more than one failure when PyAlgs are in the game.
- A bug cannot be labelled as part of the PyROOT category (see actions)
- PyROOT is fundamental for LHCb. We need to solve PyROOT issues to move forward.
- Preparing the new Gaudi release which will work for 5 and 6 releases of ROOT.
- Autodiscovery: root autodiscovers available components, for example posgre. This breaks relocatability (you can build on a machine where XYZ is present and then move onto another where XYZ is not available).
o Bill:
- Running unit tests in ROOT
- Loading of dictionaries in CMS is "sloppy". For the moment no action for ROOT.
- Inline ASM is the main blocker
- Conditions area: in 2 weeks will be probably compiling
o Andrea
- Released COOL 5.32.13 (mentioned by Valeri)
- Base class methods not visible in derived classes
Philippe: not seen because of the presence of an enum type. These classes have difficulties in being seen by the system. A workaround is available.
- Overall, things are getting clearer and more items are closed than opened
- The exceptions are going to be a problem. Tests will continue to fail.