Minutes of the meeting held on the 13.12.2000 and on the 23.01.2001
Present : Giuseppe Bagliesi, Vincent Lemaitre, Dirk Duellmann,
Andreas Pfeiffer, Guenther Dissertori
- Dirk has given a short status report on Objectivity:
Version 6.0 is there, also for Linux Redhat 6.1 and 6.2. This
version runs with the new compiler, so in principle the applications
and libraries have to be recompiled.
New features: Mainly on the lockserver side. It is now possible to
declare single databases as "read-only", which should reduce locks.
Could also be interesting for us. The actual release will happen
sometimes in January, with the new compiler as part of the release.
The IT database group will spend next year most of its time in the
evaluation of a new product by ORACLE, which offers a database with
C++ binding. The ESPRESSO development is frozen, since there is not
enough manpower.
Guenther gave some feedback about his first tries with LIZARD,
especially what concerns the physical and logical organization of
histogram files in the database.
- Vincent has shown the first full analysis for the single
W selection done in ALPHA++. The work has been done together with
his summer student Maria Perez Reale and his students in Belgium,
Olivier van der Aa and Christophe Delaere. The have implemented many
tools such as electron, muon and tau identification, updated
the jet finder (now use a jet finder written for CMS, ORCA). Some
suggestions were given to solve the problem of links to tracks
from different classes.
They have also implemented locks, in particular recursive locks.
A detailed description of all this will soon appear in an ALEPH note.
The comparison with the identical analysis in ALPHA is perfect!
Next : They will merge their code with new implementations by Rick
(see below), then they will do performance and leak checks, and finally
they will attack the problem of vertex finding.
Vincent will probably give a shorter version of this talk for a
larger ALEPH public or in an ANAPHE meeting in the near future.
- Guenther has summarized a work done by Rick. In short: Up to
now there was a wrapping of the memory BOS structure of ALPHA. This is
now changed, all the ALPHA banks are converted into C++ objects, so
the dependency on FORTRAN is weaker than ever. The PS version of
the talk is found here.
- NOTE: the talks by Vincent and Rick have been repeated
on the 23.1.2001 for a larger public. Vincents & collab. work
is summarized in this Aleph Note (ALEPH 2000-093, SOFTW 2000-004).
Guenther Dissertori
Wed 13 01:50:38 MET DST Dec 2000
GD