Minutes of the LCG Generator Services project planning meeting
25-May-2007
-----------
People attending:
Jon Butterworth (JB)
Gloria Corti (GC)
Gabriele Cosmo (GCO)
Mikhail Kirsanov (MK)
Michelangelo Mangano (MM)
Pere Mato (PM)
Andreas Morsch (AM)
Witek Pokorski (WP)
Torbjorn Sjostrand (TS)
Alexander Toropin (AT)
People attending remotely:
Paolo Bartalini (PB)
Frank Krauss (FK)
Leif Lonnblad (LL)
Peter Richardson (PR)
Patrick Robbe (PAR)
----------------------------------------------------------------------
Progress report:
The project progress report covering the period of the last six months
(since the previous planning meeting) was presented by WP and was
generally accepted by the stakeholders.
(each workpackage was discussed in details on the slides)
MM asked what was the feedback concerning the new GENSER structure,
whether the usage of GENSER has increased after the re-organization.
WP answered that the three LHC experiments have already moved to the
new GENSER installation (and are using it in production) which
certainly is a positive feedback. JB said that in ATLAS the usage of
GENSER has in fact increased because some generators which were not
present in the old version of GENSER are now installed. Also the fact
that new versions are released with no delay made it possible for the
experiments to rely more on GENSER. In particular ATLAS does not need
to have private releases of some generators (like Jimmy) anymore.
MM pointed out that he should have been contacted concerning the
comparison of D0 data and alpgen that was done by O.Zenin within the
context of the Validation workpackage. WP agreed with that remark and
promissed to make sure that the authors will be informed about any
validation (against data) work that we do.
WP asked the question to the experiments concerning the plans of
moving to HepMC 2. PB said that CMS has already moved (since CMSSW
140) and are happy with it. WP asked whether there are no problems
with persistency and PB answered that no.
JB said ATLAS was certainly planning to move to HepMC 2, most likely
before the end of the year.
GC said LHCb is also interested in moving to HepMC 2, but needs first
to make sure that it will be possible to read all the old events
(generated with HepMC 1) with the new version. PM suggested Generator
Services should help in that move (see planning). WP agreed on that.
TS said that it would be useful to have a presentation by Lynn on the
new developments in HepPDT. WP promissed to organise it.
----------------------------------------------------------------------
Project planning:
The stakeholders agreed on the general proposed planning.
GENSER - will keep on providing generators libraries for
experiments. Possible inclusion of MC@NLO will be re-investigated. PR
supports that.
GC said that the LHCb request concerning Windows binaries is not a
high-priority one but would allow to use some available resources. WP
promissed to provide Windows binaries for 1-2 main generators.
VALIDATION - WP said that with the move to the new GENSER structure
completed even more effort will be put now in validation.
JB fully supports and encourage the collaboration with Rivet
project. WP agrees with that and says that the idea would be to run
Rivet on a regular basis for every generators released within GENSER.
Beta version of Rivet is available and will be, therefore, tried out
by the Generator Services project. (WP said that at some stage Rivet
should be installed within the LCG External software suite)
TS asked a question (to start a discussion) whether the idea is only
to check differences between following versions of the generators or to
tune the generators.
WP said that the Generators Services project would be very interested
in participating/coordinating the tuning of the generators, but the
situation may be difficult as far as unpublished data is concerned.
JB says that the project could play a role in tuning using the
published data while experiments could still do some private tuning
with unpublished data.
TS says that many people are working on tuning and therefore there is
a lot of duplication going on.
It was agreed, therefore, that the Generator Services project should
certainly try to coordinate somehow the work on tuning. First step is
to regularly have monthly meetings devoted to the subject. JB said he
could try to encourage better attendance from the experiments.
JB also pointed out that it would be very beneficial if the authors
participated in the activity by communicating what should be tunned.
HEPMC - CMS has moved to HepMC2 and the two other experiments are
willing to do so in the near future. PM pointed out that the Generator
Services project should help the experiments to do the move. In
particular we could help LHCb to test the backward compatibility as
far as the persistency is concerned.
WP decided to add a new milestone concerning the move to HepMC 2 (to
be completed by the end of 2007).
WP will invite Lynn to give a presentation on the latest developments
in HepPDT.
Question from TS: is Alice planning to participate in the project?
AM: Alice not really interested in using GENSER because they install
all the generators they need by themselves. However, Alice has
presented the tuning and the usage of their generators at several
different occassions. WP will invite AM to give a talk on the Alice
generators at one of the monthly meetings.
----------------------------------------------------------------------
Milestones:
(complete list included on the slides)
- added milestone concerning tests/validation for ATLAS
- added milestone concerning having Rivet running to validate GENSER
generators
- added milestone concerning move to HepMC 2 (and phasing out HepMC 1)
----------------------------------------------------------------------
Final remarks:
PR: authors happy with the way things are going
WP: next planning meeting before the end of 2007