HSF Weekly Meeting #94, June 1, 2017
Present: Benedikt Hegner, Torre Wenaus, Dario Menasce, Patricia Mendez, Michel Jouvin, Pete Elmer, Mark Neubauer, David Lange, Liz Sexton-Kennedy, Philippe Canal, Daniel Elvira, Fons Rademakers
News, general matters
- Michel - participating in sw license task force of Belle II
HEP analysis ecosystem workshop outcomes
-
Report, actions, next steps
-
Workshop live notes and (notes towards a) report draft
-
The report draft materials will be split off to a separate document but haven’t been yet. Torre will do it the initial draft in the coming 2 days.
-
Summary in EP/SFT group meeting: https://indico.cern.ch/event/642127/
-
Agreement that it was a very good and well organized workshop
-
One comment was that we should document how the organization was (successfully!) done: Pete agrees to do it, a good template for the future.
-
CWP
-
Annecy workshop June 26-30 https://indico.cern.ch/event/613093/overview
-
There’s a very good bus connection from Geneva: https://www.flixbus.com/. To be added to Indico.
-
45 currently registered. Early bird registration ended. Need to make more publicity.
-
Overlaps the simu workshop at CERN Monday-Tuesday: push to Wednesday-Thursday the simulation discussions at the workshop: very small overlap but need to plan a room where the people can remotely participate to the CERN meeting.
-
Most WGs have some good materials already but we need to ensure that we have something closer to a draft for every/most WG so that the workshop allows to flush out the things.
-
Define a stub of what should be presented by each WG during the first day of the workshop
-
Need to decide the tool to use for the final, consolidated CWP. Candidates are GoogleDocs, github with markdown source, overleaf (not recommended for broad collaboration based on experience), github with latex source….
- Most important until we have a document is developing it collaboratively, which argues for googledoc…. then use pandoc to convert a late draft or V1 of the doc and put it in github for subsequent evolution?
Activity updates
Licensing
-
From Benedikt: a *preliminary* and simplified (Compared to the full release I suppressed all isolated packages and build-time only dependencies. The SWAN related packages I dropped as well for now. We can tackle them independently) summary of the license situation in the LCG stack https://indico.cern.ch/event/643785/attachments/1468728/2271690/lcgstack_licenses.png?from_preview=1. Confirms that most of the “GPL pollution” coming from the generators. We can influence things from our community but the more difficult problem is GSL that we cannot influence and is also used by generator…
-
Contact finally established with Fastjet developers: proposal is a meeting on June 14 at CERN as the 3 main developers should be there. But conflicting with ATLAS SW&C week in Valencia… Anyway seems to be a good occasion to start the process: Liz proposes to summarize the problem in a few slides and circulate them between the people involved in licensing at HSF and in experiments.
-
Turn generator into a plugin: GPL says that it doesn’t change the licensing problem but not all lawyers agree on this interpretation. May be something to investigate in more details if it is the only approach left…
-
Also a problem identified with HepMC. V2 is a GPLv2 but feedback that it would not be a problem to change it to LGPL (original author from FNAL and inline with FNAL licensing policy)… v3 is GPLv3 and situation not clear: Witek contacted.
-
Another problem identified is MySQL. License not very well defined: changed several times.
AOB
- Meeting next week.