HSF Weekly Meeting #144, 13 September, 2018

News, general matters

LHCC Presentation:
- https://indico.cern.ch/event/685793/
- Comments were directed to:
  - How much ROOT progress in the I/O layer?
  - When packaging WG would reach conclusions?
  - Are other FAs in supporting software (in the wake of the IRIS-HEP news)?
    - New WGs are important in defining concrete work plans.
    - LHCC has been asked to offer general support for work plans, that helps this to happen.
  - LHCC have not yet set a date for the review of HL-LHC (may not be before next summer).
COMCHA (Spanish organisation for tackling COMputing CHAllenges) Presentation:
- https://indico.ific.uv.es/event/3438/
- Presentation was well received - organisers are keen to keep in touch with us and work together.
IRIS-HEP was funded!
ACAT call for abstracts and registration is out (https://indico.cern.ch/event/708041/).
Connecting the Dots call for abstracts and registration is out (https://indico.cern.ch/e/ctdwit2019).

Formats discussion (tech list) was very useful.
- RISE extension for Juypter notebooks seems ideal for live presentations.
Dedicate one of our next meetings to training activities?
- Yes, to be followed offline with Sudhir.

Next meeting, September 19.
- Report from Hadrien Grasland about his experience using Spack to build ACTS.

Workshop at ANL arranged for discuss I/O for exascale. Recommended re-investigating OpenMP. The HPC community have invested a lot into OpenMP in recent years and report that it now has much better support for tasks (which was the main feature that led HEP to choose TBB in the past). Could be a topic for a future meeting/workshop (HSF Software Forum).

Group released its final CWP draft. Please review if you are interested in this topic.

Booked in a meeting on Spark use in LSST for 5 December.
A few slots still available this year
- https://indico.cern.ch/category/10392/
- The Framework OpenMP talk would be ideal here.
Next meeting is “HEP.TrkX: Deep Learning Tracking” on 26 September.
- Graeme may be tied up in the ATLAS IO Review - volunteers to chair the meeting? David L should be able to do this.

PyHEP 2018 workshop presentations: most of them now on Zenodo, with a DOI attributed, see https://zenodo.org/communities/pyhep2018.
Should advertise Gitter and resources more widely (hsf-forum).
Still trying to find a neutrino community representative.

Community nominations now received (thanks!).
Need to write a mandate.
Vava wanted to make sure the Reconstruction group covers also software triggers and real-time processing.
- This was in the CWP (strongly!), so yes.
- What about the other group’s names? Are they too abbreviated?
How to proceed when we have more than 3 people nominated?
- How do we make sure we get a good balance, reflecting the whole community?
  - Elections have been skewed in the past.
  - Collaborations should have significant input.
- Do we want to have terms for the conveners so there can be rotation?
  - 1 year renewable agreed.
We should call again for nominations once the mandates are available.

Indico ready.
Tweaked the announcement a little.
- It was felt that aspects of how we use generators could also bring significant gains and should be discussed.
Dates of 26-27 November at CERN confirmed.
Should be ready now to circulate widely (please look if you can).

General Matters and Roadmap
- Feedback from CSBS review was pretty good. List of comments was circulated to the coordination group.
  - CSBS review comments (commentable by everyone, writable by HSF-coordination).
  - Try to finish this by the end of next week.
  - Please take a look and comment as needed.
Publication status for Individual WG Papers
- Software Trigger and Event Reconstruction
  - Still needs to go to HSF repo.
- Data Organisation, Management and Access
  - Michel has been chasing Bo.
- Data and Software Preservation
  - Mike H gave us some LaTeX sources ages ago, now finally being added into github.
- Visualization
  - See above.
- Facilities and Distributed Computing
  - Ian ways should be done in next month. Some editorial decisions to be made regarding contradictory comments.
- Conditions Access
  - Paul has changed institute, but this is high on his priority list.

HSF Logo in vector format - reminder of action on Benedikt.
Next HSF/WLCG workshop:
- Workshop dates fixed for 18-22 March 2019.
- Offer from JLab.
- Offer from BNL to host at Stony Brook.
- JLab looks cheaper for registration ($200 vs. ~$350) and has cheaper accommodation options (staying in Stony Brook looks expensive).
  - Pere, Michel supported JLab, easy accommodation.
- Nobody with objection to choose JLab at the meeting. Final check to be done with I. Collier.
We wrote an HSF letter of intent to collaborate with an IREU proposal from David Lange and Pete Elmer (funds US students to come to CERN, working with Openlab in Big Data and Machine Learning).