GlueX Software Meeting, December 6, 2021

From GlueXWiki
Jump to: navigation, search

GlueX Software Meeting
Monday, December 6, 2021
9:30 am EST
BlueJeans: 968 592 007

Agenda

  1. Announcements
    1. New version set: 5.1.0
    2. first step toward a global filesystem for Gluex
    3. Transitioning to Zoom
    4. New version set for simulating 2019-11 data
  2. Review of Minutes from the Last Software Meeting (all)
  3. Review of minutes from the last HDGeant4 meeting (all)
  4. Report from the November 4th SciComp Meeting (Mark)
  5. FAQ of the Fortnight: How do I get notified of changes in a repository on GitHub?
  6. Upcoming Computing Review (Mark)
  7. Documentation Upgrade (Beni)
  8. Review of recent issues and pull requests:
    1. halld_recon
    2. halld_sim
    3. CCDB
    4. RCDB
    5. MCwrapper
    6. gluex_root_analysis
  9. Review of recent discussion on the GlueX Software Help List (all)
  10. Action Item Review (all)

Minutes

Present: Alex Austregesilo, Edmundo Barriga, Thomas Britton, Sean Dobbs, Sergey Furletov, Mark Ito (chair), Igal Jaegle, Richard Jones, Simon Taylor, Jon Zarling, Beni Zihlmann

There is a recording of this meeting. Log into the BlueJeans site first to gain access (use your JLab credentials).

Announcements

  1. New version set: 5.1.0 This version set was released on October 29. Mark is preparing a new version set and now that Pull Request #605, "Remove incorrect logic in TRUTH photon construction," submitted by Sean, has been merged by Alex, a new tag of halld_recon can be applied and added to the new set.
  2. first step toward a global filesystem for Gluex. Richard led a discussion of how we can use StashCache to create, maintain, and analyze a global file system. See his email for details.
    • The OSG is moving away from user X509 certificates and towards tokens for user authentication. This (and other issues) have caused a stoppage in job submission to the OSG over the past few weeks
      • The stoppage was confirmed by Thomas. "We are on the cutting edge of OSG use," he reports.
    • We should leverage efforts from the larger community on data management. StashCache is an example of an opportunity to do so.
    • Igal is interested in using this for PrimEx and SRC/CT.
    • One leading candidate for using this is for doing analysis launches on the OSG. This requires having our REST data out there on StashCache.
    • Mark: We (aka he) need education on StashCache. Perhaps Richard can give a pedagogically oriented presentation at some point.
    • Transitioning to Zoom Mark put a wiki page together on the topic and invited us to contribute content.
    • New version set for simulating 2019-11 data There will be a new version sets coming out soon to pick up the fix from Sean mentioned above (Pull Request #605) as applied to the recon-launch compatible builds (2017-01, 2018-01, 2018-08, and 2019-11).

Review of Minutes from the Last Software Meeting

We went over the minutes from the meeting on October 25th.

  • Thomas reported that the node formerly known as scosg16 is being rebuilt as scosg16-dev and will serve as a test platform for changes to OSG-related software. This avoids bringing in changes, untested, on the production Submit Host (now scosg20).
  • CentOS Stream looks to be the choice (not yet final) for a future upgrade of the JLab Farm.
  • On Package Structure, Mark has concluded that the project to render halld_sim and hdgeant4 independent of halld_recon by splitting out the pieces of code used by all three is not easily done and will require coding of interface routines, i.e., a significant refactoring effort. He has succeeded in doing the split in on a technical basis, but the extracted code represents an unacceptably large fraction of the various detector reconstruction packages. The split code, though not suitable for general use, may guide us if we want to pursue this change in the future.

Review of Minutes from the Last HDGeant4 Meeting

We went over the minutes from the meeting on November 29.

Hao Li has marked Issue #195, "Crash adding daughters on RHEL/CentOS 8" as resolved.

Report from the November 4th SciComp Meeting

Mark promised a report soon. He knows y'all can't wait. ;-)

FAQ of the Fortnight: How do I get notified of changes in a repository on GitHub?

We took a quick look at the link provided in the answer to this FAQ.

Upcoming Computing Review

We went over a few items linked from the Computing Review wiki page.

  • There will be a meeting to show slides this afternoon. All presenters in the review will present what they have.
  • A few of the slides that Mark prepared were flashed.
  • The pre-review handout document now runs to about 16 pages.

Documentation Upgrade

Beni presented his thought on how we should improve software documentation. See his slides for all of the details. Some points (for emphasis):

  • User's are well-positioned to comment on the documentation and should be encouraged to do so.
  • An overview presentation of the data analysis flow, from reconstruction to analysis to ROOT analysis is lacking.
  • Richard: there are additional valuable resources for software information:
    • The HOWTO wiki page.
    • Recordings of past Tutorials/Workfests
  • Beni proposed forming a task force
    • Those present unanimously agreed that Beni should be the chair.
      • Richard remarked that "we are all 100% behind you, Beni."
    • Mark volunteered to provide advice to the new chair.
    • Beni will call an organizational meeting.