ROOT Team Meeting

→ Europe/Zurich
32/1-A24 (CERN)

32/1-A24

CERN

40
Show room on map
Danilo Piparo (CERN), Jakob Blomer (CERN)

Shift handover
Jack->Jonas R.
Problem with Mac Beta were solved. Builds were green today.

News

Auto-merging to streamline the PR merging process

Jonas H. mentions that
- spurious test failures will likely avoid reaping the benefits of auto-merging
- if something does break, because there was another change to the repository while the tests ran, the person that approved the auto-merge, might not be in the office to fix the issue.
Decision about auto-merging postponed.

GHA Documentation

Jack suggested to unpin the docs Github Actions. Approved.

Code coverage

  • Comment from Jonas H. that current coverage percentage shows branch coverage, which is lower than line coverage and might not be definition we want. Low branch coverage may be caused by error checking code.
  • Philippe mentions that code coverage for the changes of a single PR can also be generated, but takes additional 2h to run.
  • Jack brings up the question on how the coverage would influence the development process.
  • Aaron shares his positive experience with Interop.
  • Lukas mentions that coverage info makes the life of the code reviewer easier.
  • Jack adds that there is the chance of false security if a line is green.

Graphics Meeting

  • Jack suggests to have a clear objective for the meeting since there are many different avenues the discussion could take.
  • Danilo answers that a risk assessment (both in likelihood and impact) would be a great outcome.
  • Jack suggests a risk table prior to having the meeting.
  • Jonas R. mentions 10 year olds slides that mention similar risks, yet things are still working.
  • Danilo mentions related difficulties and challenges from the past.

LLM and AI

  • Lukas mentions that "all" (forum, code, docs) data is openly available, therefore, customizations to existing LLMs may not increase the accuracy.
  • Jack does not see any actions to be taken
  • Vipul agrees to Jack points
  • Jack and Stefan want to avoid AI generated content on the forum.

Compression

  • Jonas H. describes that zlib added a change that trades off compression speed against compression ratio (patterns of three byte are not compressed, due to vectorization.)
  • Jakob asks the question about the goal? Decide which library do we want.
  • Stefan asks if the threshold are too small.

PoW

Stefan mentions that the issues maybe related to the releases that after some time

Meeting summaries and Roundtable

skipped.

There are minutes attached to this event. Show them.
    • 16:00 → 16:01
      Find notetaker 1m
    • 16:01 → 16:05
    • 16:05 → 16:10
      Shift Handover 5m

      Giacomo --> Jonas R

    • 16:10 → 16:15
      Meeting Summaries and Plans 5m

      ROOT Meetings:
      - I/O
      - Statistical Analysis Software
      - PPP
      - Planning / Stakeholders

      Related meetings:
      - ML4EP
      - LIM

    • 16:15 → 16:45
      Topics 30m
      • Liaisons' Reports
    • 16:45 → 17:00
      Round table 15m
    • 17:00 → 17:05
      A.O.B. 5m