Jan 10/24 PionLT/KaonLT GitHub Meeting Notes
                 --------------------------------------------
                              (Notes by GH)

This was a special meeting to discuss the future structure of the
hallc_replay_lt software

Present
-------
Regina - Garth Huber, Ali Usman, Alicia Postuma, Nacer Hamdi, Nathan Heinrich,
   Vijay Kumar  
CUA - Richard Trotta, Casey Morean
FIU - Pete Markowitz

Preliminary item
----------------
- Richard found that for some of the newer scripts, you wil have to specify a
  more recent python version, i.e. 3.3.5 vs 2.2.1

hallc_replay_lt discussion
--------------------------
a) UTIL_BATCH script should be updated
- used by both pi/K analyses
  - should segregate analysis scripts similar to how they are done in
    hallc_replay, don't want changes by one experiment to affect another
  - also, different channels in same experiment (e.g. pi/K/p in KaonLT) need
    different filenames, not just one set of scripts for KaonLT and another for
    PionLT

- Ali: suggest to segregate codes to make codes easier to follow/understand

- Nathan: Production branch is 544 commits ahead of Master.  It would be good
  to update Master, then it would be easier to do the segregation

- Casey: we want Master to be the branch we point people too when they are
  starting analysis

- *ACTION NEEDED*: sych fork on Master right now
  - everyone should push to Master and update it


b) what to do with defunct/archived code?
- some defunct branches should be archived

- Nathan: old snapshots correspond to various milestones
  - no need to delete them, not sure if they're working now, but they were working
    at the time the snapshots were made

- Casey: suggest to archive to Mass Storage System (MSS), as otherwise the
  utility of the old snapshots is lost

- Richard: it would be good to get all your scripts working in a container and
  save it.  That way all essential information is preserved.

- Casey: suggests naming according to Release Number, and adding tags on which
  version number of other software it relies on
 - concerned that certain branches will disappear with time and then this
   information will be lost

c) Documentation
- Nacer: need a message on top of every file on what inputs and outputs are,
  what it does
- Richard has a useful emacs script that automatically puts author and date at
  the top of every new file.  But the essential comment info is empty and needs
  to be filled in later.

Summary of Action Items
-----------------------
1) ltsep_analysis_2022 should be merged with Master and labeled with release
   tags

2) remove symlinks, some are left over from online analysis
  - getting rid of things that don't matter makes it easier for new people
  - archive old software

3) need to go through scripts, find what's used now, and make sure they work
  for everyone


Next Meeting
------------
- the one thing that was not discussed was a task list, timeline for
  accomplishing these tasks
- GH suggests another meeting in about a month, after Nathan and Alicia return
  from JLab on Feb 8