[Halld-offline] Offline Software Meeting Minutes, October 26, 2016

[Mark Ito]

Folks,

Find the minutes below and at 
https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_October_26,_2016#Minutes 
.

   -- Mark

________________________________________


    GlueX Offline Meeting, October 26, 2016, Minutes

Present:

  * *CMU*: Naomi Jarvis, Curtis Meyer
  * *JLab*: Alexander Austregesilo, Thomas Britton, Brad Cannon, Eugene
    Chudakov, Paul Eugenio, Mark Ito (chair), David Lawrence, Hrachya
    Marukyan, Paul Mattione, Nathan Sparks, Simon Taylor, Beni Zihlmann

There is a recording of this meeting <https://bluejeans.com/s/nlR5L/> on 
the BlueJeans site.


      Announcements

 1. *HDPM 0.3.1 released*. Nathan reviewed his announcement of the new
    release
    <https://mailman.jlab.org/pipermail/halld-offline/2016-October/002519.html>.
    He has re-written the system in Go <https://golang.org/>. The system
    can be distributed as compiled binaries now; no Julia
    <http://julialang.org/> interpreter needed.
 2. *sim-recon 2.7.0 released*. A wide variety of changes are included
    in this release. This release was built using upgrades to Xerces-C,
    CCDB, and HDDS. Mark referred us to the release notes
    <https://github.com/JeffersonLab/sim-recon/releases/tag/2.7.0> for
    more information.


      Memory Leak

Beni reported a memory leak that only revealed itself when many raw data 
files are processed by a single job. [Added in press: Beni opened an 
issue <https://github.com/JeffersonLab/sim-recon/issues/610> on the 
GitHub site describing the problem].


      Review of minutes from the last meeting

We discussed items from the minutes of the meeting on October 12 
<https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_October_12,_2016#Minutes>. 


  * Thomas has been re-writing portions of the Offline Monitoring
    website in Javascript to enhance the dynamism of the user
    experience. The idea is to reduce the number of click-and-wait
    actions necessary to get to desired information. Documentation to come.
  * David sent an email to the group
    <https://mailman.jlab.org/pipermail/halld-offline/2016-October/002495.html>
    about the downside of storing complete HDDS files as the solution to
    a run-dependent geometry, based on his experience with a similar
    scheme with the online translation tables. Changes to different
    detector components occur at different times, leading to a
    multiplication of the number of versions needed. To some extent the
    problem with the geometry is ameliorated by having different
    subsystems specified in different files, but the problem remains.


      Software Review 4 Talk Walk-Through

The bulk of the meeting was devoted to a quick walk-through of the three 
talks from Hall D 
<https://halldweb.jlab.org/wiki/index.php/Software_Review_4_-_All_Hall_-_October_28> 
to be given at the upcoming software review.


      Launches

There have been some issues with calibration; there are some tweaks to 
be done for the reconstruction launch. The goal is to start by Friday.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20161104/eab6933c/attachment.html>