Notes from the LArSoft Stakeholders / Partners and Project Meeting April 2, 2014
--------------------------------------------------------------------------------
Experiments
o LBNE (Tom)
Believe they are behind in nightly builds. Need someone
from expt the experiment to maintain it.
There is a cron job running. Work with Tom after the meeting
to get it running again. The LArSoft build can be used as a template.
Approve of the ROI changes to recob::Wire
-- There needs to be validation of the change: compare results of a
job with full data range to a job that uses ROIs.
-- Noted that first induction plane in uBooNE could be the real test
case.
-- This change may make it unnecessary to pursue GPU work for the
purpose of deconvolution.
GPU work: Amir Sarbin, UT Arlington
o uBoone
Nothing. Can't build prof.
Will work with Erica after the meeting to fix this.
o ArgoNeuT
Some discussion of migrating to new system in order to take advantage
of some of the new toolsLimited use case
and effort.
Not urgent.
o LAriat : nothing
Optimization
o Identified a number of issues related to uB MCC4
Fixed.
o Working on LBNE now
Release procedures
Project report (Ruth)
o Maintaining the project WBS.
Currently in excel, will migrate it to Redmine
Contains both short and long term planning/tracking
Monthy reports to upper CS management. About 5 minutes.
Will circulate the reports to this group.
o Extended mrb functionality
-- support for other shells: considerable progress, but not completed
Will require some changes to mrb that Lynn must do. She is subscribed
this week due to work deferred for the sake of LArSoft, so will miss
the target completion date of 4/4 (tomorrow)
-- Support for dependency checking
Previous estimate was 6 FTE weeks. Lynn now believes it will be less
Target completion is June 1. That date is still ok.
o Nightly builds and cvmfs
Noted that some experiment nightly builds are not working.
The project wants to keep all nightly builds in scope. We need a name from
each experiment of the person who will own the nightly build.
Lariat: Brian
uB: Herb will identify
LBNE: Tom will identify. There are people in the environment available
o Integration system
Alpha release due June 15.
Still need to specify what must be included in the alpha release.
Erica and Ruth will work on this.
Proposing a workshop at end of June to populate the alpha system
Expt feedback is positive. Idea would be to lock ourselves in a room
until we have fully implemented tests running in the alpha system.
Requires test code, something that will check the result, integration
into Jenkins (assuming that will be used for the alpha system)
We need to talk about how to employ students to the benefit of this
project. Forty to 50 people coming for the summer to work on IF
There are minutes attached to this event.
Show them.