Difference between revisions of "December 17, 2008 Software"

From GlueXWiki
Jump to: navigation, search
(Agenda)
 
(4 intermediate revisions by 2 users not shown)
Line 5: Line 5:
 
#* Anonymous svn checkouts
 
#* Anonymous svn checkouts
 
#:
 
#:
# Discussion: Formalizing build procedures (Mark I.)
+
# Discussion: [http://www.jlab.org/~marki/halld/releases.pdf Standard Software Versions] (Mark I.)
 
# Too many cdcTruthHits [http://www.jlab.org/~zihlmann/CDC/cdctruthhits_issue.pdf (Beni)]
 
# Too many cdcTruthHits [http://www.jlab.org/~zihlmann/CDC/cdctruthhits_issue.pdf (Beni)]
 
# [[Action Items From Hall-D Software Meetings|Action Items]]
 
# [[Action Items From Hall-D Software Meetings|Action Items]]
Line 30: Line 30:
  
 
==Minutes==
 
==Minutes==
Attendees: ''TBD''
+
Attendees: David L. (chair), Simon T., Alex S., Mark I., Elliott W., Elke A., Beni Z., Herun Y., Jim S., Micah V., Curtis M.
 +
 
 +
===Standard Software Versions===
 +
See Mark's slides above that were used to prime the discussion. There was a general consensus that we should have at least a nightly build and build for each stable release. The "production" build will point to the last stable release. It was suggested that the release notes be posted on the wiki in a way taht makes it easy to scan through them when looking for major changes.
 +
 
 +
There was some discussion of external package management. In particular, things like root, cernlib, and xerces. At a minimum, the release notes for a release should contain a list of the versions for the relevant packages that were known to compile on a specific platform. The issue of binary distribution was deferred until a later meeting.
 +
 
 +
===Too many cdcTruthHits===
 +
Beni presented evidence of multiple truth hits in the CDC. These are events in which more than 1 truth hit can occur for the same straw. Mark first noticed this while working on tracking code. The double hits do not occur in the raw CDC hits, only in the truth hits. This issue was understood by Richard (not at the meeting) who has since, addressed it.
  
 
==[[Action Items From Hall-D Software Meetings|Action items]]==
 
==[[Action Items From Hall-D Software Meetings|Action items]]==
# Update the start counter geometry to reflect the current 40 paddle design (Seems D.)
+
# Follow up on anonymous svn checkouts (David L.)
# Follow up in anonymous svn checkouts (David L.)
+
# Formalize/document software release creation procedure (David L., Mark I.)
# Turn on Moeller and Bhabha scattering by default in hdgeant (Alex S.)
+
# Update documentation on Wiki for turning on/off processes in GEANT (Alex S.)
+
  
 
==Next Meeting==
 
==Next Meeting==
 
''TBD''
 
''TBD''

Latest revision as of 10:53, 14 January 2009

Agenda

  1. Review minutes from December 3, 2008 Software meeting
  2. Announcements
    • Anonymous svn checkouts
  3. Discussion: Standard Software Versions (Mark I.)
  4. Too many cdcTruthHits (Beni)
  5. Action Items

Participating

Location

The meeting will be at 14:00pm in Cebaf Center F326.

Telephone

To connect by telephone: 1.) dial:

800-377-8846 : US
888-276-7715 : Canada
302-709-8424 : International

2.) enter participant code: 39527048# (remember the "#")

Video Conferencing

We will attempt an EVO connection for this meeting. We have a reasonable expectation that this will work.


Minutes

Attendees: David L. (chair), Simon T., Alex S., Mark I., Elliott W., Elke A., Beni Z., Herun Y., Jim S., Micah V., Curtis M.

Standard Software Versions

See Mark's slides above that were used to prime the discussion. There was a general consensus that we should have at least a nightly build and build for each stable release. The "production" build will point to the last stable release. It was suggested that the release notes be posted on the wiki in a way taht makes it easy to scan through them when looking for major changes.

There was some discussion of external package management. In particular, things like root, cernlib, and xerces. At a minimum, the release notes for a release should contain a list of the versions for the relevant packages that were known to compile on a specific platform. The issue of binary distribution was deferred until a later meeting.

Too many cdcTruthHits

Beni presented evidence of multiple truth hits in the CDC. These are events in which more than 1 truth hit can occur for the same straw. Mark first noticed this while working on tracking code. The double hits do not occur in the raw CDC hits, only in the truth hits. This issue was understood by Richard (not at the meeting) who has since, addressed it.

Action items

  1. Follow up on anonymous svn checkouts (David L.)
  2. Formalize/document software release creation procedure (David L., Mark I.)

Next Meeting

TBD