Date

June 10, 2014 13:30-14:30 PDT

< previous (2014-05-29) | next (2014-06-24) >

Attendees

 

Kian-Tat Lim
  •  
Mario Juric
  • (for first part)
Jacek Becla
  •  
Unknown User (ciardi)
  •  
Andrew Connolly
  •  
Gregory Dubois-Felsmann
  •  
Unknown User (mfreemon)
  •  
Donald Petravick

Unknown User (rlambert)

Robert Lupton

Unknown User (xiuqin)
  •  
Unknown User (robyn)
  • (lurking)
Unknown User (shaw)
  • (lurking)
Russell Owen
  • (lurking)

 

Goals

  • Confirmation of weekly meeting time.

  • Clarification of roles of SAT, TCT, DMLT, DMPM/DMPS.

  • Action requests: DM-784, DM-785, DM-811, DM-830, DM-831

  • Q&A, if needed, about DM-721 current hardware and needs.

  • DM-31 drill-down into data interfaces and transfer.

Discussion Items

TimeItemWhoNotes
2 minWeekly meetingKTL
  • Proposal is Thursday 10 AM PT, but that is impossible for Xiuqin.
  • New proposals are Tuesday 12-2 if Andy can be convinced to change or Monday 11-1 if Don can be convinced.
  • Next meeting the week of June 23.
2 minRolesKTL
  • Clarification of distinctions between teams with nearly the same membership.
  • TCT has a reserved time on the first Thursday of the month, but typically votes by E-mail.
5 minAction Requestgpdf
  • DM-784 - Getting issue details... STATUS
  • Packages can be named by their authors; they must get approval from their local gitolite admin to install them under LSST/DMS.  The SAT will review all new package names at the next meeting and may request renaming, splitting, or merging then or at any later time.  The SAT will prepare guidelines for the gitolite admins.  The SAT will periodically review the dependency tree and package names, perhaps during a release retrospective.
10 minAction RequestJeff
  • DM-785 - Getting issue details... STATUS
  • Jeff's proposal seems reasonable; no objections.
1 minAction RequestRussell
  • DM-811 - Getting issue details... STATUS
  • No objections; the SAT recommends that the TCT accept this change.
5 minAction RequestRobyn
  • DM-831 - Getting issue details... STATUS
  • No objections; the SAT recommends that the TCT accept this change.  Creating a more specific identifier is suggested to avoid collisions.  In the case of "filter", "filterName" may be appropriate; for filter objects, something like "filterObj" might be appropriate, but more discussion may be needed.
  • Robyn is looking into adding coding standards checking to buildbot.
  • Russell suggests that we prepare configurations for common editors that would check some of the standards.
10 minAction RequestJim
  • DM-830 - Getting issue details... STATUS
  • The sense of the SAT is that underscore ("_") is the better choice.
  • No translation is needed at any point.
  • Distinguishes from VO datasets that are "." hierarchies.
  • Is similar to package naming hierarchy.
2 minDevel/Integration HardwareKTL
  • DM-721 - Getting issue details... STATUS
  • Any questions?
30 min Design Deep-DiveKTL
  • DM-31 - Getting issue details... STATUS
  • Process of developing deep-dive page needs to be more open.
  • Add link to Confluence page in ticket.
  • Level of detail seems reasonable.
  • Kian-Tat Lim will send out message to lsst-data.
  • Use JIRA issues and links for TBDs? Gregory Dubois-Felsmann will check on installation of the Tony Johnson macro for generating action items.
  • DM-33 - Getting issue details... STATUS will be next.
  • Kian-Tat Lim will resolve all issues, either as Done or as In Review by TCT.
  • Kian-Tat Lim will add issues for preparing package naming guidelines and preparing editor syntax-checking configurations.
  • Kian-Tat Lim will start a discussion on lsst-data about the best names for variables containing filter object instances.
  • Kian-Tat Lim will send out message to lsst-data about design deep-dives.
  • Gregory Dubois-Felsmann will check on installation of the Tony Johnson macro for generating action items.
  • Kian-Tat Lim will prepare at least an outline for DM-33.