- There are many sources of project documentation (Docushare, Confluence, lsst.io, CLO, Slack, LCRs, etc etc).
- Docushare is the official system of record.
- But Docushare also contains myriad obsolete or unnecessary documents; shouldn't all simply be transferred to Operations.
- Expectation that Operations will use Docushare, but there remains work to do to sort this out.
- Software documentation, etc, belongs on lsst.io.
- No concrete action emerging from this.
- Wil O'Mullane has asked Jeff Kantor to organize a review of networking.
- Likely carried out by telecon.
- Discussion of (alternate) standard visits, and interaction with satellite trails.
- Ongoing conversation with AURA leadership on coronavirus; more news when we have it.
- Leanne Guy requested not to make an announcement re the algorithms workshop until Victor has met with AURA leadership today.
- Request from leadership to make sure that anybody transitioning from construction into commissioning is clear what their roles are.
- DOE review of Camera finances in May (the DOE equivalent of the EVMS audit).
- Camera schedule contingency dropping; now at ~30 days.
- Review the list of action items from last week's meeting.
- J. Swinbank
- F. Mueller
Test Plans due in the next 45 days
Milestones due in the next 45 days
- Deferred from last week's meeting.
- RHL will need to perform rapid processing in Chile (“rapid”, as distinct from “prompt”).
- Some of that processing will need to be reproduced at NCSA.
- e.g. need to be able to generate quick-look ISR images both on the mountain and at NCSA.
- And the results of this processing should then be available as permanent data products.
- We should expect that we don't know what types of processing will be necessary early in commissioning; need a general approach to running pipelines.
- Possible to imagine this as a prototype of prompt processing, but it's not clear how that meshes with expected schedules.
- Kian-Tat Lim identifies three issues:
- Does the generalization of processing fit into the existing prompt processing framework?
- Do we need a new framework?
- How long does it take to build such a new framework?
- Consider e.g. deploying OCPS at NCSA, or generalized prompt processing.
- We note that LDF is already busy, and there may be an opportunity cost to rolling out a service like this.
- Agreed that:
- Robert Lupton will provide use cases, including soliciting them from the Commissioning Team;
- Based on those use cases, Kian-Tat Lim & Unknown User (mbutler) will consider how much effort would be involved in standing up such a service.
- The DMLT will then decide whether this overrides other priorities.
- Robert Lupton — mediating with the Commissioning Team, provide Kian-Tat Lim & Unknown User (mbutler) with a list of example use cases for rapid processing at the LDF.
- Kian-Tat Lim & Unknown User (mbutler) — propose a design and schedule for deploying a rapid processing system at the LDF.
- Wil O'Mullane en route to Spain (visa renewal; mostly on leave this week in Ireland and UK - talk in Bath on Friday).