Phone Meeting 2005-02-15 15:30 UT

Participants: PrebenGrosbol, PeterLinde, BillCotton, PerryGreenfield, RichardHook, DougTody, KariVasko, NicolaCaon

Agenda and minutes

1. General information
   • Annual OPTICON report to EU
     • We have submitted the input from our Network to the OPTICON annual report to EU. The report for the full OPTICON is 120+ pages and can be found at: ftp://ftp.roe.ac.uk/pub/jkd/opticon/draft_report_140205.doc but only less than 5 pages are relate for our Network.
   • Discussions on possible ALMA associated member
     • PrebenGrosbol reported that he had a discussion with the ALMA software IPT management who was interested in having a more direct link to the our discussions on future software environments. They will shortly propose a person for this role.
   • Additions to agenda
     • On request by BillCotton and PrebenGrosbol two items were added to the architecture discussion (see below)
2. Next face-to-face meeting
   • Indicate you preference on PollMeetingDates
     • At this moment 4 of 5 people indicated a preference to June 23/24 for the next face-to-face meeting. Thus, rooms have now been reserved in ESO. We would first have to make a final decision approx. 2 months before the meeting.
3. Architectural concept
   • Report on Radionet activities
     • BillCotton summarized the Radionet activities to build a data calibration framework. It would partly be based on classic AIPS tasks but also provide access to other tasks through a Python top layer. In general terms, the architecture follows the concepts suggested by DougTody.
   • Architectural concept paper
     • There has only been one formal reply to the architectural concept document namely by Markus Dolensky (AVO). The main issues were answered by DougTody in his reply e-mail.
   • Discussion on Component-container paper by DougTody (see ArchitectureAndFrameworks)
     • PrebenGrosbol raise the issue that 'tasks' were defined as stateless while stateful components in several cases, such as for interaction with computational tasks, would be useful. DougTody clarified that although tasks were regarded as stateless other types of components could very well be stateful.
     • PeterLinde found that it would be very useful soon to be more concrete rather to continue too far with a more abstract design.
     • In order to make things more concrete, the next issues to address would be the parameter model and the minimum container-component interface.
4. Actions
   • Status of open actions (see ActionList)
   • Summary of new actions (see below)
5. Time for next phone meeting
   • Suggestion: 2005-03-22 15:30 UT (16:30 CET)
6. Any other business

Due to my cold, my notes were worse than normal so please add what I have forgotten and correct my errors.

-- PrebenGrosbol - 21 Feb 2005
to top