Subsystem Scientist Activities (R. Lucas, and all)
C. Wilson: Introduced herself to Kemball. Will meet face to
face during the ASAC meeting or later. Attempting to get caught up
on revised memo #11. Kemball asked her to think about requirement
prioritization.
J. Richer: One meeting was held so far of the telescope
calibration group (~6 involved people). Largely an introduction. A
related item is the work by Pardo on ATM. He was surprised that it
was described more briefly than in the subsystem
agreements. Glendenning noted that the ATM work is one level deeper
and hence internal to the telescope calibration. Glendenning to send
Richer a pointer to the telescope calibration agreement. Language
issues (F90) might need wider software group discussions.
R. Lucas will now compile these reports in advance of the meeting
and append them to the agenda (and the minutes). So please, send
them in due time.
Status of memo 11 review R. Lucas
The review meeting could not be held one month ago. It seems
necssary to have an oral meeting, not just an e-mail review. It will
probably have to be held in September.
Status of Aips++ auditing
(Steve Myers, Francois Viallefond, Koh-Ichiro Morita)
Steve reports that a first pass through all requirements is almost
completed at this point. Nothing is too surprising. A draft document
is built up as the audit proceeds. Both A. Kemball and Tim Cornwell
have seen it but have not reacted yet.
The main concern in the discussion that followed was about the
level of checking that is done (by Wright, Handa, and others): due
to the timescale available, only the availability of tools and
features is checked, based on the available documentation,
e.g. checking that the corresponding tool exists and that the
corresponding interfaces can be brought up. What should be done is
`pushing real data through each function'.
Thus the reporting to the ASAC should clearly mention this point.
The next step should involve data. Indeed this is foreseen in
the benchmarking to be done this fall. As much as possible results
should be available before the PDR in December. Thus we should start
to define benchmark data sets in September.
The most direct way to start this benchmarking process is Phase
III of the AIPS++/IRAM test. Details of how to do this will be
discussed by R. Lucas and A. Kemball.
Considerable progress in Phase I of this test has been
demonstrated when Athol and Kumar visited IRAM last month. Phase II
will occur in late August / early September.
Date of next phone meeting
Due to a conflict with the Software architecture meeting, the next
meeting is postponed by one week:
September 18th 15:00 UT
$Id: 2002-08-14.html,v 1.5 2002/08/16 14:32:03 lucas Exp $
