Minutes:
Attendance: Warren Anderson, Joe Romano, John Whelan, Greg Mendell, Kent
Blackburn, Duncan Brown, Patrick Brady, Jolien Creighton, Teviet
Creighton, Alan Wiseman.
Actions Items from last week:
----------------------------
Changes to the LALWrapper code. Duncan has not done them since they would
change the prototype of the LALConditionData() routine. I suggested that we
modify the interface layer and wait until right after the next release to make
this change. Too many people are just coming up to speed and I would prefer
not to change the interface under them. Duncan will do this and modify the PR.
Discussion of the Conventions document:
--------------------------------------
The question of scope was raised. The desire was that it would be global to
the LSC; this is not what is in the current document. Alan said that the
the SCCB can make this decision for the collaboration. After some discussion,
the decision was made to change the scope statement to: "The conventions in
this document will apply to all data and software in the LSC and LIGO
Laboratory."
John Whelan suggested that we should multiply the Delta t into the vector. No
other concrete suggestions were forthcoming at this time. Kent did mention
that Peter Shawhan should be brought on board too. It may impact the design
of the database.
Kent also suggested that we could add something like "Where there is overlap, this
document supplants the functionality ....... Disclaimer ..... in areas where
this overlaps with previous documents ......"
Patrick will e-mail Daniel and John telling them that they have been added to
lal-discuss.
Stochastic DSO:
--------------
Warren has started to work on this code having taken over from Joe. The
make-dvi fails, but they dont understand the reason yet.
Patrick will take care of documentation PR's.
Joe asked if there was an example response function with complex entries.
Patrick will try to track down the response function for E2.
One more stochastic thing. LLO-Allegro correlation. It was a combination of
educational and defining who is doing what. Idea is for Siong to write
shared object. Brought him up to date .....
Patrick and John will e-mail Siong Heng .......
CW DSO:
------
Greg has bee trying to contact David Chin but has not heard anything. Not
much progress on that. Still working on how to deal with the amplitude in the
folding problem. Greg will try to get code into LALWrapper. He is worried
about the inputs into some functions: whether they should be frequency series
or vectors.
LDAS:
----
LDAS release is nearby. Functionality will all be there. Kent will announce
it soon. He will be away next week.
Duncan:
rpms will be available soon for all of ldcg. This will effect everything.
This is good. This was requested. Several people wanted rpm's of /ldcg.
They want to mirror ldcg to institutions. Sam is doing that. Joe has made a
request to Stuart to do the same thing there. Stuart is behind it. It might
be a good way to go.
Warren asked if it is one package or multiple. Rpms will be useful for
people who want to have things more useful for a week.
Kent continued to say that coupling the rsync's to the Laboratory and the
sites for stability. Biggest news for the inspiral group. Have put in
changes to the dataconditioning API so should be ready. Wanted to get it
into this release.
The mpiAPI has been better than it has been. Timeout issues with the loop to
receive data through TCPIP. 0.5% of jobs used to fail. That has increased
to 7%. This does not have anything to do with mpiAPI, but has a lot of
impact. There is an html server that will be incorporated into LDAS which
will fix this, they hope.
Other Business:
--------------
Joe asked about John's document about timeSeriesData, in particular including
information about the phase in a heterodyned time series. When will this be
added to LAL? This should be put upward to the SCCB.
Jolien: If we decide we will make major modifications to LAL. Have to freeze
LAL. Etc. It would be nice to do them all together.
Action items:
------------
Duncan will modify LALWrapper to allow Comm in conditionData()
Patrick will e-mail Daniel and John
Patrick will take care of documentation PR's
Patrick and John will e-mail Siong to join the MPI group
Greg will think about putting his code into the CVS archive
Next meeting: Thursday, 28 June 2001 @ 10am PDT
|