Agenda and Minutes, 31 May 2006
Agenda: Wed, 31 May 2006
Minutes
Attendance:
Brian,
Kevin,
Patrick,
Brian O',
Keith Th.,
Keith R.,
Igor,
Stuart,
Peter,
Nick,
Sergey,
Kent,
Ed,
John,
Ben,
Albert
Calibration:
Brian explained that there is a release of ASCII files. These are
validated. Frames cannot be read using standard tools.
Two ways forward. He would prefer to have one standard frame format
that is agreed to and can be released by the calibration team
Brian believes ASCII files are consistent with spec.
Albert: Factorize this problem. More frame generation outside
calibration team.
Ed asked about Virgo. Brian responded will use h(t).
Brian indicated that h(t) will be frames will be validated by
calibration team and its review committee with help from data
analysis working groups
Keith Th: Will map functionality to matlab, but it may take some
time.
Patrick reiterated the old question about frame users:
Pulsar: h(t)
Burst: ASCII - frame.
Stochastic: ASCII - Frame
Inspiral: frame.
Position was: Support one frame format. But Brian does need one
person who does LAL|LALApps for verification.
Stuart: ?????? Yep.
Patrick summarized why we have arrived at Jolien's suggested frame
format.
Igor: What is the reason to change frames?
Patrick: Because API and frame format were tied together. We want
to maintain that and have a simple interface that can be used by
multiple people without having to know the guts too well and without
having to validate use of calibration every time.
h(t) roll-off in low frequencies. How is that information going to
be propagated? Patrick asked about document about h(t)?
Albert: Stagnated. There should be better QA. Need to get some more
documentation. They have posted some of these things. He will be
restarting this process with the calibration team.
DASWG Meeting to discuss h(t)? Brian would like to have a big
meeting of all interested parties. Does not care about the venue.
Kent: How to make sure frame files versions can be accessed when
querying datafind? This needs some enhancements of LSCDataFind.
Albert: If not familiar with calibration of h(t), cal lines disappear.
Patrick: How to proceed?
- Brian will vet new frames. Discussion with calibration team,
release.
- Patrick wold prefer to have document before release. Jolien
will need to update the calibration document to make this
happen.
- Fr & FrContrib: Peter would release tools there.
- Keith Thorne will try to mimic whatever is done with the c-code.
Stuart: Asked if bounced by Benoit..... Wonder how they distribute
information about whitening. Patrick to e-mail Benoit once we are
closer.
DMT:
No report.
Glue:
Updated # semaphores in DBZ at LHO and LLO. Ben reported this.
LAL|LALApps:
No report.
LDAS:
tclGlobus 1.1.1 released. RPMs available.
LDR:
Core functionality of metadata service. Still on track for major
release next month.
LIGOTools:
Pending update to frames package. Re-released segments,
segwizerd, metaio.
Onasys:
Web stuff released.
MetApps:
Not much activity. Static data issue to wrap.
AOB:
Nick: Is there a plan to transition to NumPy in place of numarray.
Patrick explained that he'd been hesitant because the documentation
is not free. Nick has ported his frvect package to use NumPy which
provides some pressure to transition by the end of the summer.
Nick explained that it works well with metplotlib.....
Patrick to get word back to Nick & others next week about this.
Stuart asked about lscsoft repository. Patrick explained that Adam
Miller will be taking this on. He should start to come up to speed
this week and will be able to do basic release management by next
week.
$Id: 060607.html,v 1.1 2006/06/13 05:21:01 patrick Exp $