Attendance: Kent, Warren, Lingqin, John, Duncan, Jolien, Patrick, Greg 


Meeting discussion:
------------------

Warren:         explained contents of email.
Kent:           warned about relying on contents of frames to get data
Warren:         would resort to a lookup table if data was available,
otherwise constructred from frames.
Patrick:        how do we get this data out of the frames?
Kent:           Can specify other things than ADC channel in frameAPI and
these are retrieved from the frames and sent on as ilwd.
This has never been tested.
Patrick:        Do we need a special structure for detector information?
Kent:           No need for specialised structure. Major rewrite of code
needed: prefer to continue to use multidimdata.
Duncan:         Will new code be needed for frame header structures?
Kent:           Pretty much yes, most of the code that is written is for ADC
                channels. Don't know if FrDetector structures will come out in
                a format that the wrapper will handle.
Patrick:        Get everybody together with the frame people and 
Kent:           Has Warren verified the frame spec for detector structure?
Warren:         No, got info for lal spec.
John:           LAL spec doesn't match current frame spec.
Kent:           Talk to John Z and Stuart about what will actually in the
                frames.
Kent:           Frame builder needs to be verified, talk to Alex.
                FrameAPI, datacondAPI, wrapperAPI need to be verified.
Patrick:        Have mpi meeting two weeks from now to discuss this.
                Get Alex, (John Z, Stuart), Sam, Albert to join.
                Will get Alan to convene an sccb meeting next week to ratify
                frame spec and then implementation can be discussed in mpi
                meeting.

Kent:           Trying to get system to run at sites. Linux boxes with RH 7.1
                and 2.4.*. No big changes to wrapperAPI or mpiAPI. Tracked
                ocasional job fail down in mpiAPI, added some extra code to
                make it more robust, about 100x better. Will steal some code
                from the a tcl web server to make this more robust.

Greg:           Periodic upper limits discussion on known searches. AEI group
                has Hough transform and Glasgow has heterodyne method, so
                perhaps should write shared objects for existing LAL code from
                these groups. Have small shared object that calls a LAL
                function. Function ready for LAL will be contributed.
Patrick:        Is anyone from Glasgow writing LAL?
Greg:           Have written tested code that needs to be "LALed"
                Could build shared object for AEI groups code. Will probably
                write small shared object anyway. Pulsar group is worried
                about duplication of work.
Patrick:        Important to get a LAL/LDAS periodic source shared object
                quickly.
Greg:           Has been told to help one of these groups get a shared object
                as soon as possible.
Patrick:        Write a simple shared object with existing code. Make contact
                with one of the other groups and get them to join.
Jolien:         Will put Gregs pulsar code into LAL.

Warren:         Not had chance to do much as moving to LLO.  Working on
                writing output to multidimdata to go into frames
Kent:           How big are spectra to go into frames?
John:           Don't know yet.
Kent:           Use database or not?
Warren:         Some frame format that is more useful.
Kent:           Write to FrProc structure in framespec. A subset of frame
                format the is supposed to caputre processed data.

Lingqin:        So far not much done. Will have a meeting to look at
                heirarchical search code. Modify fct engine and shared object.
Patrick:        Time line for useful search code? Heirarchical or not.
Lingquin:       No timeline set. But will try and get work done for E6?

Duncan/Patrick: Nothing done on power or inspiral.

Patrick:        Documentation is a disaster. Have to fix this for the next
                release. August 24th. Code freeze two weeks prior to this.
Kent:           Target day for burst/stochastic is week after labour day. Need
                guidance as to what new stuff needs to be in LDAS.
Patrick:        No solid date for inspiral science mdc. Will inform Kent when
                we have a plan.
Kent:           In favour of mdc, but worried about scheduling.
Patrick:        Will discuss coordination of mdc's with upper limit chairs.
Kent:           E6 3rd - 17th October.


Next meeting:  19 July 2001 at 10am PST
------------