Teraminutes of November 27th 2006 Present: Chiara, Yannick, Henry, Emmanuel, Fred Absent: Mireille, Laurent.

Emmanuel's summary document

Notes on Emmanuel's summary:

WIRCAM processing
 Jean-Gabriel Cuby will try to get the raw J data and reprocess it himself.
 Yannick has suggested that Loic Albert comes to TERAPIX for two weeks to work on the preprocessing. Unfortunately Loic is not available before March. The preprocessing at WIRCAM CFHT is not progressing because Loic has a lot of other duties. Another possibility is that Chiara goes to Hawaii. The problem has been communicated to the board.
 WIRCAM photometry calibration. The meeting with Herve was quite useful. It's clear that the problem is that the magnitude range in which the calibration can be carried out is quite narrow (because there are not many faint sources and at the bright end there are problems with saturation and non-linearity.) Herve will recalibrate the bh36 data with his method. Still waiting for the results.
 Technical note on the WIRCAM processing: the sky subtraction on mix9 becomes progressively longer (7 minutes becomes half an hour) as the processing continues; this happens bizzarely enough on Saturday morning.

Mireille/Emmanuel/Mathias: PI Processing
 The new version of SCAMP produces better results than the old version, but there are some fields which still pose problems. 2MASS is better (lower astrometric residuals) in crowded fields.
 Mathias will continue processing (all swarps).
 Chiara and Henry mentioned that scamp can crash at the end of the processing, possibly due to problems generating some of the check-plots.
 Chiara will teach Mireille how to update the P.I. status pages (dl:12/08)

Other matters
 Emmanuel would like more information in Ganglia on mix9 to help in diagnosing problems with processing->assigned to Fred (dl:12/01)
 Henry found a bug in swarp which seems to be happening with the static library on mix9, possibly due to an incompatibility with the threads library and that on Kiravix->may disappear while moving to icc?

Other matters

 The University has confirmed we can hire a CDD until the concours is open once more in one year. We cannot contact candidates directly ourselves.
 Emmanuel's webservice stage-candidate has withdrawn his candidature
 In Leiden Emmanuel explained the situation with panorapix; we really need a skilled, professional openGL engineer. Yannick may have some money to pay an engineer to continue the project. However VST is delayed. (1 December 2007 is still first light date, but not certain)
 mini-power cut; some of the machines which were not restarted which were not on UPS. Nothing unusual happened on the other machines which were on UPS.
 The only order still to be sent is that for the storage machines (750gb disks each, 16 tera storage total).
 RTRA meeting 29th 'Physique des deux Infinis'; large amount of money available for Ile-de-France / IAP. Possibility to use this money for a new network connection for IAP/TERAPIX for example.
 QF out still to be transferred.

Dot project update
 W1-u extended; Henry will check out the astrometric errors in the new stacks compared to the old.
 Fred we will see Christophe Gobet this week to try to fix the problem of flipping breakers when machines are re-started (dl:12/01).
 Fred will make a 'demo' on the successful shutdown of the system.
 Henry will see with Stephane Arnouts and JCC about flux conservation in MEGACAM pre-processing and what protocol has been adopted (dl:12/01).

Next Terameeting: Monday 4th December 2006