Minutes for the teraminutes of February 5th 2007

Present: Henry, Laurent, Emmanuel, Chiara, Yannick, Mireille, Fred, Mathias (by videoconference), New TERAPIXan, Gregory Semah. Note: the TERAPIX web pages need to be updated with Greg

Emmanuel's summary document

T04 CFHTLS release

 No QF has been been done for the past week after the ups went off. Some confusion arised during the La Rochelle meeting about restarting the processing. No mail was received from the stopped machines. Fred checked that sending mails would work on the machines.
 Fred has volunteered to supervise the operations of release T04 during the coming months. This means restarting jobs which have crashed or gone wrong, and interact promptly with other TERAPIX members in case of problems.
 Laurent explains that config-spica2.pl exists only on nodes with two raids and have both data to process.
 Many images cannot be transferred to the nodes because a problem JC's verifyfits program prevents datatransfer to transfer the data to the nodes. Laurent investigates on this.

Processing of PI data
 Henry will be Mireille's interlocutor for science-related decisions with the processing.
 Barger data is on hold because of the aforementioned problems with datatransfer: only 14 barger images have been QFITSed. Mireille will contact Barger once all the data have been transfered.
 Mireille has shown to Fred examples of exposures for which CCDs exhibit a variation of FWHM stronger than expected. Fred has transmitted the link to Emmanuel to check PSFEx' behaviour.

Soft
 Problems with CSV export in SPICA found by Herve Aussel should be fixed by the end of the week (Laurent did not propagate his corrections to other nodes). Each scripts is duplicated three times (clix (private and public) and ftpix (archive))
 Simple Image Polynomial is a new format for describing higher-order distortions which is being used at astrometry.net, and they would like to have this capability in SCAMP / SWarp. Emmanuel would like to do this, but implementing it is likely to be a lot of work, and it's not clear that this format will be around for very long.

Hardware
 Gregory has an account on the TERAPIX machines.
 Fred looks into the .htaccess mechanism for protecting the intranet to allow Gregory to access it.
 Fred will explain to him how the TERAPIX cluster works and Gregory will write up a document describing everything!
 Mandriva 2007 installed on Greg's machine. Greg will change his screen with that from another computer (the "nice" 21' screen should be in Martin's room).
 pix7 was restarted with the last spare alim
 Olivier Ricou from LRDE will work with the pix1,pix3,pix4 trio to make an web service.

Other matters
 Gregory Semah has started to work at TERAPIX, and is currently in Henry's office.
 He's learning about the FITS image format and astronomical image processing.
 TERAPIX is involved in KIDS/VIKING and we will be processing part of their data, at least in the AstroWISE framework.
 Mathias is still waiting for news about the request to the PNPS for money to buy a machine for processing.

WIRCAM
 Yannick mentions 2 other possibilities to finance a new WIRCAM machine: a machine lent by J.-P. Beaulieu and/or a grant by CSA (decision given very soon).
 Bouvier data has been processed (image Y) 118 GB.
 Chiara shows her problems with WIRCAM data reduction: Loic's crosstalk correction is still causing a lot of trouble in the final stack.
 Willot stack ready but Mr. Willot is on vacation.
 Following a question by Chiara, it is decided that PI data older than one year will be archived; intermediate data products will be deleted after this time. The TERAPIX web pages (Yannick) and the FAQ (Henry) will be updated with this information
 Henry will check Chiara's English before the WIRCAM processing document is made public.
 Chiara will eventually release her processing scripts to the community. Some acknowledgement formula should be prepared for potential users.

After the meeting Laurent demonstrated how the QFs can be cleaned up and restarted.

Next terameeeting: Monday 12th February 10.00 in the council room.