Minutes for the Terameeting of Friday 03 September 2004

Present: everyone.

SPICA (Laurent:)

 Everything seems to work, Laurent is making some final tests and hopefully we can start Spica on all the stacks from Dougados in the near future. D1g* stack produced.

 Henry identified a problem with swarping images using headers produced by astrometrix/spica: the flux scale keyword FSCALE generated by Photometrix is NOT the same as the fluxscale keyword read by default by swarp (FLXSCALE). Mireille found Phot_c Phot_k were not used correctly in Dougados data and confirmed the FSCALE issue found by Henry.

SCAMP (Emmanuel, Yannick:)

 Scamp "block" solved but there is now another problem with crashes when attempting to compute the astrometric solution for very large numbers of images (for example in the W1+D1+prew1+all filters field).

 USNO-B reference catalogue works better now once the problem with dense source catalogues has been resolved and this will be the default reference catalogue for future version of Scamp and in Spica.

 work in progress: XML output, proper motions, and possibly a better estimation of positional errors in sextractor

QualityFITS (Fred):

 Fred will add a plot of the PSF computed in many regions over the entire STACKED field (previously in single image mode this worked CCD-by-CCD).

Veronique Cayatte PI Data (Mireille):

 Mireille has stacked all the Veronique Cayatte data in all filters. Data have been distributed and preliminary feed back from Cayatte are positive. Need further input from Cayatte, in case problems detected.Yannick has also prepared some colour images.

Skywatcher (Jean-Christophe:)

 Many optimisations have been made and it is now possible to draw all of the survey in around 30 seconds. Limiting factor seems to be SQL requests. Perhaps mysql can be statically linked to optimise performance (maybe gain 10%-30%). Is skywatcher optimized for very large surveys which have many pointings? For example in DENIS/2MASS there are 1,000,000.

 Fred suggested several changes that may speed up Skywatcher.

 Jean-Christophe will investigate various methods to optimise the software: at the same time he will continue work in integrating skywatcher in the Terapix pipeline environment.

AVO Credits (Delphine, Yannick):

 All of the AVO credits have to be used up by the 31st October! This includes Gerard's salary. One solution from Yannick is to pay Gerard 80% of his salary on a mission to Manchester (!). The remaining will have to be spent before the 30th of October, around 8000 eur to spend!

Dbterapix (Gerard):

 Gerard is working with Laurent to understand the lock condition observed

 Gerard's priority is now DBterapix: we would like to have this well documented before he leaves at the end of March.

Data transfer (Jean-Christophe):

 Many images and files in the outgoing fibre channel array were marked by CADC as 'invalid' Jean-Christophe is trying to understand why.

Hardware (Fred, Gilles):

 Fibre channel array arrived today, but without the necessary cable, which should arrive on Monday.

 Need to think about a way to implement automatic shutdown of all the machines if there is a power failure: need to make some tests, Laurent will be warned well in advance as these tests will involve completely stopping the cluster.

Panorapix (Fred):

 Fred processed the Panorapix code with "doxygen" which produces automatic documentation of all function calls and subroutines.

Offices

Gerard will move into the office of Gilles Jean-Christophe will move to Laurent's office. Chiara will move to Henry's office (which will become the visitor's office). Fred keeps its second desk as a guest office. Anthony will be in room 210 (replacing Ludovic who is moving by Sept. 10).

Next Tera-Meeting: Friday Sept. 10. 2004