Minutes for the terameeting of 13th april 2006

Present: Chiara, Henry, Laurent, Anis, Yannick, Emmanuel, Fred

Status PI data (Laurent, Yannick)
 Data PI Annie Robin 205 (04AF09) Qfits-ed/validated, 169 (05AF14) to be processed through QF, but still locked at CADC. Need email from Annie to CADC to give authorisation to Terapix
 Data Trentham processed by Mireille, relased and ftp-ed/
 PI data from two of the three Qrunids Korean( 04AT05 04BT03 05AT01, including names 05AT01-1 and 05AT01-2) have been processed. The last one needs email from Lihwai Lin to CADC to give authorisation to Terapix.
 Yannick will grade 05Af14 and 05AT05 once Qfits-ed

Data processing (Laurent, Yannick)
 Henry will be re-added to the mailing list so he will receive updates for the ongoing processing and the status of the new Wide W3 images.
 The T0003 VW wuill start once W3 new ready. About 25 days of processing.
 400 L02/L05 images have been transferred. Yannick will grade them once Qfits-ed.

Spica for Dummies (Laurent, Yannick)
 Laurent will prepare a cookbook and will teach to Chiara, Mireille, Yannick and Henry how to run it (operation mode)
 He will also teach Fred high-level SPICA handling and design
 The idea is to have several Terapix-ian beeing able to run / handle SPICA when Laurent is not available.

Spica-new (Laurent)
 no progress

WIRCAM (Chiara, Emmanuel, Yannick, Henry)
 Cosmos stacks with the 'correct' photometry is on the way.
 There seems to be a 5% error in the flat-field produce by Loic
 Several uncertainties on the photometric calibration/accuracy over the whole Cosmos field discussed. Some are related to the pre-processing and photometric calibrations processes at CFHT, other may be related to the way the intercalibration is handled in Scamp. Emmanuel will look at this issue.

Eye, weightwatcher (Chiara, Emmanuel)
 megacam.ret and wircam.ret now calculated
 should allow better rejection of cosmic rays in megacam images. Fred will use them in Qfits.
 Chiara and Emmanuel fixed bugs that hampered accurate handling of overlapped polygons, or of very large numbers of polygon
 Still a bug for very large number of polygon-segments in mask file
 big speedup in weightwatcher polygon parsing code

Scamp, SWarp (Emmanuel)
 There are still some pending issues: the Scamp plots on Qfits-out show an offset of 0.05-0.10" arc-second that is not understood. SWARP used with Schlegel maps offset dust images. The offset depend on (RA,DEC).

Departure of Jean-Christophe (Jean-Christophe, Laurent, Fred, Yannick)
 Before leaving, Jean-Christophe will explain to Laurent and Fred how SNOOPIX and DataTransfer work and will explain to Yannick how Skywatcher works.
 The Qfits-webservive will be made public before April 29. Yannick will try it with WIFI images prior to give a green light to JC.

Panorapix-new (Anis, Emmanuel)
 An interface opengl has been constructed which allows image display
 Anis thinks (99.99% Condifence Level!...) he can produce a working interface which will allow displaying images using shaders for the 3th of May, with both luminance and contrast handling running

Hardware (Fred, Emmanuel, Yannick)
 Four power supplies have arrived
 3 graphic board beeing bought by begining of next week
 Regarding new nodes, several options discussed, Fred will go ahead for
 2 bi-pro dual core processors 5TB, 16GB (mix8/mix9)
 1 fcix-like storage bay
 1 bi-pro dual core 5TB, 8GB (from EFIGI)
 new 3ware controller (24 ports) allows more disks than before, so in that case 3500 euros from CNES will be used to buy disks
 Need to buy a UPS 5-KVA
 a rack is already available
 It then envisioned to add up to the cluster mix7, the 2 bi-pro dual core (16GB, 5TB). Since mix7 diks must be empty for SPICA, all mix7 data will be transferred to mix1/mix2/mix5 prior to move mix7 to the SPICA cluster.
 mix5: the motherboard and processors have been changed but the machine crashes! we will go back to the original configuration

Basic image analysis/procesing tools, stage EPITA (Anthony, Emmanuel, Fred, Henry, Chiara, yannick)
 There is a possibility to have a student for three months to work on some programming tasks.
 The idea is to propose a 1-year EPITA student to develop a basic environement to hanlde simple image processing tools similar to imarith and imstat.
 need first to check whther tools like ftools works.

Other matters (Emmanuel, Fred, Henry, Yannick, Anis, Chiara, Laurent)
 some discussion about the best OS to put on the cluster, Emmanuel asks if some new solutions have been developed. Fred propose to test and challenge FEDORA core against GENTOO on mix8/mix9
 Forum terapix.iap.fr suggested by Emmanuel, Fred will investigate and will install a tool by the end of April.
 poste IE will be available beginning of december
 A heated discussion followed concerning office space, with a proposition being made to Henry which he promised to consider.