Teraminutes for meeting of Friday, 27th of February 2004 Present: everyone, except Fred, who's in Brazil.

Scamp (Emmanuel) deep/ultra wide are being used as test fields. Major bug corrected. Results are good, produces 'Mario-like' solutions. Problem however with the ultra-wide; very challenging field to produce astrometric solutions for, because the field-to-field overlaps are very small (and the problem with the deep is that the dither offsets inside in the fields are small). Scamp demonstrated with wide data. Works, but astrometric solution not good enough for swarping! Very large residuals at the field boundaries. Discussion about proper motions in the catalogues: can this be an important effect? perhaps there are numerical problems: ATLAS/num rec don't give the same matrix solution. Check with A2029 data, to make direct comparison with results from Mario's software. Mireille, Henry, Laurent will provide the LDAC files of A2029 to Emmanuel for these checks. Need to check the effect of adding rejection of outliers to the fit (not implemented currently).

Data Sophie Maurogordato (Gilles). Swarping and astrometric calibration completed, but photometric calibration still has to be done. Four cfh12k pointings, total size several gigs. Will provide cut data. Be careful about the zero-points! Mireille will explain the method to Gilles.

DBclient (Gerard). Memory leaks have been fixed. consumption of memory is now optimal: Dbclient consumes 5MB of ram more than the size of the input catalogues. Now start work on tools to perform remote execution of SExtractor (via web interface) on images from Aladdin. Visit to Strasbourg next Friday for this.

ICD document (Jean-Christophe) release date in *stacked images* will be defined by the date of the LAST observation, and this information will be injected into the stacked image. Laurent can do it! keyword to add is RELEASE_DATE. Software version numbers should be added as well. This request comes from CADC to make data more homogeneous with other datasets currently stored at the CADC. Fred should write the software version in the binary tables he creates. Other matter: we should let CADC know how much data they should expect from us: Jean-Christophe, Laurent and Yannick will do this estimate.
 Need to get skyprobe data , note only png or jpg image of skyprobe values. Jean-Christophe and Yannick will check with CADC and CFHT that these meta-data are available.

Data processing for the wide,deep,B2029: TO ALL: PLEASE REPLACE A2029 by B2029 in order to avoid confusion and possible wrong identification by AVO/ADS tools. Mireille will provide masks for Henry for wide data in WCS; convert WCS masks to pixel masks and then run the quality assessment tasks (mostly galaxy autocorrelation at large scales). Mireille showed tests with NARROW=N option in wide data. Henry showed latest version of quality assessment tests on d4 ( 2hrs) and d3 ( 4hrs integration). Mireille will run all W3 data sets with Narrow-No option. Yannick will provide new mask for the r-band data of B2029.

COSMOS i* data (JC). data now at TERAPIX, QualityFITS will be run on the data.

Spica (Laurent) not much to Report, still waiting for Scamp. Discussion about release date of stacks.

SkyWatcher/sky location of the field on the sky request arrived from Canadians. Jean-Christophe is now moving of this project.

Defectix (Anthony) added sky subtraction, added reduction of dynamic range of the images, decreases learning time. configuration files done. Need to define a formate for masks.

Opteron status (Emmanuel). Reassuringly, Emmanuel has found many messages on the net from people who have *exactly the same difficulty* as us (instability with raid file systems on the opterons), with the same hardware configuration. In summary, no known configuration seems to works correctly with opterons + mult-tera disk config. Several points:

 1. First some set of our problems seems to be come from the linux kernel; need to upgrade to version 2.6.
 2. It seems likely we will have problems with *all* 3ware cards on Opterons, even when we change the motherboards.
 3. new 3ware firmware out, new bios, need to test this update.
 4. Waiting for Airma test motherboard from Mr. Landat.

Cosmix (Gilles): accounts opened. Check with Stephane Colombi the account needed for his PhD students and/or postdocs.

ACI/MDA: in progress.

Next tera-meeting: Friday 5th March 2004 at 14:30 .