Minutes for the terameeting of June 10th, 2008

Present: Lila, Patrick, Gregory , Fred, Henry Joy, Mathias M. (short period), Yannick.

Please read the summary : Yannick's summary document

Extra actions discussed during the meeting . dotprojects not discussed, so please have a look at them.

WIRCam:

 Patrick will contact Ibata to get information on the way Rodrigo envisioned the sky subtraction in its fields. In particular what data attached to its observing strategy were supposed to be used for the sky subtraction. Note yannick: Rodrigo Ibata will be at the photometric meeting next Friday.

 Patrick will ask Jerome Bouvier what astrometric accuracy he need for its second data set, in particularat the edges of the fields. Patrick need to know whether he can stick to the present data and produce a stack based on this series of images only, or he must merge them with the previous run in order to produce a global astrometric solution with better astrometric accuracy at the edges of the fields that overlap with the previous data set.

 Patrick will explore an astrometric solution with SDSS instead of 2MASS: Scamp crashes with the most recent data sets during the matching when 2MASS is used. The reson for this is unknown at this stage.

 Henry decide to no longer wait for cross-talk corrected images from Loic and move on on his own to get W3 WIRcam data ready quickly.

SPICA2

 Mathias M. will add an option in the menue to allow users to select the path for the output data. This is a request from Yannick and Patrick to avoid that all data of any jobs be copied at the same palce in the disks, to avoid extra file copy on more suitable disk space after a job is finished (e.g. move the weightmaps and ldac to a work area located the the input image location), and to avoid conflict between different jobs running in parrallel.

 Mathias S. will tests the most recent user interface tools developed by Gregory.

 Gregory plan to implement SExtrator this week.

 Gregory will read and comment the Latex document written by Mathias M. The document will then be sent to all for comments. After this step, the document will be sent to Henry to re-shape it from Broken-english to God-Save-The-Queen-english.

Development WIRCam

 Patrick did a presentation of the WIRCam processing steps and his processing tools to Gregory and Mathias M. Mathias and Gregory are now going to implement the WIRCam processing sequence and tools into SPICA2. They plan to start from the .py scripts written by Patrick.

Development (general)

 There is a strong demand from all to add up the basic image operation tool developed by Emmanuel several years ago into the Terapix cluster as a whole. This will allow Terapixian to easily and quickly add/subtract/divide/multiply images with a very simple user interface. Patrick and Gregory expressed their interest to get the code and add new operations , like Log(image), etc...

Hardware

 Regarding the mix5, mix7, efigix disk problem. Fred checked that there is no risk to loose data , provided disks are not in error mode. However, it is preferable to change the disk. Fred will replace efigix first, then mix7 and mix5.

 The mix3 mother board is arrived! mix3 should be back soon.

 Fred is still testing mix11.

 No news about fcix4 . Yannick expressed his concern about this issue and the endless story of fcix4.

 Fred will switch cosmix5 and mix21 is order to get one single raid for cosmix and two for mix21, as it was initially planed.

 There are still problems with Condor: Mathias M. did check it is working. However, Henry's job were stuck to "old" mix's and failed to execute tasks on new nodes. Condor messages for these mix's are "failed for unknown reason". Fred will explore this, though it is not easy to fix something with such messages.

NEXT meeting: Monday June 30 (Yannick in Victoria-Canada from June 25 to 28)