............................... Date: Sat, 20 Feb 1999 12:28:53 Who: Tim Conrow Subject: Last chance for input to DA (prelaunch to do list) Original file: /home/tab/jjl.t ............................... From listmaster Sat Feb 20 12:29 PST 1999 Date: Sat, 20 Feb 1999 12:28:53 -0800 From: Tim Conrow Organization: IPAC/Caltech To: WIRE mail list , Terry Herter , Gordon Stacey , Paul Graf , bts@ipac.caltech.edu, mww@ipac.caltech.edu CC: tim@ipac.caltech.edu Subject: Last chance for input to DA (prelaunch to do list) --------------EF0CCC7424F1A9C87D2582EE Content-Type: text/plain; charset="us-ascii" X-Sun-Content-Length: 1234 Attached is the text from some of my vugraphs at the last ST meeting where I showed my list of deficiencies and my pre-launch to-do list. It is now the last minute. All decision are being made with or without benefit of ST input starting immediately. Much progress has been made on many of the issues presented, but your comments are still worth while. Some decisions underway: - See previous email about upcoming additions to the source list columns. - Re coadder outlier rejection: Absent instructions from the ST on this I've decided for purposes of speed and convenience to do full, three pass outlier rejection on the segment level and turn it off for multi-segment and area coadds. Since the 3rd pass actually blanks out the outliers in the frames the later multi-seg coadds will still benefit. The more complex possibility of doing three pass rejection at the multi-seg level needs alot more discussion, which there isn't time for before launch. These and other smaller changes will take effect at the next and putatively the last pre-launch pipeline delivery, coming up pretty damn soon. All other issues will be resolved on a similar time scale. -- -- Tim Conrow tim@ipac.caltech.edu work:626-397-9578 fax:626-397-7354 o DA Status: Known or Possible Deficiencies - No tools exist for identifying moving or variable sources - Reference documentation of pipeline and tools inadequate - Final decision on multi-pass segment analysis not made - Median BG subtraction untried on real sky. Always expect surprises from the IR cirrus - Algorithm for identifying new dead/flakey pixels may be simplistic. - Reporting of new bad pixels is weak - Method for handling surprising levels of stray light not codified. Do we need specialized tools? - DAOPhot is a bit finicky and will doubtless require tuning in flight. More serious action is not out of the question - Experience with position refinement very limited. Response to low reference star density highly uncertain - Bandmerger behavior in areas of high source density is unclear - No procedure identified for updating source lists in DBMS - Assignment of DA area number from SO area ID is insufficiently robust - Tools for monitoring survey progress are primitive o DA Status: Pre-launch To-Do List - Must be completed before launch . Make assignment of area number more robust and recoverable . Test new data privacy features and recent pipeline changes . Track down and clear median 12um BG estimation error . Complete position refinement and coadder upgrades . Develop absolute quality grade and accept/reject criteria . Continue tool and pipeline reference documents . Codify a daily QA task plan for scientists and engineers . Finish pipeline product RTB . Perform H/K TLM verification from recent ops test data . Complete backup plan for software and analysis data . Confirm H/K TLM monitoring tools meet requirements . Implement tiered access on web page . Add ability to assign terminal data state to WireObs page . Perform load testing of WIRE subnet . Upgrade DA delivery process to include reference data - Important, but could slip into ops . Develop easier to use and view survey progress tools . Write tools to assess post-facto constraint violations . Complete Wdaophot debugging . Upgrade source list contents . Refine plan for loading source tables to DBMS . Get auto-startup of Pipeline working and train Rosalie . Complete tools for estimating C&R from repeated observations . Test Input running as two parallel processes . Test alternative web page access through cable modem . Prepare backup ftp site (disabled) . Implement recommended security improvements . Continue improving independence of WIRE subnet . Re-estimate disk usage for ST, AIs and standard products . Expand DBMS space, /wire/font and /wire/well . Codify rules for use of analysis disk space . Complete various usability upgrades to web page . Finalize DAOPhot parameters . Arrange for instrument-only BG subtraction for AIs