User Tools

Site Tools


difx:d202wishes

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

difx:d202wishes [2015/10/21 10:08] (current)
Line 1: Line 1:
  
 +====DiFX 2.0.2 Wish List====
 +
 +This list assumes DiFX 2.0.2 is to be released relative soon (a few months) after DiFX 2.0.1
 +
 +  * <​del>​Bug:​ when nfftchan is more than xmacstridelength in pulsar mode, only the first xmacstridelength channels appear properly populated</​del>​
 +  * <​del>​Bug:​ FATAL Mk5DataStream::​calculateControlParams:​ bufferindex>​=bufferbytes:​ bufferindex=170311680 >= bufferbytes=170311680 atsegment = 7</​del>​
 +
 +  * vex2difx (or other) support for creating .machines and .threads (from the eVLBI perspective)
 +  * Chase up and eliminate the cause of occasional "​Somehow Visibility X ended up with Y+1 subintegrations - was expecting only Y" errors
 +
 +====Longer term wishes====
 +
 +These are probably beyond the horizon of DiFX 2.0.2 unless a champion comes in to make them reality
 +
 +  * vex2 support (relies on vex2 finalization and some support from field systems)
 +  * Generalized GUI with unified database backend (Work progressing at MPIfR and USNO)
 +  * Amplitude correction for primary beam effects for off-center correlation (To be addressed at 2011 meeting at Haystack)
 +  * Move to a binary format for pcal files and name by datastream not by antenna name (To be addressed at 2011 meeting at Haystack)
 +
 +====Bug Fixes====
 +
 +  * LBA format data now scaled roughly correctly (removing the need for large ACCOR corrections)
 +  * There was a bug when xmaclength was > nfftchan for pulsar processing - this has been corrected
 +  * guardns was incorrectly (overzealously) calculated in mpifxcorr
 +  * "​Mk5DataStream::​calculateControlParams:​ bufferindex>​=bufferbytes"​ bug fixed
 +  * Low weight reads could result in uninitialized memory
 +  * Streamstor XLRRead() bug work-around installed several places (read at position 0 before reading at position > 0)
 +  * Fix to pulse calibration data ordering for LSB or reordered channels
 +  * Pulse Cal amplitude now divided by pulse cal averaging time in seconds
 +
 +====New Features====
 +
 +  * Mark5-based correlation:​ easy access to SMART data (can be viewed with getsmart)
 +  * Mark5-based correlation:​ emit multicast message containing drive statistics after each scan
 +  * VSIS interface added to mk5daemon
 +  * Experimental support for non power-of-2 FFT lengths
 +  * New utilities: mk5map, fileto5c, record5c
 +  * New features added to existing utilities:
 +    * mk5cp: copy without reference to a directory, send data over ssh
 +    * vsn: get SMART data
difx/d202wishes.txt · Last modified: 2015/10/21 10:08 (external edit)