TODO in vex2difx x. Capitalization of Antenna names and obs code x. # sign x. load params file x. Pulsar support x. Spacecraft support x. Handle files (i.e., not Mark5 modules) x. Handle multiple phase centers x. full paths to files x. guardblocks, blocks per send x. what to do about modes that don't support all antennas? (c081a.skd currently fails) - Adam suggests populating a bogus datastream entry with no corresponding baselines x. support mark5b x. support VDIF x. support baselines= x. implement maxLength x. use antennas, not modules, in antenna count x. support time formats other than mjd x. support EVLBI 18. reimplement oversampling/decimation support in a cleaner way 19. Clean up antenna names/defs as done for sources x. Currently corrparams is used both during vex parsing and at input building time. It would be nice to eventually make the vex parsing independent of corrparams. 21. vex2 support 22. option for not splitting scans where a non-member antenna has a clock break Concerns outside vex2difx proper x. Need ant velocities and pos'n epoch in vex file - what is a year? not everyone thinks it is 365.25 days x. Need to get module shelf location from track 3. Calcode/Qualifier in vex x. Get exper_nominal_{start,stop} in vex files After vex2 is in good shape 1. Renormalize how clocks are handled - get rid of offset,rate,accel,jerk and replace with vector - clarify in .v2d whether supplied clock relaces or augments .vex values - vex2obs: properly link 2. EOPs - generalize to all parameters - vex2obs: properly link 3. Multi-phase-center - update wiki and code: require source name - pass through cal code vex2 issues to resolve: * currently calCode is affixed to a source, but vex2 associates with a phase center within a scan * EOPs: need to support different "eop origin" values per EOP value; keep individual measurements separate, or bundle to degree possible?