Notes for Scantable vs. Measurement Set Discussion

scantable issues

• most serious
  • polarization type - global: only one type per scantable - not real problem...
  • currently scantable assume some globals (NIF, etc) - but scantable intrinsically can handle...
  • difference in spw - IF: can be put in aux. table in MS? (SPECTRAL WINDOW table have 'freq group')
• others
  • scanno renumbering - (mostly) reader issue
  • antenna dish diameter info gets lost - reader issue
  • POINTING sub-table truncation
  • weather info- reader issue
  • the issues presented can be resolved by exntend current scantable data format
  • data size?- even we copy some missing meta data, scantable probably be much smaller in size

To keep scantable

• direct MS->scantable filler + scantable -> MS filler (~3+1mon.) - Tekeshi
• adding more meta-data in scantable
• Malte and Takeshi will work on a new scantable format proposal (include more meta data)
• plotter issue?
  • if we stay on scantable, we can still write out to MS to use plotms
  • still need to add SD specific functionality