Agenda

CASA/JIVE VLBI Coordination Meeting

Minutes in Bold

2020 February 19 @ 1500UTC

Attendance: George, Justin, Mark, Des, Ilse, Ryan

Logistics

  • Future Meeting dates (usually last Wed of the month):
    • NEXT MEETING: 2020 March 25 @ 1400 UTC. (Note daylight-savings!!)
    • 2020 April 29?
    • 2020 May 27?

General Admin

  • Recent/upcoming meetings:
    • Space VLBI 2020 in Charlottesville, VA (2020 Jan 28-30): George (w/ Fomalont) to present poster on compensating for OD errors
    • Interesting meeting, trying to bridge from science to technology challenges (cf prior meeting ~2 years ago in NL) but still mostly aspirational and idealized discussion; no real convergence on how/what to pursue. Next proposed meeting: late 2021.
  • New CASA git repository. See: casa6 Any issues?
    • Need plotms build instructions (George will ask Pam for details...TBD)
    • Unclear how to make CASAshell work (George will followup...TBD)
    • CASA5 build possible within casa6 repository structure (George to provide simple instructions...TBD)
    • Need instructions for running c++ unit (google) tests (Work ongoing on CASA side, stay tuned...)
  • Current VLBI Integration branch: CAS-12325.
    • Will make new VLBI integration branch from new repository soon, modulo items that move out of CAS-12325 in current work (these should be satisfied in 5.7/6.1 release)
    • (Mark) Waiting for dust to settle on casa6 migration; new casa6 branch will likely be needed by summer.
  • Michael Janssen Socorro visit (Feb 10-11)
    • Exercised rPICARD pipeline with Jay Blanchard mainly, troubleshooting pipeline and CASA on VLBA data. Should get report from Jay (feed into requirements). George discussed general planning issues. One new item: Using MS intents (via convention established by ALMA + VLBI-specific adaptations as needed). Mark notes next VEX version will recognize intents. A script that parses the VEX file and adds intents to the MS will probably be needed until coreelators propagate this information into their output.
  • (Mark) Requirements collation ConsolidatedRequirements....Any news?
    • Mark has completed initial draft and has shared with original authors of earlier docs for a first round of comments. Will cast a wider net (Jansssen, Blanchard, Linford, others) for comments soon.
  • Work plan describing how/when CASA can replace AIPS (FY2020 NRAO/DMS Program Operating Plan)
    • Req doc should be in good shape by our March meeting; should formulate plan for generating the corresponding VLBA readiness doc (minimum requirements, timescales for completion, etc.) at March meeting, expecting completion ~June (e.g., in time for NRAO UC, and 3rd quarter POP review).
  • Stakeholders' deliberations for 5.8/6.2 upcoming
    • CASA 5.8/6.2 decoupled from ALMA/EVLA pipeline planning (though 5.7.x patches likely for them). Casa6 repo will roll over to 6.2/5.8 in April, anticipating an October release.
    • George will supply Urvashi with VLBI-specific todo list (from below) for stakeholder deliberations over the next ~month or so, thereby formalizing CASA team commitments,
  • (Ryan) CNGI Status
    • F2F meeting Feb 4-6 in Socorro to discuss framework principles with principal imaging and calibration developers. Good meeting, exciting times, etc.
    • Expecting CASA 6 and ngCASA to be "interoperable" at least modulo physical data format conversions. NB: MSv2 (or v3?) will remain the essential user-facing (science) data model, but disk format will be quite different as per CNGI demands.
    • CoDR as soon as ~late April/early May, possibly coordinated with CASA F2F meeting?
    • NB: CNGI concerns the framework itself and related tools; efforts here ramping up ~now through 2021. "ngCASA" will be the new implementation of CASA functionality using CNGI; efforts here ramping up later in 2021. Support for CASA 5/6 to remain for many years, but general support commitment likely to taper down as early as Summer/Fall 2021, modulo imperatives.
    • JIVE concerns (Mark):
      • MS format (including current use of Table system) being encouraged for new/existing instruments/correlators (move away from UVFITS and similar, especially)
      • VLBI not a big "scalability" customer, so principle benefits of CNGI not as attractive (Ryan: smallish MSs should be directly servicable, in memory)
      • Port of fringefit and other VLBI specific code represents a non-trivial work commitment
  • Other admin?
    • JIVE folks welcome at CASA F2F (dates still TBD, but possibly as soon as last week of April). Will provide for remote participation, as needed.

Development/Features

  • Items for CASA 5.7/6.1 (Stakeholder inputs)
    • (George, CAS-695) Support correlation-dependent flagging, with user option.
      • Coding work done, will push thru casa6 repo workflow this week; new verification tests pending. On track for 5.7/6.1.
    • (George, CAS-12735) Frequency meta-info consolidiation (see also CAS-12553)
      • Full intended refactoring work probably deferred; essential fix within FringeCal to be included in 5.7/6.1.
    • (Mark/George, CAS-12591) spwmap fix
      • Coding done and pushed to casa6 repo; new verification tests pending. On track for 5.7/6.1.
    • (Des, CAS-12365,12653) Deploy dispersive term solve, including user term control
      • Working through testing workflow
    • (Mark) MS concat bugfix on casacore github to CASA? (CAS-12872)
      • Dirk Petry and Mark now verifying that current casacore submodule in casa6 repo includes this fix.
  • Items for future cycles (future stakeholders discussions should include these items, at least):
    • (George) Items above dashed line to be included in 5.8/6.2 planning
    • (Mark, CAS-8546) GAINCURVE definition, deployment of standardization (deployment --> 5.8/6.2)
    • (Des, CAS-12765) Wide bandwidth (deployment in CASA --> 5.8/6.2)
    • (George/Des, CAS-xxxxx) Convert CC tables to image or ComponentListfor resolved model-setting
    • (George) Per-scan interpolation option?
    • (George) Overlapping solution intervals (CASR-467, CASR-468)?
    • (Mark) EOP: formalize import of (DifX) correlator model, update EOP (or other, e.g., troposphere) part, and difference to obtain required delay calibration
    • (George) Finish frequency meta-info consolidation (CAS-12735)
    • - - - - - - - - - - - - -
    • (As of completion of requirements doc and related, the list below will be removed, and we'll maintain only a list of current in-progress items)
    • (Des/George) PCAL (Des has ad hoc scripts, need updated prioritization from VLBA since this is not a high EVN priority; not a high priority at NRAO/VLBA in the near term
    • (George) Polarization calibration w/ resolved calibrators
    • (Des) CAS-12485? fringefit weighting control (test data options pending)
    • (Des) Baseline stacking?
    • Opacity for high-frequency VLBI, derived from Tsys(sec(z))
    • TEC
    • (Walter) Advanced VLBA Amplitude calibration a la Walker memo from a few years ago?
      • Bandpass normalization flexibility: voltage vs. power

Verification/Validation/Documentation

  • n/a

AOB?

  • n/a

-- GeorgeMoellenbrock - 2020-02-19
Topic revision: r2 - 2020-02-20, MarkKettenis
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding NRAO Public Wiki? Send feedback