HSA with Additional Resources (AmyMioduszewski). An HSA observation with Y27 (phased array), but simultaneously use the full bandwidth of the VLA correlator; that is, use DiFX and WIDAR. We need more discussion for this class of proposals where multiple backends are used from different instruments. This is not expected to be an edge case; many folks are asking for this feature. This may also arise with HSA observations using the GBT as a station (e.g., use GUPPI). One option is to have one Allocation Request where the multiple backends would be specified within a given Allocation Target. This is in contrast to the use case where we have HSA using Y1 and the VLA is used simultaneous. Here there are two Allocation Requests (one for HSA and one for VLA).
Astrometry Project (4.3.1 in Document). Measure the water maser emission of 10 sources to determine the parallax. Technically only need 2 measurements separated by 6 months. The Allocation Constraint formalism developed earlier should work here: a list of times with deltas for each source. There would only need to be one Allocation Request with potentially multiple SBs. In the past, astrometry projects have often been observed with fixed time depending on the priority. But since the constraints may be different for each source it is not clear how this fits into our current structure: for each Allocation Request there are one or more Allocation Targets and Allocation Constraints. We agreed to keep this simple and require different Allocation Requests for Targets with different constraints. This evolved into a philosophical discussion about complexity. We want to make sure that the software interface is designed to be simple for simple projects so as not to deter non-experts. We also do not want to encourage people to use the software in complex ways because they can do so. Nonetheless, we have to accommodate the more complex observing strategies.
VLBA Pulsar with a VLA Trigger (AmyMioduszewski). The VLBA is dynamically scheduled to observe a pulsar. When this is observed the VLA is triggered within a week. This would be two Allocation Requests: one for the VLBA observations and one for the VLA. The VLA will have an Allocation Type of Triggered.
Update Document (LorantSjouwerman). DanaBalser has incorporated comments from MarkClaussen (version 0) and LorantSjouwerman (more recent version). He will add a date to the wiki page that corresponds to the most recent version that has been uploaded (which is also listed in the PDF).
Timeline (LorantSjouwerman). We agreed to a goal of drafting a complete document by the end of the calendar year (31 December 2018).