dtody at nrao.edu
Tue May 15 02:19:39 PDT 2007
Minutes DAL-2, Beijing, 15-May-2007, 10:30-12:00
Chair: D. Tody
Minutes: M. Dolensky (edits by D. Tody)
DAL Session Page and Presentations
I) SLAP [Delgado]
- more informative error messages, possibly defined error codes
- score/TOP, optional MAXREC to limit large query set
- need way to specify precision of returned line model metadata
- units handling is an issue; there is a need to return uniform
units, but it is often desireable to propagate units from data source
The issue of SLAP and SSAP compatibility (consistency with 2ndGen DAL
interfaces) was discussed. The decision to support format=metadata
in SSAP simplifies things for SLAP as this is what the current WD and
applications use. Can make SLAP V1.0 service profile fully compatible
with SSAP with minor changes, and then add getCapabilities etc. in a
V1.1 version as for SSAP, and we have achieved a uniform interface.
II) Spectrum DM [McDowell]
Clarification: Long string convention should be used for TTYPn and TUTYPn FITS keywords
5 implementations (e.g. http://vo.cfa.harvard.edu/soft)
missing from the list on the slide: ASPID-SR
Q: Why is RESPOWER not part of the resolution object?
A: Specific to spectral axis but not the other axes like time.
Characterization- and Data-Class do differ to some extent (query response vs. data serialization):
This implies two utypes depending on the context, but while developing the DM it turned out that the two
are different and shouldn't arbitrarily unified => no roadblock for PR V1.0.
- incorporation of DMs: SSA currently has basically a copy of certain pieces rather than directly inheriting from
various specific Schema. This eases configuration control. Upgrades are deliberate with all pros and cons this implies.
Conclusion: Advance Spectral DM doc to PR RFC.
Discussion on advancement of SSA IF doc to PR.
- Strategy is to move SSAP V1.0 PR forward immediately after interop,
based on discussions of remaining issues from the sessions, using the
format=metadata mechanism to provide interface introspection.
- A backwards-compatible V1.1 will follow which includes getCapabilities
and getAvailability, as per agreement with GWS on VOSI compatibility.
- [Bonnarel] protocol info is missing from SSA query response
- [Bonnarel] to provide a paragraph about SIA extension mechanism
- [Dobos] discussion of protocol extension to support theory data
Conclusion: Advance SSA IF doc to PR RFC following interop.
More information about the dal