Post-STC1 (and realistically usable) Space-Time coverage in the registry; sect. 3.2
New resource types to support multi-resource services (as per Discovering Data Collections); sect. 3.1
Waveband is now governed by a vocabulary of messenger types; promoting this to REC would also make the vocabulary http://www.ivoa.net/rdf/messenger accepted (with Photon and Neutrino included as new terms).
Deprecated STCResourceProfile and replaced it with spatial, temporal, and spectral children of coverage; spatial has a frame attribute not further defined so far to allow for future extension to e.g., solar system bodies.
Dropped vs:Waveband and changed waveband to being controlled by a vocabulary that initially grows a generic Photon and a Neutrino concept over what the previous Waveband had.
Introduced new DataResource and CatalogResource resource types and wove them into the inheritance hierarchy to CatalogService; these are to be used for "dependent" resources.
New (optional) table/@nrows attribute gives an estimate for the size of a table
extendedType is now defined to correspond to VOTable xtypes in the absence of extendedSchema.
Required inclusion of quotes for delimited identifiers in a SQL context.
Sanctioning the use of footprint/@ivo-id to indicate the footprint standard used (this goes against the semantics of the ServiceReference tye underlying footprint, because the reference was originally intended to go to a registered service returning footprints – but no such services ever came around, and instead people used ivo-id as specified now).
Now allowing any vs:DataType element to define vs:InputParams.
In order to still ensure schema validation of type names, now advising to have an explicit xsi:type in param's dataTypes.
Deprecated TAPType.
DataType/@arraysize no longer defaults to 1, and the interpretation of arraysize=1 as a scalar is withdrawn. Use empty arraysize for scalars now.
Added a SHOULD requirement on CatalogResources to have a tableset.
Deprecated DataCollection and StandardSTC (which are no longer needed).
Adding a summary of the Discovering Data Collections EN.
BaseParam's delim attribute no longer defaults to blank. That conflicted with VOTable, where other conventions are in place (e.g., for string arrays).
Now discouraging the use of delim outside of InputParams.
No longer requiring unique table names within a tableset; uniqueness is now required within a schema. Many services have been in violation of the old unique-within-tableset rule for a long time without operational difficulties, and this change affects little as fully qualified names are now required for the main uses of tableset.
Ported source to ivoatex.
nrows
quantity in TAP metadata display where applicable.
taplint
validator release >3.4, or failing that the latest pre-release, will validate against the schema coming with this PR.
nrows
attribute gives the approximate number of rows. I'm quite happy that that's a reasonable thing to supply, but I wonder if the name "nrows" might be confusing, and conceivably clash with a future requirement for recording the exact number of rows. I would suggest at least that the nrows Meaning entry should read "The approximate size of the table in rows" rather than "The size of the table in rows". And maybe rename the attribute nrows_approx
or similar? Not sure.
tableset
still says "Each schema name and each table name must be unique within this tableset." My understanding from the change log and (I think?) the CatalogResource type definition is that this is no longer exactly the case; table names must be unique only within their host schema.
delim
attribute is defined as '<xs:attribute name="delim" type="xs:string" default=" "/>'. Should that be 'fixed=" "' instead?
Group | Yes | No | Abstain | Comments |
TCG | ||||
Apps | ||||
DAL | ||||
DM | ||||
GWS | ||||
Registry | ||||
Semantics | ||||
DCP | ||||
KDIG | ||||
SSIG | ||||
Theory | ||||
TD | ||||
Ops | ||||
StdProc |