slight inconsistency (?) in document "IVOA Registry Interfaces v1.0" in Keyword Search Query section (2.1.3)

[Markus Demleitner]

Hi Menelaus,

On Fri, May 02, 2014 at 09:53:29PM +0200, Menelaus Perdikeas wrote:
> Hi,
> 
> With reference to document "IVOA Registry Interfaces - Version 1.0":
> 
>     http://www.ivoa.net/documents/RegistryInterface/20091104/REC-RegistryInterface-1.0.pdf
> 
> Near the top of pg.18 (Section 2.1.3 Keyword Search Query) we read:
> "The operation only queries for active resources noted by
> status='active'".
> 
> Near the bottom of the same page we read: "For each active or
> inactive resource record ..." (incidentally, it is not clear to me
> why "resource record" is underlined as if to perhaps juxtapose
> "resource record" against "resource" earlier in the page - it is
> the only occurrence of "resource record" that's underlined in the
> document).
> 
> My interpretation is that the first provision takes precedence so
> that keyword search query is consistent with constraint-based
> search query (Section 2.1.2) where we read (end of pg. 15) "deleted
> and inactive records, therefore, should not be returned".

I am fairly sure this is the only sensible strategy.  Deleted records
are basically an artifact of incremental harvesting with OAI-PMH
anyway and certainly should never be user-visible.

For inactive records, the case appears less clear to me, but then
I have to admit I have never felt much affection for that status
anyway (meaning: I think we'd be doing ourselves a favour if we
severely discouraged  its use).  Anyway, I don't think clients should
be presenting inactive services to their users, so even if they're
not pure artifacts, from a usability PoV users should not see them.

Finally, RegTAP says only active records are to be ingested.  By
stipulating that RI1 only makes active records visible, too, we're
preventing one more thing that might confuse people when moving from
RI1 to RegTAP.

Cheers,

       Markus