VEP-007: datalink-core#detached-header

[Markus Demleitner]

Dear colleagues,

On Tue, Jun 15, 2021 at 10:13:57PM +0200, Baptiste Cecconi wrote:
> http://volute.g-vo.org/svn/trunk/projects/semantics/veps/VEP-007.txt

Thanks, Baptiste, for preparing this VEP; just as a procedural
remark, in general it's better if people just send the proto-VEP to
the semantics chair (right now, yours truly), who'd then do a formal
check, assign the VEP number, do the necessary changes to the
vocabularies, and then make the announcement.  Not that it matters
*too* much, but in case multiple persons prepare VEPs concurrently,
having the semantics chair as a semaphore will prevent race
conditions (like multiple VEPs having the same number for a while).

Oh, and of course the Semantics chair also puts the stuff into the
vocabulary repo, which I've just done; see 
http://www.ivoa.net/rdf/datalink/core for how the new term looks in
context.

On the proposal:

> New Term: #detached-header
> Action: Addition
> Label: Detached Header
> Description: Machine-readable metadata for #this, which in general 
> will be necessary for its scientific use.  Examples include FITS 
> headers distributed without their data blocks or PDS label files. 

I like that a lot, in particular because it has very clear
"pragmatics", meaning: "What do people want clients to do when they
encounter this piece of semantics?"  Where headers are stored
and delivered separately, clients clearly need a reliable way to find
them.  There is also a lot to be said for letting clients read
headers up-front, before pulling potentially Gigabytes of data.

If there were some takeup of this, I'd probably make it easy on
DaCHS (i.e., server-side) to offer such links for FITS files (where
it'd probably be "primary header, except for compressed images, where
it's the first extension header").

In case you see problems or ways to improve this, please speak up
soon.  If I don't hear from you, I'd forward this for TCG review on
July 5.

        -- Markus