param-link-to-field

[François Bonnarel]

Hi all,
    Considering the intial proposal in the draft (by example)

> <RESOURCE type=”service”>
> <PARAM name="resourceIdentifier" datatype="char" arraysize="*"
>          value="ivo://example/datalink" />
> <PARAM name="standardID" datatype="char" arraysize="*"
>          value="ivo://ivoa.net/std/DataLink#links" />
> <PARAM name="accessURL" datatype="char" arraysize="*"
>          value="http://example.com/datalink/mylinks" />
> <GROUP>
> <PARAM name=”ID” />
> <FIELDref ref="datalinkID" />
> </GROUP>
> </RESOURCE>

and markus' following argument
> it's .... (not) explicit (enough); sibling elements within an (unadorned) GROUP
>      have no explitict semantics, a LINK child of a PARAM does
which I think makes sense, I am still reluctant to use the LINK scheme. 
actually Markus' proposal is "forcing" LINK structure beyond the limits
I fully agree with Pierre's caveats there.

A simple "ref" on the SErvice parameter could work as long as it's 
distinguished  by an ad hoc (DataLink specific utype) this way, eg :
> <RESOURCE type=”service”>
> <PARAM name="resourceIdentifier" datatype="char" arraysize="*"
>          value="ivo://example/datalink" />
> <PARAM name="standardID" datatype="char" arraysize="*"
>          value="ivo://ivoa.net/std/DataLink#links" />
> <PARAM name="accessURL" datatype="char" arraysize="*"
>          value="http://example.com/datalink/mylinks" />
> <PARAM name=”ID” utype="dl:serviceInputParam" ref="datalinkID" />
> </RESOURCE>

  Best regards
François


Le 18/02/2014 09:37, Markus Demleitner a écrit :
> Hi Pat,
>
> On Mon, Feb 17, 2014 at 09:01:14AM -0800, Patrick Dowler wrote:
>> Thanks Markus; that clears things up. Since I didn't hear otherwise,
>> I am going to change the WD to use a child LINK instead of the
>> sibling FIEDLref (with href fragment).
>>
>> When I do that, I need to explain what content-role="ddl:id-source
>> means. What is that "ddl" prefix looking thing and where is it
>> defined?
> Is it possible to be honest and just say "The content-role is an
> opaque string to be compared case-sensitively."?  I'm not wild about
> that concrete string either, and I'd like "id-source" just as well (I
> wouldn't even loudly object to
> clsid:8AD9C840-044E-11D1-B3E9-00805F499D93) -- I'm rarely worried
> about name clashes in such fairly tightly controlled circumstances.
> Meaning: If you like some other string better, just change that.
>
>
>
> Since it kind of fits in here, allow me a brief indulgence in
> told-you-so: In Waikoloa, I proposed a model written in VO-DML to
> describe the Datalink response and thus have "principled" way of
> figuring out its serialization.  I'm not saying we should dig this up
> now that the standard has already progressed that far, I'm only
> saying the questions of such string forms, referencing in general,
> and all that comes with it are examples of why something like VO-DML
> and its standard serialization(s) would make our lives a good deal
> easier.  So much would suddenly be automatic...
>
>
> Cheers,
>
>           Markus