VO-DML specification document

Fri May 9 02:44:47 PDT 2014

Hello Gerard,
Hello all,
On 09/05/2014 09:15, Gerard Lemson wrote:
> Dear Francois.
> Markus has given a very thorough response to the main point you were trying
> to make (I think).
Markus' answer doesn't cover all the issues  I was adressing (as  he 
admits himself ) and I will come back to his answer later.
Let's go to the "discussion process" now.
> But that point was not about the VO-DML specification document. I will try
> to respond to the points you make that relate to that.
In Hawai, the decision was taken to have two drafts,  related but 
independantly driven to (potential) recommendation. That's my point. 
Apparently this is not how the things work up to now.
>   
>> Starting from the last TCG teleconf time and from that email from
>> Gerard below there has been  a lot of discussions around VO-DML those
>> days and there are some aspects which give me some concerns:
>>
>>
>> *	    I think many people are still mixing two aspects which were to
> be
>> separated in two different drafts according to the conclusions of our
>> controversal discussion held in Hawai as they were summurized by Jesus
> here:
>> http://wiki.ivoa.net/internal/IVOA/PlenarySessionsSep2013/DM_Closing_Hawaii
> 2013_JSalgado.pdf (slide 6 and 7)
> Sadly (for other reasons that missing a controversial discussion) I was not
> in Hawaii, but the aspects you talk about were already separated into two
> different documents by the time of the Heidelberg interop. This had been
> announced before that meeting.
Two documents, yes but no independance of the possible results for the 
two discussions.
> It was a decision made near the end of the utype tiger team that it would be
> best to create 2 documents, one for the data modelling language, one for its
> use in annotating tabular data sets using utypes, the "mapping document".
> See the last 4 minutes or so on
> http://wiki.ivoa.net/twiki/bin/view/IVOA/UtypesTigerTeam
>
These minutes are more than one year old. Since that time there have 
been two discussion in very open interop sessions and one in TCG. 
Obviously there is no consensus on the proposal of the TIGER team for 
utypes. The TIGER team cannot do as these objections could be ignored 
and not discussed seriously.
>> I see that a lot of work has been done for the update of the first
>> draft "VO-DML: A Data Modeling Language for the VO" but nothing new
>> had been done
>> about the second draft "Mapping of Complex Data Models". The title of
>> this second draft reflected the difficulties appearing in using vo-dml
> description
>> to map the models into VOTable, making an extensive but seriously
>> modified usage
>> of the utype attribute.
> The title of the utypes mapping document is "UTYPEs: Portable Data Model
> References".
I know this document.
> It can be found from reference [3] in the VO-DML spec but it has been linked
> to often enough elsewhere as well.
Yes I know. And (as some other people) I object the mechanism described 
here should REPLACE the current utype mechanism.
The document is stable since May 2013 and has not been modified or 
replaced so far. That's the points
> Note that even if the title of that document was "Mapping of Complex Data
> Models"
> (which it is not), the adjective "complex" is associated to the "data
> models" object, not to the "mapping" subject. The reason why the majority of
> the tiger team was in favour of the proposed mapping mechanism is because it
> is remarkably simple even for more complex models, as long as those are
> designed well and written in a properly designed modelling language.
That's the tiger team opinion. Again this has been openly discussed in 
Heidelberg and Hawai and there has been no consensus around this opinion.
The name proposed in Jesus' slides was  given for this very reason, I 
think.
The relationship between the VO-DML mechanism and utypes had to be 
rediscussed according to the objections made.
>
>> However,  I can read several sentences which show that for many people
>> nothing has changed since the time where the introduction of the draft
>> VO-DML draft was first written.
>>     Here is a quotation of the abstract of the vo-dml document "VO-DML
>> a consistent modelling language or IVOA data models"
>>
>>
>> 	"Arguably the most important use case for VO-DML is the UTYPE
>> specification [2]
>> 	which uses it to provide a translational semantics for VOTable
>> annotations.
>> 	These annotations allow one to explicitly describe how instances of
>> types from a
>> 	data model are stored in the VOTable."
>>
>>    and all the introduction still emphasize that the main use case for
>> VO-DML
>> modelling language is utype specification in VOTABLE. It also Implies
>> that the
>> VO-DML GROUP mechanism is THE (unique) way to do the mapping.
>>
>>      Last but not least I see nothing like a ""Mapping of Complex Data
>> Models" document in the repository.
> See above for links to the doc you must be referring to.
> And I still stand by these conclusions that the utype tiger team drew, but
> may well have repeated it too often in the text.
> And note that the utype mapping is also arguably (!) the most important use
> case for this spec, because it was the direct motivation to propose VO-DML
> in the first place.
I remember this. But it happens sometimes that by working on a 
predefined goal people develop a "detail" or a "first step" which appear 
very usefull, but however the intial goal his not reached. This is 
"historical hazard in discovery". Exactly this for me: the unified 
description of data models you have been proposing looks nice and 
usefull for me but I don't think the related utype mechanism you derive 
from this is convenient or usable as is IN GENERAL.
>>       It must be clear to everybody that apart from this, most of the
> effort done
>> under building a consistent modelling language for IVOA looks very
> promising to
>> me. Having a description language with xml serialization alllows to
>> share diagrams and models built with different modelling softwares and
>> allows to help generating interoperable documentation and code. This
>> is a real progress and I appraciate the effort done By Mark (and now
>> Arnold) to map various models in the work, done by Gerard, Omar and
>> others. For me this is core of
>> "VO-DML a consistent modelling language for IVOA data models" and this
>> is a progress.
>> Probably we have things to discuss still (the utype attribute stuff
>> and the ivoa
>> datatypes among others) but I see no objection in going forward along
>> this path towards recommendation
>>
> Great. So let's focus on that.
Probably I have to be more precise : "I see no objection as long as I am 
not buying the VO-DML utype mechanism WITh the standardized description 
of model in VO-DML language and xml."
>   We want to get the wrinkles in the spec
> ironed out.
> As I have repeatedly pointed out, there was a whole slew of models expressed
> to VO-DML by me during tiger team. This was a proof of concept and I hoped
> that authors of the original models could maybe use them as a starting point
> for a refactoring effort.
> I included Characterisation, which you were involved in. It would be good if
> you and co-authors could try to mimic the effort of Arnold and to come up
> with a VO-DML representation of that model. Particularly because attempts
> are made to use it in the other efforts.
I see no objection in helping in that (for Characterisation 2, not 1.33) 
as long as I am not buying the VO-DML utype mechanism WITh the 
standardized description of model in VO-DML language and xml"


Cheers
François
>   >
>> *	      So let's talk about "Mapping of Complex data models"
>>
> I'll comment on your points related to the utype spec in another email
> though I will not be able to improve on Markus very eloquent response.
>   
> Cheers
> Gerard