[just now home from Ireland, hence the delayed response]

> > It would help if the WG agreed upon some functional requirements for 
> > templates. Any plan for that?
>  
> +1  I found myself wondering what requirements each solution satisfied.

I agree that the WG should define requirements first.  I will kickoff a thread 
on that shortly!

>  I would like to thank the group for the fruitful discussion on various 
> proposals on templates. When reading the drafts further, i have few questions 
> on the draft draft-ma-netmod-yang-config-template-00 - YANG Templates 
> <https://datatracker.ietf.org/doc/draft-ma-netmod-yang-config-template/> . 
> Since, it is marked as standards track, I would like to understand if the 
> intention is to standardise the ‘ietf-template’ model mentioned in section 
> 8.1 or to standardise the metadata object stmt-extend/operation-tag or both.  
> In my view, the stmt-extend/operation-tag is something which could as well be 
> achieved with existing RFC 7952?
>  
> I have similar queries on slides-121-netmod-sessb-16-yang-templates-01 
> <https://datatracker.ietf.org/meeting/121/materials/slides-121-netmod-sessb-16-yang-templates-01>,
>  are we proposing to standardise the module ‘yang-template’ in slide 3 or the 
> metadata object apply-templates/exclude-templates or both?
Please do not read anything into the Intended Status field of any of the 
current draft proposals.  All ideas expressed in current I-Ds and/or slides are 
effectively “Informative” at this point.

The show of hands performed at the end of that session clearly expressed that 
the WG would like to define a *standard* templating mechanism.

At some point the WG will adopt an I-D that has the Intended Status "Standards 
Track”, but that won’t occur until after the requirements are settled and a 
group of authors formed to put together such an I-D.

> In both the proposals, were the point on mandatories and defaults(which was 
> touched upon in the meeting) considered? I believe, this needs to be 
> considered when applying a template.
I recall that presentation.  I raised my hand to say the following (but lowered 
it when noticing that we were running low on time and wanting to giver others a 
chance to speak):  

As a contributor, I feel that the concern is unwarranted.  My belief/hope is 
that templates are not themselves validated when defined, but rather only the 
fully expanded/flattened results of templates are validated.


> IMO a solution using anydata + a deterministic method of identifying the 
> object for the root of template is required.

I'm trying to avoid discussing solutions too much now, but I tend to agree.  
Implicit in this statement, and I think different from any of the proposals, is 
that the data model (e.g., YANG) for the template may not be the same as the 
data model of the datastore the template resides in, but rather any node within 
that datastore.  The reasons for why this seems good to me (as a contributor) 
include: 1) better focus and 2) templates can be referenced anywhere a grouping 
is used.


Kent


_______________________________________________
netmod mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to