Re: [core] ? WG adoption of draft-veillette-core-yang-cbor-mapping-00

[Andy Bierman <[e-mail]>]

Hi,

I am not advocating a requirements document but it is helpful
to understand them before trying to agree on the details.
It is not just data model or nothing to consider, but what
data modeling features are really needed.

For example, YANG allows decentralized model extension.
Any module can augment another one, without any process
other than writing the module. YANG has many ways
(grouping, leafref, typedef, etc.) to easily reuse other modules.

YANG has the concept of a datastore, not just API input and output.
It has the notion of configuration vs. state, that is also relevant to NM.
Cross-object validation (e.g. must-stmt) and conditional usage (when-stmt)
is important for configuration management. It has advanced support
for module conformance (what is optional to implement? what is optional to
use?)

Is it OK if CoAP is used for purposes outside of IoT?
I think constrained networks are within the charter,
so real network management of  devices on constrained networks
seems relevant to this WG.



Andy




On Thu, Apr 21, 2016 at 1:46 PM, Juergen Schoenwaelder <
[e-mail]> wrote:

> On Thu, Apr 21, 2016 at 07:47:53PM +0000, Michel Veillette wrote:
> > Hi Juergen
> >
> > I'm not sure I can share a LWM2M definition files since peoples need
> fill a form to get access to them, see the following link.
> >
> >
> http://technical.openmobilealliance.org/Technical/technical-information/release-program/current-releases/oma-lightweightm2m-v1-0
>
> Then I can't help.
>
> > I didn't go through a formal mapping between these two modeling
> languages but following is what this mapping might look like:
> >
> > LWM2M                | YANG
> > ---------------------+---------------
> > <Object>             | module
> > <Name>               | module name
> > <Description1>       | description
> > <MultipleInstances>  | list if true, container if false
> > <Mandatory>          | mandatory
> > <Item>               | leaf if <MultipleInstances> is true, leaf-list if
> false
> >  <Name>              | leaf name
> >  <Operations>        | config
> >  <MultipleInstances> | See <Item>
> >  <Mandatory>         | mandatory
> >  <Type>              | type
> >  <RangeEnumeration>  | range or enum
> >  <Units>             | units
> >  <Description>       | description
> >
>
> The devil is usually in the details but since there is no public
> specification I do not know. It sounds somewhat surprising that an
> '<Object>' would equate a YANG module but then I simply do not know.
> Only someone who has access to the specifications and who can get
> clearance to write about a mapping according to IETF rules can work
> on this.
>
> /js
>
> PS: You shall [...] not [...] use all or any part of a Document as part
>     of a specification or standard not emanating from the Licensor
>     without the prior written consent of the Licensor [...]
>
> --
> Juergen Schoenwaelder           Jacobs University Bremen gGmbH
> Phone: [phone]         Campus Ring 1 | 28759 Bremen | Germany
> Fax:   [phone]         <http://www.jacobs-university.de/>
>
> _______________________________________________
> core mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/core
>