Re: [yang-doctors] Do we need a YANG doctor lunch meeting at the IETF meeting?

Andy,
>
>
> On Wed, Jun 21, 2017 at 2:57 AM, Ladislav Lhotka <lhotka@nic.cz 
> <mailto:lhotka@nic.cz>> wrote:
>
>     Andy Bierman <andy@yumaworks.com <mailto:andy@yumaworks.com>> writes:
>
>     > Hi,
>     >
>     > I have a topic that is probably not workable but the issue is
>     very common.
>     >
>     > It would be nice if YANG library repositories were more
>     consistent and
>     > official.
>     > There is currently a ban against packaging YANG modules (e.g.,
>     RPM or
>     > Debian)
>     > because of concerns the IETF Trust copyright will be violated
>     somehow.
>     > Maybe this could be changed somehow.
>     >
>     > The packages could correlate to the YANG catalog, e.g. an
>     > ietf-routing_1.0.deb package.
>     > I have been advocating YANG packages for a few years. IMO it is
>     hopelessly
>     > futile to
>     > attempt to manage a device as a random collection of individual
>     > modules.
>
>     I would support this. I believe it is necessary to sort of
>     "emancipate"
>     YANG library, i.e. treat it as metadata that can exist on its own
>     rather
>     than plain state data provided by a server, or perhaps as a
>     module-level
>     data modelling language.
>
>     Also, I believe it would be extremely helpful to step back and somehow
>     streamline and simplify all what we have, and in particular integrate
>     YANG library with datastores and the "use-schema" part of schema mount
>     into a coherent whole that is (relatively) easy to understand. Not
>     doing
>     so and just increasing complexity along several axes has IMO the
>     potential to kill the YANG business as such.
>
>
>
> I think the YANG catalog module attempts to provide off-box metadata.
You will get a presentation of it during the OPS-AREA meeting.
> The OpenEmbedded project is far superior to anything the IETF could do
> on its own, but serves as an existence proof that building blocks can
> be managed somehow (layers, feature-groups, packages for yocto).
>
> Networking equipment is still built with a monolithic mindset, so it
> will be difficult for vendors to provide real modularity. The 
> first-level goal
> is to let YANG designers and operators organize YANG APIs at a higher
> level than individual YANG modules.
>
> I think the YANG catalog work is properly motivated and heading
> in the right direction to define the packages.
Yes.
This is the notion of release bundle in 
draft-openconfig-netmod-model-catalog/

Note that they YANG module behind yangcatalog.org diverged from 
draft-openconfig-netmod-model-catalog, for good reasons.
Joe and I are preparing an IETF draft to explain those.

How many YANG doctors will we see in the hackathon ...

Regards, Benoit
>
>
>
>     I already made a concrete initial proposal in Chicago:
>
>     https://www.ietf.org/mail-archive/web/netmod/current/msg18045.html
>     <https://www.ietf.org/mail-archive/web/netmod/current/msg18045.html>
>
>     Lada
>
>
>
> Andy
>
>
>     >
>     > I have been working on Yocto/bitbake recently and the amount of
>     high-quality
>     > tools (see OpenEmbedded) is quite impressive. We could have
>     something like
>     > a bitbake recipe
>     > that describes the YANG requirements for some service or feature
>     package.
>     > Vendors could
>     > append the recipes to optimize for their platforms. Recipes
>     could point at
>     > open-source code
>     > or even vendor libraries that provides instrumentation for a
>     YANG package.
>     > We need distributed recipes, not a centralized catalog server,
>     although
>     > they could work together.
>     >
>     >
>     >
>     >
>     > Andy
>     >
>     >
>     >
>     >
>     >
>     >
>     > On Tue, Jun 20, 2017 at 9:26 AM, Benoit Claise
>     <bclaise@cisco.com <mailto:bclaise@cisco.com>> wrote:
>     >
>     >> Regards, Benoit
>     >>
>     >> _______________________________________________
>     >> yang-doctors mailing list
>     >> yang-doctors@ietf.org <mailto:yang-doctors@ietf.org>
>     >> https://www.ietf.org/mailman/listinfo/yang-doctors
>     <https://www.ietf.org/mailman/listinfo/yang-doctors>
>     >>
>     > _______________________________________________
>     > yang-doctors mailing list
>     > yang-doctors@ietf.org <mailto:yang-doctors@ietf.org>
>     > https://www.ietf.org/mailman/listinfo/yang-doctors
>     <https://www.ietf.org/mailman/listinfo/yang-doctors>
>
>     --
>     Ladislav Lhotka
>     Head, CZ.NIC Labs
>     PGP Key ID: 0xB8F92B08A9F76C67
>
>