Re: [Netmod-ver-dt] Schema-comparison

["Joe Clarke (jclarke)" <[e-mail]>]

You may want to do a search and replace for “stuf” to “stuff” before you have to use the rev-ext:nbc tag :-).

Does it make sense to take a stance on points 1 and 2 here?  I think it makes sense to define the extensions in yang-versioning.

With respect to rev-ext:nbc, I think this is useful when the tooling would otherwise not be able to determine an nbc change (like in description).  So I think we could say nbc and renamed-from are the useful extensions.  I also don’t think it would be bad to allow a namespace in the value of renamed-from in case a vendor wanted to move something to a different module.

Joe

On Mar 20, 2020, at 16:29, Reshad Rahman (rrahman) <rrahman=[e-mail]<mailto:rrahman=[e-mail]>> wrote:

Hi Rob, all,

I’ve attached slides, it’s basically the slides Rob did for 106 (pruned a few) + small addition.

I’ve kept the 1st 3 points which were already part of next steps (slide 7), not sure if they should be rediscussed or not?

Inline.


From: "Rob Wilton (rwilton)" <rwilton=[e-mail]<mailto:rwilton=[e-mail]>>
Date: Thursday, March 19, 2020 at 12:42 PM
To: "Reshad Rahman (rrahman)" <[e-mail]<mailto:[e-mail]>>, "[e-mail]<mailto:[e-mail]>" <[e-mail]<mailto:[e-mail]>>
Subject: RE: Schema-comparison

Hi Reshad,

Some comments inline …

From: Reshad Rahman (rrahman) <rrahman=[e-mail]<mailto:rrahman=[e-mail]>>
Sent: 18 March 2020 20:12
To: Rob Wilton (rwilton) <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>
Subject: Schema-comparison

Hi Rob, all,

I took a look at schema-comparison, it had been a while…. Regarding the issue<https://github.com/netmod-wg/yang-ver-dt/issues/25> we briefly discussed during the meeting:


  *   Ideally I believe any path should work with renamed-from (e.g. module to module). Helpful when comparing schemas. But TBH I don’t know how common this is. As for a solution, what comes to mind is we need an import statement for the “from-module” in the “to-module” and the identifier used in rename-from stmt would need to include prefix of “from-module”.

[RW]
For me:

  *   Renaming a node within a module is easier and we should make sure that we have that one covered.
  *   If we want to allow renaming from other modules then I think that we want to avoid using an import statement because it will create unwanted dependencies, and perhaps instead rely on a schema node path that uses qualified module name, e.g. perhaps like RESTCONF paths, with the default being the current module.

 <RR> Ack.


  *   The question about moving across module made me wonder whether a revision/version would be useful with the renamed-from extension.

[RW]
Not sure.  I think that this probably needs more further thought/discussion.  It may be that it turns out not to be required.
<RR> Missed the fact that this was already in the issue and presented at last IETF.

Regards,
Reshad.

Some comments/questions:

  1.  Actions aren’t mentioned in section 3, oversight?

[RW]
Yes, probably.


  1.  Filtered YANG schema comparison. In theory, some filters could be applied even for module comparison. Maybe take the filters out of YANG schema comparison and have a separate section?

[RW]
Yes, but I think that it might be more work than just pulling out the filtered section, since it also talks about constructing a full schema tree.

Perhaps:
Generic filtering should be a subsection of section 3 (or a new section 4).
The existing section 4 should also mention that the filtered comparison can be used.
The existing section 5.2 should be extended/renamed to indicate that it is comparing a full schema tree (or perhaps this is what we should be referring to a package schema).



  1.  Section 3 mentions minor and editorial and refers to yang-packages. Should refer to semver instead?

[RW]
Yes, good catch.


  1.  Section 3.1, this is where we should add descriptions of the extensions?

[RW]
Yes, as a starting point.


  1.  Examples would really help

[RW]
I agree.


  1.  Would be good to get feedback from tooling folks. Martin didn’t give any. I think we should ask yanglint (Radek?).

[RW]
Yes.  We could also take a look at what pyang does here.

Thanks,
Rob


Regards,
Reshad.


From: Netmod-ver-dt <[e-mail]<mailto:[e-mail]>> on behalf of "Rob Wilton (rwilton)" <rwilton=[e-mail]<mailto:rwilton=[e-mail]>>
Date: Tuesday, March 17, 2020 at 12:49 PM
To: "[e-mail]<mailto:[e-mail]>" <[e-mail]<mailto:[e-mail]>>
Subject: Re: [Netmod-ver-dt] Possible choices for editors

Notes from today’s DT meeting.

Bo, Joe, Reshad, Rob attended

We plan on keeping the 107 meeting slides format to broadly be the same, but raising more issues.

+ Joe to publish the ietf- version of the drafts..

+ RobW to ask about what we should do with the Reqs draft.

  *   Take to WGLC?
  *   Publish as informational?


+ RobW to raise with NETMOD/NETCONF chairs on the best way forward for progressing these drafts.

See further notes inline below.

From: Netmod-ver-dt <[e-mail]<mailto:[e-mail]>> On Behalf Of Rob Wilton (rwilton)
Sent: 03 March 2020 15:10
To: [e-mail]<mailto:[e-mail]>
Subject: [Netmod-ver-dt] Possible choices for editors

Obviously conditional on input from WG chairs & adoption results.

From discussion today:

6) draft-verdt-netmod-yang-schema-comparison-00 - Reshad?
  - Schema comparison tooling, unchanged from the version presented at 106.
[RW]

  *   Keep this discussion fairly short.


5) draft-wilton-netmod-yang-ver-selection-02 - Joe/Jason?
  - Version selection, updated since 106, as per notes below
[RW]
Issues:

  *   Move to NETCONF WG?
  *   Does it need to be split due to updating both NETCONF and RESTCONF?



4) draft-rwilton-netmod-yang-packages-03 - Bo?
  - YANG packages draft, updated since 106
[RW]
Bo producing slides
Discussed proposed issues to raise from the github issues list.


2) draft-verdt-netmod-yang-module-versioning-01 - Balazs?
  - Base module versioning solution, unchanged from the version presented at 106.
[RW]

  *   Feedback from Martin.
  *   Reshad to produce slides for 107.


3) draft-verdt-netmod-yang-semver-01 - Joe/Jason?
  - YANG Semantic version numbers, unchanged from the version presented at 106.
[RW]
-          Do we mandate IETF drafts use YANG Semver?
-          Joe’s going to produce slides.

1) draft-verdt-netmod-yang-solutions-03 - Rob?
  - Solution overview, updated since 106 to cover updates to version selection and schema comparison drafts.
[RW]

  *   Not planning to say much on this at all, at this stage.


0) requirements draft - Joe
  - Already adopted.
<ietf-107-netmod-yang-schema-comparison.pptx>_______________________________________________
Netmod-ver-dt mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/netmod-ver-dt