Wen Bo Li

It's mainly intended for pure OpenConfig or non-pure (augmented) OpenConfig models where there is a partial conformance with the OpenConfig-style. I intend to add a generation flag: ```go compressOCNames =...

Ok, so we're explicitly saying that ygot will maintain generated-code backward-compatibility if the YANG is also changed in a backward-compatible way? I see field additions as the main example, so...

@robshakir Your comment on "OpenConfig only" gave me an idea that by segregating the name generation and looking at the originating module namespace of a node (I.e. whether the node...

Hi @rimpsh Thanks for sharing us your fix. There is currently a workaround for the problem you described, which is to call the function `ygot.PruneConfigFalse`. I will admit this is...

Apologies for the late reply -- If by "merge" you mean resolving any differences between the two leaves and merging their properties, then this is not something we intend to...

After some discussions with @greg-dennis where we considered "ordered-by" and implicitly indexing keyless lists, the current idea is we'd be essentially treating keyless lists as a leaf when it comes...

The complexity tradeoff with 2. is that it is impossible to know for a gNMI path deletion **which** of the keyless values is being deleted. This means that while it's...

The current proposal is that this is a low-priority feature since we would prefer gNMI `PROTO` encoding to not use JSON. https://github.com/openconfig/public/pull/750 contains a discussion ([link to conclusion](https://github.com/openconfig/public/pull/750#discussion_r1053625675)) as to...

By the way thanks for posting this contribution! This looks like something that can benefit others and overall this implementation looks good.

identities and enumerations are represented as strings per https://datatracker.ietf.org/doc/html/rfc7951, `2` is ambiguous.