[netconf] Re: WGLC for trace-ctx-extension-03

[Per Andersson <[e-mail]>]

Hi Roque,

On Thu, Nov 7, 2024 at 1:34 PM Roque Gagliano (rogaglia)
<[e-mail]> wrote:
>
> As can be seen below, the YANG module validation the recommended
> namespace value is
>
>     urn:ietf:params:xml:ns:yang:ietf-netconf-otlp-context
>
> This ties better to the YANG module name and the presented xmlns prefix
>
>     xmlns:ietf-netconf-otlp-context
>
> It is noted that the namespaces for
>
>      urn:ietf:params:xml:ns:yang:traceparent:1.0
>
> and
>
>      urn:ietf:params:xml:ns:yang:tracestate:1.0
>
> correspondingly should have longer names containing the YANG module
> name. However, this would be very long. Perhaps a shorter, but
> corresponding namespace such as
>
>     urn:ietf:params:xml:ns:yang:otlp-context-traceparent
>
> could be used? If so maybe the shorter "otlp-context" namespace can be
> kept instead of the recommended "ietf-netconf-otlp-context", which would
> signal relation?
>
> (Roque) This is great feedback but I believe we should not tie ourselves to OTLP, context is independent of the exporting protocol. We are updating the module for next release with “ietf-trace-context” (which would work for both RESTCONF and NETCONF but also for future protocols). Would that be ok?

Yes, good to me.


> Nits:
>
> * nc_trace as short title. Suggest using something more descriptive.
>   (Also noticed short illegible titles in in
>   draft-ietf-netconf-restconf-trace-ctx-headers and
>   draft-ietf-netconf-transaction-id.)
>
>
>
> (Roque) I am not sure of the process to change draft titles. Could you please advice as WG Chair?

I suggest you change this in the preamble of the RFC source.

https://github.com/netconf-wg/trace-ctx-extension/blob/main/draft-ietf-netconf-trace-ctx-extension.md?plain=1#L3

Perhaps something like

abbrev: NETCONF Trace Context Extension

(Please do this for draft-ietf-netconf-restconf-trace-ctx-headers as well.)


--
Per, as a contributor