Hi Mahesh,

> I think the reason that Rob is suggesting that IANA defer picking up
> any module, including IANA module after IESG approval, is to address
> some of the questions you have regarding the steps in Section 5.2. If
> IANA were to defer pulling out the IANA module from the I-D, it will
> allow for all the edits, formatting, validation to happened on them,
> just like they happen on IETF modules. It will also address your
> question on what happens when edits are applied after IESG approval
> but before publication.

I wouldn't argue that we post the IANA-maintained module at approval, since 
some aspect of it could theoretically change before publication.

But my impression, like Sandy's, is that the RFC Editor wouldn't have any 
reason to touch text that won't be published in the RFC. I was under the 
impression, though, that modules based on IANA registries are generated by a 
script, and unsure as to what kind of editing might be required if there's 
something wrong with script output that isn't caught by the authors. Our 
experience with YANG is limited enough that I wasn't sure what the possible 
categories of things that could go wrong might be.

<snip>

> Manually editing and adding is one option. But I believe that there
> was a script (from Kent?? ) that could be used to update the module
> with new entries in the registry.

We ran the script Kent included in RFC 9644 (and possibly a similar script in 
at least one other RFC?), but ran into two issues: an unanticipated reference 
type (a draft string associated with a recent registration) that required an 
edit to the script itself, and a bug that resulted in several lines having 
indentation issues. The next time we were asked to use a script from a 
document, the script referred us to a private Github, which wasn't OK  for us. 
I was under the impression that authors have used that script to generate 
IANA-maintained modules since then, but I'm not sure why I think this.

If authors have a bespoke script that they've used, or a general one, I wonder 
if it makes sense to have IANA run it at publication instead of asking the 
authors run it again. If it's trivial, then we won't have to go back to the 
authors to ask them to confirm that the content is correct; we can just 
validate and post. On the other hand, if there's a chance that something could 
go wrong, as with the RFC 9644 script, it's not clear to me that it makes sense 
to have IANA (distinctly non-expert in YANG) do it in the first place. 

If it doesn't make sense to have the authors re-create the module (assuming 
registry updates), at publication or earlier, we should probably just manually 
update. 

thanks,
Amanda

_______________________________________________
netmod mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to