Re: [netconf] Adoption call for draft-kwatsen-netconf-http-client-server-04

[Martin Bjorklund <[e-mail]>]

Kent Watsen <[e-mail]> wrote:
> 
> > There's a thread in opsawg about renaming drafts...  Conclusion (I
> > think): keep the name of the draft, but change the title and name of
> > the module.
> 
> I saw but didn't read the thread.  Did it cover crossing from an
> individual draft to a WG draft (i.e., kwatsen-05 --> ietf-00)?

Yes.

> The
> draft name must change, right?  If this is a Datatracker history
> issue, is it understood that the replace-by behavior enables diffs to
> cross the individual to WG draft boundary, and it really doesn't
> matter what names are used?

Personally I don't care; just pointing out a discussion that may be
relevant.


> > If the name is too generic, and the issue is that the content is more
> > specific than the name suggests, then renaming can help resolving the
> > issue.
> 
> In principle, yes, but for this case, how is the "http" name too
> generic?

I think it is the name "ietf-http-server" that seems to indicate that
this a module that can be used to configure any HTTP server.  (See
below!)


> Do we have a single concrete reason that has been vetted?
> How can the draft explain to readers why the most obvious name wasn't
> used?  What is the distinguishing characteristic that potential
> consumers should use to ascertain if the modules are [in]appropriate
> for them?
> 
> As for the title and module name, it seems the simplest thing is:
> 
> 	OLD:
> 		title: 	Groupings for HTTP Clients and Servers
> 		module: 	ietf-http-client
> 		module: 	ietf-http-server
> 
> 	NEW:
> 		title: 	Groupings for RESTful HTTP Clients and Servers
> 		module: 	ietf-restful-http-client
> 		module: 	ietf-restful-http-server
> 
> But isn't there also an issue of names being overly specific?  Some
> may ask why the draft is limited to RESTful HTTP, being that it only
> configures endpoints and has nothing to do with if the application is
> RESTful or not.

I wonder if the names should be "ietf-http-server-groupings" instead?
(and same for tcp / ssh / tls, but not netconf / restconf).  We
already have some "-types" modules.  Or even "ietf-http-server-types",
if by "type" we mean "typedef and/or grouping".

This could also be a way to make the name less problematic.  It makes
it more obvious that these modules provide building blocks, rather
than a complete solution.


/martin