On 6/26/2018 5:44 PM, Juergen Schoenwaelder wrote:
On Tue, Jun 26, 2018 at 05:31:27PM +0200, Balazs Lengyel wrote:
Hello Juergen,

Sorry the wording was misleading. I want these capabilities both as state
data AND as instance-data-files, because
there is a need to know this information before you ever see the real
network node. How about the following?

"YANG servers SHOULD document server capabilities that are available via Netconf/Restconf (as YANG defined data) also using instance-data-files."

I am against this. I think we should standardize the format and not
how the format is used. And it remains unclear why this would be a
SHOULD or how one implements this SHOULD and in which cases one can
ignore it. Again, I do not like to mix specification of mechanisms
with the specification of policies how mechanisms are to be used.

/js
BALAZS: The current draft really says two things:
  1. this is the format for instance data
  2. instance-data SHOULD be used to document server capabilities
IMHO both are important. I fear that if we exclude the second one from this draft it will never get documented, even though it is an important statement. Or would you support a 2 page draft stating that you SHOULD document ietf-yang-library content with instance-data?
regards Balazs
-- 
Balazs Lengyel                       Ericsson Hungary Ltd.
Senior Specialist
Mobile: +36-70-330-7909              email: [email protected] 
_______________________________________________
netmod mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/netmod

Reply via email to