Re: [netmod] Text in import to indicate whether a module is needed as import-only or as implemented

Andy Bierman <andy@yumaworks.com> writes:

> On Wed, Dec 18, 2019 at 6:08 AM Balázs Lengyel <balazs.lengyel=
> 40ericsson.com@dmarc.ietf.org> wrote:
>
>> Hello Mahesh,
>>
>> I was asked by the group to include in each import statement whether the
>> imported module is needed as import-only or as implemented. IMHO
>> netmod/netconf group should agree on some standard text for model designers
>> to use. Maybe the text proposed below can be used everywhere. The sentence
>> starting with Revision yyyy-mm-dd  may not always be needed.
>>
>>
>>
>>
>>
>> I propose the text
>>
>>
>>
>>   import ietf-netconf-acm  {
>>
>>     prefix nacm;
>>
>>     description
>>
>> *       "The module ietf-netconf-acm is OPTIONAL to implement.";*
>>
>>   }
>>
>>
>>
>>   import ietf-yang-library {
>>
>>     prefix yanglib;
>>
>>     description *"The module ietf-yang-library is REQUIRED to*
>>
>> *      be implemented. Revision 2019-01-04 or a*
>>
>> *      revision derived from it is REQUIRED.";*
>>
>>   }
>>
>>
>>
>> Regards Balazs
>>
>>
>>
>> P.S. In Yang-Next this could be a candidate for a formal substatement
>> instead of a description text.
>>
>>
>>
>
>
> IMO this is a bad idea.
> The IETF keeps adding CLRs to YANG that make it harder to use over time.
> The YANG syntax defines what must be implemented from the imported module,

I don't see how YANG syntax defines this. If a module imports ietf-netconf-acm, it could be because

- it just uses a typedef, such as "node-instance-identifier", and then
  ietf-netconf-acm needn't be implemented (but can be),

or

- it augments ietf-netconf-acm, which makes sense only if the latter
  module is implemented.

It it the YANG library that specifies whether a module is implemented or not, but the "import" statement itself doesn't tell you anything.

> so extra description-stmts just add more busy work for writers and
> reviewers.
> The text specifying the revision overrides (and breaks) YANG rules.
> YANG 1.1 has specific import-by-revision that cannot be changed with a
> description-stmt.

This is one of the problems that the semver effort aims to solve. What happens quite oftem these days is that a module is written for NMDA and thus needs NMDA versions of modules to be imported. But specifying an exact revision is really a bad idea.

Lada

>
> Andy
>
>
>
>>
>>
>> *From:* Mahesh Jethanandani <mjethanandani@gmail.com>
>> *Sent:* 2019. december 18., szerda 3:20
>> *To:* Kent Watsen <kent+ietf@watsen.net>
>> *Cc:* Balázs Lengyel <balazs.lengyel@ericsson.com>
>> *Subject:* Re: [netconf] New Version Notification for
>> draft-ietf-netconf-notification-capabilities-08.txt
>>
>>
>>
>> Hi Balazs,
>>
>>
>>
>> Additionally, it would be important to address some of the normative text
>> in the module. Specifically, we were looking at the following description
>> statements:
>>
>>
>>
>>   import ietf-yang-push    {
>>
>>     prefix yp;
>>
>>     description
>>
>>       "This module requires ietf-yang-push to be implemented for the
>>
>>         two subscription-capabilities containers.";
>>
>>   }
>>
>>   import ietf-yang-library {
>>
>>     prefix yanglib;
>>
>>     description "This module requires ietf-yang-library to
>>
>>       be implemented. Revision 2019-01-04 or a
>>
>>       revision derived from it is required.";
>>
>>   }
>>
>>
>>
>> The requirement in the description statement feels and smells like words
>> one would use to signify requirements in a specification. However, you are
>> not using any of the words like REQUIRED etc. to describe it. Why is it?
>>
>>
>>
>> Mahesh Jethanandani
>>
>> mjethanandani@gmail.com
>>
>>
>>
>>
>>
>>
>> _______________________________________________
>> netmod mailing list
>> netmod@ietf.org
>> https://www.ietf.org/mailman/listinfo/netmod
>>
> _______________________________________________
> netmod mailing list
> netmod@ietf.org
> https://www.ietf.org/mailman/listinfo/netmod

-- 
Ladislav Lhotka 
Head, CZ.NIC Labs
PGP Key ID: 0xB8F92B08A9F76C67