Re: [netmod] How to constrain a leaf to a read-only list of supported values?

[Vladimir Vassilev <[e-mail]>]

One alternative for getting rid of unsupported "identity"-s is to use 
"if-feature" as sub-statement which is now supported in YANG 1.1. This 
would be an elegant solution for the example with the 3 compression 
methods. However for standard models one will either have to live with 
the many irrelevant options presented to the user or use some extension 
similar to 
https://tools.ietf.org/html/draft-vassilev-netmod-yang-direct-must-augment-ext-00 
but for identities instead of "must" statements. Then clients and 
servers supporting it will be able to take the light out of the 
"identity"-s the implementation does not support by augmenting those 
with relevant if-feature statements.

On 09/01/2016 09:48 AM, Balazs Lengyel wrote:
> Hello,
>
> The proposed iana-entity.yang seems to take the same approach: one 
> file defining 28 identities.
>
> And I share all your concerns about iana-if-type.
>
> Balazs
>
>
> On 2016-08-31 14:16, Vladimir Vassilev wrote:
>> On 08/31/2016 12:38 PM, Ladislav Lhotka wrote:
>>>> On 31 Aug 2016, at 11:10, Vladimir Vassilev 
>>>> <[e-mail]> wrote:
>>>>
>>>> If you design your models using identityref and define the 
>>>> identities in separate modules e.g. compression-zip.yang, 
>>>> compression-gzip.yang, etc. you can just chose not to load the 
>>>> particular YANG models containing the identities not supported when 
>>>> your device starts.
>>> Right, and I have proposed this approach several times in the past. 
>>> However, some people prefer that the modules defining identities 
>>> mirror IANA and similar registries. In the case of 
>>> iana-interface-types it also means that implementations have to deal 
>>> with obsolete, obscure and experimental interface types that happen 
>>> to be in the IANA registry but nobody will ever want to use.
>>>
>>> Lada
>> +1
>>
>> The 275 identities defined in iana-if-type.yang appearing as possible 
>> /interfaces/interface/type tab completion options in a YANG aware cli 
>> or drop-down menu in gui is annoying and stands out as an obvious 
>> problem.
>>
>> It is not late to split the file. No standard RFC YANG model includes 
>> iana-if-type.yang yet. The actually referenced identities in current 
>> drafts is less then 16 (grep-ing in my known YANG model archive) 
>> {ethernetCsmacd, l2vlan, ieee8023adLag, ifPwType, pos, atm, 
>> atmSubInterface, sonet, otnOtu, frameRelay, bridge, 
>> macSecControlledIF, fastdsl}
>>
>> If not single instance per file maybe dividing the file into 
>> categories so if your device is atm aware you import 
>> iana-if-type-atm.yang and get {atm, atmSubInterface}.
>>
>> However we can probably agree the iana-if-type.yang exception is not 
>> a valid excuse for new models like the one in the example where there 
>> are 3 compression methods to not modularize the identity definitions 
>> into separate files and not load identities the implementation does 
>> not support but instead resolve to workaround solutions.
>>
>> Vladimir
>>
>> _______________________________________________
>> netmod mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/netmod
>>
>