Hello Einar,
I answered  most comments in other mails. But see one below!
regards Balazs

On 6/25/2018 3:32 PM, Einar Nilsen-Nygaard (einarnn) wrote:
  • Server capabilities as mentioned by Alex; specifically, it’s not clear to me yet how the correct versions of the models used inside instance data are defined.
  • BALAZS: 
    I do not want to make a strong connection between the instance-data-set and the individual model's revision/version. Obviously the same instance-data-set can be used with many module versions even if those versions are incompatible. 
    Without semver in the YANG Modules (YAMs) I see no formal solution at all. Connecting the instance-data-set to a strict revision-date of the YAM would be extremely tedious and require continuous updates.

    IMHO the problem is similar to the import-by-version in the module-update/versioning draft.  Maybe I can reuse that solution. E.g. Include a set of used YAMs and allowed semantic-versions as metadata under container instance-data-set.

    My proposal is to list this as an open issue and see what the versioning draft comes up with. Still I fear this might end up as a warning and a statement that it must be solved by other means.
    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