Re: [netmod] WGLC on draft-ietf-netmod-rfc6991-bis-11

[tom petch <[e-mail]>]

From: Jürgen Schönwälder <[e-mail]>
Sent: 27 June 2022 15:08
On Mon, Jun 27, 2022 at 11:41:35AM +0000, tom petch wrote:
>
> Two additional thoughts.  Martin Durst has seen the YANG for language tags in the I2NSF I-D and says that it is 'complete overkill'.  Also, 10 IPR claims have been submitted against I2NSF I-D, albeit for older versions, with 'Possible Royalty/Fee' so I have gone off the idea of using anything from that WG.  I see that others use just a length constraint for the basic format which is hardly worth a type.
>

Yes, even a typedef like this

  typedef language-tag {
    type string;
    description "A language tag according to RFC 5646 (BCP 47).
  }

would do. Note that RFC 5131 defines a LangTag TC for the SMIv2 world.
Petterns are nice to have to detect wrong values "early" but once
patterns become very complex, the likelihood increases that they are
wrong.

<tp>

Weellll Francesca threw a DISCUSS on 16feb22 against -i2nsf-nsf-facing-21.  Paul produced -22 20mar22.  Francesca threw a further DISCUSS 22mar22 since the pattern in the I-D was not a valid regex.  Paul revised the I-D again and Francesca declared herself satisfied.

So regex are hard to get right!

What we have now is overkill (Martin Durst).

I recall seeing similar DISCUSS on YANG I-D from other WG so I expect this discussion to continue but see Jurgen's proposal above as the best so far (unless and until an AD says otherwise).

I had forgotten that SMI had tackled this problem - I should have guessed.

Tom Petch
/js

--
Jürgen Schönwälder              Jacobs University Bremen gGmbH
Phone: [phone]         Campus Ring 1 | 28759 Bremen | Germany
Fax:   [phone]         <https://www.jacobs-university.de/>