Hi Peter!

Thanks for the revised -08 draft.  

I need help understanding how this revising is responsive to my DISCUSS 
feedback about the registry (now) in Section 12.3.

I see this block of new text:
==[ snip ]==
   When a new constraint is defined, the rule associated with that
   constraint MAY be inserted at any position.  Backwards compatibility
   is guaranteed because nodes which don't support the new constraint
   will not participate in an algorithm where the FAD specifies a
   constraint they don't support.

   The relative ordering of existing rules MUST NOT be altered.  Doing
   so has the potential to create backwards compatibility issues.

   Deletion of the rules MUST NOT be done.  Given that the rules are
   only used conditionally based on the information carried in the
   winning FAD, deletion of the rule is not necessary.

   Merging or repeating of the rules MUST NOT be done.

   There is no upper bound on the number of rules that the registry
   supports.
==[ snip ]==

A few comments:

(a) What does this have to do with IANA?

(b) Is this guidance to designated expert on evaluating new registrations?  It 
doesn't say so.  If it is, what does it mean to "modify ordering of existing 
rules"?  Is deleting "rules" removing their registration?

(c) RFC2119 shouldn't be used in the IANA considerations sections.  See 
https://datatracker.ietf.org/doc/statement-iesg-statement-on-clarifying-the-use-of-bcp-14-key-words//

Regards,
Roman



-----Original Message-----
From: Peter Psenak <[email protected]> 
Sent: Monday, July 21, 2025 4:09 AM
To: Mohamed Boucadair <[email protected]>; Roman Danyliw 
<[email protected]>
Cc: [email protected]; 
[email protected]; [email protected]
Subject: draft-ietf-lsr-igp-flex-algo-reverse-affinity-08

Warning: External Sender - do not click links or open attachments unless you 
recognize the sender and know the content is safe.


Mohamed, Roman,

I posted the new revision of the draft that should address your comments. 
Please have a look and let me know if you have any more comments.

thanks,
Peter
_______________________________________________
Lsr mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to