Re: [Int-area] updated to draft-atlas-icmp-unnumbered-03

[Danny McPherson <danny at tcb.net>]

On Oct 1, 2007, at 12:40 AM, Joe Touch wrote:

> MUST is the directive to implementers of this specification, and is
> appropriate (I didn't check, but presumably it is paired with "MUST be
> ignored on receipt").

I did look for that, didn't recall seeing it (or any other error
handling, etc.., for that matter).

> Future specifications can set it to a different value and 'not  
> ignore' it.
>
> SHOULD would allow implementers to change the value under the current
> spec, which would render the field useless to future specs.

Yeah, you're right, I should have know that..

Thanks for the clue!

-danny


_______________________________________________
Int-area mailing list
Int-area at lists.ietf.org
https://www1.ietf.org/mailman/listinfo/int-area