Re: Last Call: draft-dusseault-impl-reports (Guidance on Interoperation and Implementation Reports) to BCP

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Overall, I like this document and support it going forward.

One thing it doesn't mention (and did come up when I was an AD) is the
following.

The goal of interoperability testing is to demonstrate that our
specifications are good. I.e., that the text in the final RFC is
complete and can be implemented.

We sometimes have WGs where folk are implementing while the
specification is being developed. What they end up implementing is not
(strictly speaking) based on the text in the final RFC, but may be
implemented "from the mailing list". In such cases, implementation
issues that come up may be resolved on the mailing list (or during
interoperability tests where the implementors just work it out), but
somehow the final spec is still not clear on the point at issue.

It would be good to point out in the document that the intention is
that implementations be based on the final spec. In cases where there
is reason to suspect this isn't the case, and that the implementation
is not actually based on the RFC itself, some caution is advised is
using those implementations to demonstrate readiness to advance to
draft. I.e., insist on multiple implementations, rather than just
using 2 (the absolute minimum required to go to draft).  Clearly some
judgement is needed here, as we don't want to make it too hard to go
to draft either.

Thomas
_______________________________________________

Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf

[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]