Re: [arch-d] I-D Action: draft-iab-protocol-maintenance-08.txt

[Martin Thomson <[e-mail]>]

On Fri, Jul 15, 2022, at 07:09, Bless, Roland (TM) wrote:
> Please reread RFC1122 section 1.2.2.

You might be surprised to hear that I have read that text.

Joe's summary is right, "It IS about the fact that no specification is complete and the ways that different people interpret the same specification."

Yes, this talks about how - for security - you need to handle arbitrary inputs.  That's always been true; RFC 3552 does a better job of treating that.  No debate there.  But it's advice regarding the handling of unexpected error codes is just bad.  You've read my draft, I won't repeat the part where it explains how relying on robustness as the basis for extensibility is a terrible idea.

There were other points in your email I strongly disagree with.  I consider these superficial points.  The central point is that the very concept itself (as Joe stated it) is not a good basis for building a protocol ecosystem.