Bad checksum frames should be ignored

[Teufelchen1]

Currently, configuration frames with a bad checksum are reported as Error::BadFraming where the rfc states they should be silently ignored.

[chrysn]

I'm not sure this really needs a code change. Error::BadFraming could documented as "should be silently ignored, but may be relevant for link quality assessment", shifting responsibility but also not discarding possibly useful information.

[Teufelchen1]

Yes, that looks like the right thing to do here.

Honestly I would like to blame it on the draft. CoAP messages with incorrect CRCs are silently discarded. Why? I am building a parser, let me inform my user of what is going on. It's none of the drafts business to tell me what to do outside the parsing. I think the authors intention was to say "there is no recovery mechanism, the message is just lost". In that case that sentence should be changed to something like: CoAP messages with incorrect CRCs should be discarded. Slipmux does not provide recovery nor retransmission.

[chrysn]

Well, Internet Drafts or RFCs rarely take a stance on the separation between library implementation and application, and tend to describe the whole. Still, your expression is what I'd prefer to be in the draft. We tend to give examples for "should" lines nowadays, that could be should be discarded (except, for example, to keep statistics).