@d For better or worse, plain language explanations are expected in RFCs. I think the reasoning is that (a) it should be simple enough so that this is not a problem (b) we can distinguish between "we just happened to handle it in this way" vs "any compliant implementation has to handle it in this way" (c) we'd anyway want to call out subtle parts using plain language, so you wouldn't want to implement the thing based on some sort of pseudocode alone. At least IETF's editors complain (probably rightly) a lot about specifications being simple enough so that a typical programmer, with no expertise in the area, can implement them.

If I completely missed your point, then I would very much appreciate a more verbose description of the failure mode that you see.

@chjara@mk.absturztau.be