id2xml output is usually very far from a reasonable approximation of author intent, much less valid v3 rfcxml. #12
Description
See the thread at #11.
My understanding is that the RPC currently spends a lot of time correcting the misguesses that id2xml makes, and struggles with turning artwork back into sections.
It will output malformed dates that will cause xml2rfc to crash (which is also a separate issue that should be dealt with at xml2rfc).
We either need to produce a better text to xml converter.
I had a much easier time converting the text to kramdown, fwiw.
(But, the RPC has a lot of manual fixup for submitted kramdown based xml as it avoids trying to build things like dictionary lists, and puts them in artwork at the moment as well).