On Sat, 19 Oct 2013 20:33:36 -0400, Ken Hornstein said:
It is my reading of RFC 2046 that text parts should always represent line breaks as CR LF. And I believe that we get this wrong when the text is encoded as base64. While we will never encode text parts with base64 (although that might change in the future), I also think we get them wrong on display or storing. Does everyone agree that text parts encoded in base64 should be decoded and converted to Unix line conventions? It looks like this would be as simple as never outputting a CR when decoding a text part.
I admit being insufficently caffienated at the moment to know whether this would break S/MIME or PGP, but ISTR that the signatures are done over a canonical CR/LF. So we probably need to treat decoding and converting to local line-end conventions as two separate steps.
pgp89Hkyjddmo.pgp
Description: PGP signature
_______________________________________________ Nmh-workers mailing list Nmh-workers(_at_)nongnu(_dot_)org https://lists.nongnu.org/mailman/listinfo/nmh-workers