On 3/3/20 7:47 PM, Stephen J. Turnbull wrote:
I think we should check if this is the same in Mailman 3, and change it if it is. If we have an attribute whose main function is to populate a header field, it should accept any valid value for that field (POLA). Unless there's a Very Good Reason[tm], in which case "Errors should never pass silently." :-)
I just filed <https://gitlab.com/mailman/mailman/issues/689> for this. It suggests parsing the explicit reply to address with email.utils.getaddresses instead of email.utils.parseadder.
Arguably, it shouldn't parse it at all and just copy it, but this creates potential issues if we aren't first stripping the original Reply-To: or if we're adding other items to Reply-To: due to DMARC mitigation.
-- Mark Sapiro <mark@msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan