On Jun 9, 2014, at 1:30 PM, J. Gomez <jgo...@seryrich.com> wrote: > On Monday, June 09, 2014 8:01 AM [GMT+1=CET], Matt Simerson wrote: > >> On Jun 8, 2014, at 10:32 PM, Brandon Long <bl...@google.com> wrote: >> >>> The message is already corrupted, or there wouldn't be a problem to >>> be solved. >> >> When the message arrives at the list, it's unlikely that it's already >> corrupted. What has been described is corrupting the From header by >> the same entity that is about to break the DKIM signature by altering >> the the message. This should be called the "break it worse" method. > > So, when the MLM relaying the message adds a subject tag, that alteration is > a welcomed "decoration" - but when it changes the mailbox in the Header-From > to itself, it is an unwelcomed "corruption".
> I can understand the welcomed vs unwelcomed thing, but I do not agree with > calling the alteration "decoration" in one place but "corruption" in the > other. > > Loading the language in such a way is asking for a given conclusion even > before the debate has started. That's not fair (I'm not predicating that from > you, Matt, just talking in general terms). You'll note that I didn't introduce either the terms decoration or corruption. I described *both* message alterations as breakage. In the context of DKIM, and consequently DMARC, I think that's a perfectly fair description. Breaking DKIM has been a liberty that mailing lists have thus far been able to get away with. For a time, DMARC early adopters danced around it the DKIM breakage problem by splitting their message streams into "transactional" vs "personal." For some, that recently changed and now there are consequences for mailers that break DKIM signatures. I think that's a *very* fair assessment. Matt _______________________________________________ dmarc mailing list dmarc@ietf.org https://www.ietf.org/mailman/listinfo/dmarc