[ietf-dkim] [dkim-dev] need an interpretation of the base spec

[Murray S. Kucherawy]

On Tue, 3 Jun 2008, Tony Hansen wrote:
> Right now, I'm leaning towards thinking that #2 is correct. What say the
> rest of you?

Discarding FWS and tokenizing is probably a common implementation tactic, 
which favours #2.

I'm inclined to say it's an errata issue (and would offer #2 as the 
preferred implementation) unless someone can show precedent for resolution 
of something similar in the past.