> Rather that re-tread worn ground, how about a considered reply to my
> suggestion that all headers be known to nmh and the user having to add
> to that list,
Is "user having to add to that list" in your original proposal?
I'm missing it. Is it a profile entry to allow the user to add
known header names? If that's the case, I don't like maintenance
responsibility, even if it is the user's. Though I think I'm missing
> or pass it as the value of a `wire' header. :-)
I don't feel as strongly about your proposal overall. The
"directive: header: value" syntax for unknown headers is different
from usual headers/pseudoheaders, and that's both good and not so.
But more important, we'd have to maintain a list of known headers.
I don't think that's as easy as it sounds. It doesn't change
often, but it does, and would introduce incompatibilities for
users who use different versions of nmh (as in home and work). We
sometimes insist that users use previous versions at their own peril,
but I'd rather minimize that, given the slow release cycle of nmh.
Nmh-workers mailing list