Re: [Nmh-workers] Changes to forw(1)

[Paul Fox]

ralph wrote:
 > Hi Paul,
 > 
 > > > > +1.  The `Forward' header is grabbing another one for nmh's use,
 > > > > in addition to the existing `Attach'.  Should we be using
 > > > > `Nmh-Forward' if the user isn't likely to have the hassle of
 > > > > typing them most of the time?
 > > > 
 > > > Sigh.  I think when we hashed this out last time, the (rough)
 > > > consensus was that not puttting in a "Nmh-" prefix was fine.  Attach
 > > > had prior art (I think mutt used it), and Forward seems to be
 > > > similarly named.
 > > 
 > > i vote for presenting the user with a user-friendly component name.
 > > if conflict is an issue, could we make the names of these "special"
 > > headers tuneable via a profile entry?
 > 
 > Isn't that just another level of code, documentation, and grokking by
 > the user when Nmh-Forward would just sit there, be spotted and
 > understood by the user, and typically left alone.  Either the message

well, speaking "typically", this would also apply to "Forward".

adding the "Nmh-" prefix would create the very first such header.  as i
understand it, we already have 4 nmh-specific headers:  Resent,
Forwarded, Replied, and Attach.  i don't see much value in creating a
unique prefix for the fifth.

 > numbers might be edited, or the whole line deleted.  I actually think
 > it's an advantage to see that this is a Nmh header and not one that may
 > have general purpose interpretation.

thinking about this further, i think i might rather teach Attach about
mh message numbers and sequences than add a new Forward header. 
Attach is already examining its file arguments to decide how to a
attach any given file -- teaching it recognize message specifiers
isn't a big stretch.  this would clearly lead to these two
invocations having different results:
    Attach: cur
    Attach: ./cur

but i think i'd be okay with that.

paul
=----------------------
 paul fox, address@hidden (arlington, ma, where it's 45.5 degrees)