For better or worse, I'm trying to get my attribution format to match
Outlook (sorry!).  I can nearly get it, but one thing that seems
missing is differentiating "To" and "CC" receipients.  The "%T" format
contains all recipients:

   T - full names of the recipients of the message, in a comma-separated list
       If a full name cannot be found, the corresponding address is used
       instead.

Using this the attribution is incorrect since I'd list all recipients
as "To:".  Is there a format to get only "To" and only "CC"
recipients?  That way I could create an attribution that accurately
represent the attribution.

I'm using VM 8.2.0b.

thanks
-Kyle

Hi Kyle, what is the Outlook attribution format?

I think the original design of the specifiers was probably not ideal.  

If you can look through the manual section on the Summary Format and suggest
a solution, I can implement it.

Cheers,
Uday

Sorry for the late reply on this.  I'm noticing two hurdles in trying
to conform to Outlook attribution.  First off, here is the Outlook
attribution format:

From: [author name]
Sent: [date]
To: [to field recipient full names, semi-colon separated]
Cc: [if any CC recipient full names, semi-colon separated]
Subject: [subject]

I can get pretty close to that with:
(setq vm-included-text-attribution-format "\n:From: %F\nSent: %w, %m %d, %y %h\nTo: %T\nSubject: %s\n\n")

The first thing would be separating the "To:" recipients from the
"CC:" list.  The %T field groups them all together.  If I could supply
an attribution format that has the CC list conditionally present (ie-
do not display if no CC recipients) that would be cool, but I could
get by without that behavior.  I can live without that, I'd really
just like to breakout the CC list from the To list.

The other thing is the %T field uses a comma separated list which
makes the Outlook name format of "last, first" difficult to parse with
multiple recipients (Outlook does not surround names with quotes).  So
if I reply to "Smith, Bob" and "Doe, John", that will produce a
recipient list like: "Smith, Bob, Doe, John" (again, the quotes are
not in the actual email).  This is confusing since everything is comma
separated.  If I had the option of speciyfing the separator that would
be one way of doing it.

Looks like a lot of this stuff is in or around vm-su-do-recipients.  I
haven't played around with that yet.

So what I'd like to do is have something like:

(setq vm-summary-fields-concat-string ("; ")
(setq vm-included-text-attribution-format "\n:From: %F\nSent: %w, %m %d, %y %h\nTo: %O\nCC: %C\nSubject: %s\n\n")

That would introduce a new variable (vm-summary-fields-concat-string)
to override the default ", " concat string.  Then I used %O to
identify the To: field and %C to identify the CC recipients.  The %O
is a but unfortunately named with %T taken.

-Kyle

On 10/6/2012, Uday Reddy wrote:

The included-text-attribution-format was never meant for producing elaborate
headers.

Why don't you use vm-included-text-headers, which is what is meant for doing
this kind of thing?

In any case, in the trunk version of VM, I have stripped out the CC
recipients from %t and %T specifiers, and added new specifiers %r and %R for
all recipients.  There are independent reasons for doing this change.  (I
think that in the old days, there weren't enough fields in the cached data
header to keep 'to' recipients separately.  It was an unnecessary
limitation.)

Cheers,
Uday

re: why not use vm-included-text-headers?  Obviously I'm trying to be
difficult.  :)

Actually, I wasn't aware of that variable.  Setting it has solved my
attribution problem.  Thanks Uday!

On 10/22/2012, Uday Reddy wrote: