Re: What should go into 1.5.7?
On Wed, Jan 26, 2005 at 09:13:29AM -0600, Charles Cazabon wrote:
> TAKAHASHI Tamotsu <ttakah@xxxxxxxxxxxxxxxxx> wrote:
> >
> > Hmm, as Thomas said, ignore/unignore pair and alternates/unalternates
> > pair are inconsistent.
> > ignore x- ; unignore x-label
> > is similar to
> > grep some | grep -v ribosome
> > but
> > alternates '@.*\.domain'; unalternates '@not-mine\.domain'
> > doesn't work.
>
> Doesn't it? I wrote the notalternates patch to do almost exactly this, though
> the patterns I'm using match on bits of the local-part of the address rather
> than the domain. But I see no reason why it wouldn't work exactly as you have
> it above.
Well, sorry for my poor English.
With your patch,
alternates '@.*\.domain'; notalternates '@not-mine\.domain'
does work.
But I don't think we need notalternates. Because we don't need notignore.
==== "ignore" behaviour ====
While reading muttrc...
Mutt stores REs of "ignore" command.
Mutt stores REs of "unignore" command.
While reading a message...
First, mutt ignores headers matching REs of "ignore" command.
Mutt recalls headers matching REs of "unignore" command.
==== current "alternates" behaviour ====
While reading muttrc...
Mutt stores REs of "alternates" command.
Mutt removes REs which are exactly the same as strings of
"unalternates" command.
**** Note that here is only one variable stored in mutt's memory ****
While reading a message...
Mutt thinks the message is from you if it matches any of the REs.
**** Note that here is only one step. ****
==== ideal "alternates" behaviour ====
While reading muttrc...
Mutt stores REs of "alternates" command.
Mutt stores REs of "unalternates" command.
While reading a message...
First, mutt thinks the message is from you if it matches any of the
"alternates" REs.
But mutt changes his dicision if the message matches any of the
"unalternates" REs.
Got it? Anyway, Thomas will do the best choice.
He has already realized this inconsistence.
Please ignore me if my English is too broken to understand.
--
tamo