* Vincent Lefevre <[email protected]> [07-26-16 06:52]: > On 2016-07-26 11:22:43 +0200, Oswald Buddenhagen wrote: > > On Tue, Jul 26, 2016 at 10:26:16AM +0200, Michael Tatge wrote: > > > On Mon, Jul 25, 2016 04:39PM +0200 Olaf Hering ([email protected]) muttered: > > > > Am 25. Juli 2016 14:52:27 MESZ, schrieb Michael Tatge > > > > <[email protected]>: > > > > > > > > >No. Sending from the command line, in scripts, the error might come up > > > > >later during delivery,... > > > > > > > > Have you actually tried it with smtp_pass= ? There is no interaction > > > > and as a result no password and as a result no mail gets created. > > > > > > using the internal smtp agent is a special case, mutt's scope is wider > > > then that. > > > > > uhm, so? when delivery is done via sendmail and fails later on, it was > > successful from mutt's perspective. > > Agreed! > > If I use > > set sendmail=/bin/false > > this is indeed wrong: each time I try to send the message, a copy of > the message is stored in $record, which gives the false impression > that the mail has been sent.
so mutt becomes responsible for what sendmail does? I think not. If mutt's internal's fail to complete an action, it should provide notice, but not an action after handing off. If mutt hands off a mail to an app to provide a special format/action and that "app" fails, the "app" would provide notice but mutt is not responsible. If in this case sendmail or it's representative fails and provided mutt notice, then mutt should act on that notice. -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://linuxcounter.net
