On Wed, Jun 23, 2010 at 14:05:32 +0300, Peter Pentchev wrote: > On Tue, Sep 15, 2009 at 07:08:25PM +0200, Tino Keitel wrote: > > Hi, > > > > sorry, my observation was wrong. My mutt always generated a From: line > > (with "realname <u...@host>"), and exim used to translate this into the > > real address. Dma seems to only use virtusertable if no From: line is > > present. > > > > I fixed my mutt configuration to not generate a From: line, and now > > virtusertable is used. > > So... erm, sorry, but I'm a bit unclear - do you think this is still > a problem? :) > > After a quick look in the dma code, I don't think it tries to modify > the From: header in any way. What it does is change the envelope > sender to the virtual user; I guess Simon's thinking was along the lines > of "the From: header should be set by the MUA; if somebody wants virtusers, > he'll have told his MUA to set the proper From header". > > On the other hand, the behavior you've seen is completely explained by > the fact that if the MUA sends a message *without* a From: header at all, > dma will *add* a From: header with the e-mail address of the envelope > sender. Since this is done after it has substituted the envelope sender > address with the virtual one, It All Just Works(tm) :) > > I don't think it would be dma's job to *modify* the From: header if a MUA > has set one. IMHO, what it does right now - always modify the envelope > sender (if a virtuser is present) and always add a From: header if the MUA > has not - should be the extent of its interfering with the user's message.
Hi, yes, I just had to realize that dma behaves different compared to exim. I think this bug can be closed. Btw., I'm using 0.0.2010.06.17-1 for a few days now and it looks fine. Regards, Tino -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
