Time to bring this bug report back to light. This time with Exim 1.80.
The example below, with directors using 'unseen' TESTS out fine with
-bt, but does NOT deliver for both directors. So, try the example
config listed in the prior message (below) and try to actually get both
directors to deliver on the address.
I'll start digging through the code for where this happens.
- T
On Thursday, December 04, 1997 4:59 AM, Troy Cobb wrote:
> I just downloaded and did a clean compile of 1.750, and the bug goes
> away. So, the difference between our setups appears to be that 1.750
> works, and my 1.73 doesn't. :)
>
> A quick diff looks like there were changes in direct.c between these
two
> releases.
>
> - T
>
> On Thursday, December 04, 1997 4:20 AM, Philip Hazel wrote:
> > On Thu, 4 Dec 1997, Troy Cobb wrote:
> >
> > > However, if the director is also set as unseen, to allow the
address
> it
> > > is directing to be tried by other directors in line, then the
> address
> > > that is passed along as unseen is the STRIPPED local_part, not the
> > > original. My expectation would be that further directors would be
> given
> > > a shot at the prefixed/suffixed address, but this original address
> does
> > > not seem to be being pushed back onto the director stack, instead
> being
> > > replaced by the stripped address.
> >
> > Your expectation is entirely correct.
> >
> > > Here's a quick example:
> > >
> > > director_one:
> > > unseen,
> > > prefix=local-,
> > > transport=local_delivery,
> > > driver=localuser;
> > >
> > > director_two:
> > > prefix=local-
> > > driver=aliasfile;
> > > file=/path/to/special/aliasfile/for/additional/addresses
> >
> > I tried this exact example (except that I had to add
> "search_type=lsearch"
> > to the second director).
> >
> > > In this case, an address of "local-somebody@???" will get
> > > matched by director_one, which sends it to local_delivery. In
> addition,
> > > it is marked as unseen and sent along to director_two, which WILL
> NOT
> > > MATCH the address, because the address being sent along is just
> > > somebody@???.
> >
> > This did not happen for me. The -d9 output was:
> >
> > address local-ph10@???
> > local_part=local-ph10 domain=xoanon.csi.cam.ac.uk local=1
> > >>>>>>>>>>>>>>>>>>>>>>>>
> > directing local-ph10@???
> > stripped prefix local-
> > calling director_one director
> > director_one director succeeded for ph10
> > transport: local_delivery
> > address local-ph10@???
> > local_part=local-ph10 domain=xoanon.csi.cam.ac.uk local=1
> > >>>>>>>>>>>>>>>>>>>>>>>>
> > directing local-ph10@???
> > stripped prefix local-
> > director_one director skipped: previously directed local-
> > ph10@???
> > stripped prefix local-
> > calling director_two director
> > director_two director: file =
> > /path/to/special/aliasfile/for/additional/addresses search type = 48
> > file=/path/to/special/aliasfile/for/additional/addresses query=ph10
> >
> > > Of course, this could all be by design, and I'm just
misinterpreting
> the
> > > documentation. :) I'll look into the code later to determine the
> place
> > > this is occuring.
> >
> > I was trying this with the latest code, but I don't think there has
> been
> > any change in that area recently.
> >
> > So: What is different between your setup and mine??
>
>
>
> --
> *** Exim information can be found at http://www.exim.org/ ***
--
*** Exim information can be found at
http://www.exim.org/ ***
