• Andrew C Aitchison via Exim-dev [2024-12-15 23:19]:
> On Sun, 15 Dec 2024, Exim Bugzilla via Exim-dev wrote:
>
> > https://bugs.exim.org/show_bug.cgi?id=3071
> >
> > --- Comment #3 from exim-bugs@??? ---
> > (In reply to Jeremy Harris from comment #2)
> > > The cutthrough facility uses the verify callout machinery (recipient, with
> > > the use_sender option) to place the ongoing connection.
> > >
> > > Changing the envelope sender would be a breaking change. I think supporting
> > > this will need an option on the ACL control=cutthrough_delivery -
> > > perhaps "/sender=transport", with "original" signifying the existing
> > > behaviour of using the received envelope-from?
> >
> > Docs don't mention anything about the sender address, do they? There are some
> > limitation mentioned, but non of them apply to the sender address.
> >
> > I'd assume that people using cutthrough_delivery with SRS would expect SRS to
> > be applied with cutthrough_delivery, and myself I'd certainly appreciate a way
> > to set envelope sender upon cutthrough_delivery.
>
> I would not expect to use SPF or SRS with cutthrough-delivery,
> just as i would not expect a secondary MX to do SPF.
But if you _did_ use SRS, would you expect cutthrough to use the
rewritten envelope sender?
I guess it depends on the use case.
SRS is mostly relevant for doing forwarding so some external sites, such
as to gmail in my case, and gmail in particular requires valid SPF or
DKIM for all senders (for bulk senders, I think, they require SPF, DKIM
and DMARC).
> > I like the proposed idea. Would sender=transport also take into consideration
> > errors_to router settings (as those would affect the transport's return_path)?
> >
> > But this bug can be fixed simply by documenting that envelope sender does not
> > change.
>
> although, earlier you said:
> > FWIW more options could be affected...
I probably did, yeah, but I had forgotten about that. Yes, debug_print
is also not used. My point is that bug fix could be that documentation
should mention that cuttrhough_delivery is not a normal delivery, and as
a result, certain options, such as return_path and debug_print, are not
used. Then the bug could become a wishlist item, though.
--
## subscription configuration (requires account):
##
https://lists.exim.org/mailman3/postorius/lists/exim-dev.lists.exim.org/
## unsubscribe (doesn't require an account):
## exim-dev-unsubscribe@???
## Exim details at
http://www.exim.org/
## Please use the Wiki with this list -
http://wiki.exim.org/
