dovecot - Feb 2019 - smtp_address_cmp does not ignore case of localpart

Hi,

in src/lib-smtp/smtp-address.c line 759 the localpart is compared with 
null_strcmp, which does not ignore the case. While this is the correct 
behavior (according to the RFC) it causes problems in real life.

In my case the vacation notices are not sent, if someone writes to 
First.Last at ruhr-uni-bochum.de because :addresses contains (only)
first.last at ruhr-uni-bochum.de.



Regards,
Andreas

-- 
! AJ3630-RIPE @ RUB8-RIPE                        Network Operation Center !
! Dezernat 5.I                                   Ruhr-Universitaet Bochum !
! The amount of energy necessary to refute bullshit is                    !
! an order of magnitude bigger than to produce it.  -- Alberto Brandolini !

On 11/02/2019 22:43, Andreas Jobs via dovecot wrote:
> Hi,
>
> in src/lib-smtp/smtp-address.c line 759 the localpart is compared with
> null_strcmp, which does not ignore the case. While this is the correct
> behavior (according to the RFC) it causes problems in real life.
>
> In my case the vacation notices are not sent, if someone writes to
> First.Last at ruhr-uni-bochum.de because :addresses contains (only)
> first.last at ruhr-uni-bochum.de.
>
>
>
> Regards,
> Andreas
>
In practice if you convert the username part to lower case in the MTA
then you will accept mail with mixed case but only have to deal with
lowercase in Dovecot.

John

On Tue, Feb 12, 2019 at 08:12:43AM +0100, John Fawcett via dovecot
wrote:
>On 11/02/2019 22:43, Andreas Jobs via dovecot wrote:
>In practice if you convert the username part to lower case in the MTA
>then you will accept mail with mixed case but only have to deal with
>lowercase in Dovecot.
Yes, but the vacation action (also) compares :addresses against the 
header to|cc|sender... And changing the headers is a bad idea.

Maybe my subject line was a bad choice :-)


--Andreas

-- 
! AJ3630-RIPE @ RUB8-RIPE                        Network Operation Center !
! Dezernat 5.I                                   Ruhr-Universitaet Bochum !
! The amount of energy necessary to refute bullshit is                    !
! an order of magnitude bigger than to produce it.  -- Alberto Brandolini !