On 7/17/2014 3:30 PM, Todd Lyons wrote:
> On Thu, Jul 17, 2014 at 2:22 PM, Evgeniy Berdnikov <bd4@???> wrote:
>> On Thu, Jul 17, 2014 at 02:49:44PM -0600, rblue wrote:
>>> I can't tell if $acl_m_use_ses is or not equal to 1. I assume it is due to
>>> these lines:
>>>
>>> expanding: $acl_m_use_ses
>>> result:
>>> expanding: 1
>>> result: 1
>>>
>>> but I'm unsure why this is resulting as false:
>>>
>>> condition: eq{$acl_m_use_ses}{1}
>>> result: false
>>>
>>> Any ideas?
>> The $acl_m_use_ses is expanded to empty string, surely it is not 1.
>> Empty string may be converted to number 0, but never to 1, IMHO.
> Oops, my previous long-winded answer was slightly incorrect WRT the
> above two expansions. Evgeniy is exactly correct.
>
> That's why I suspect this is the router being called for address
> verification in the MAIL or RCPT acl, instead of the actual routing
> that would be used to process delivery of the message in the DATA acl.
>
> ...Todd
>
Sorry I am new with exim - easy to tell :). Are you saying that the
send_via_ses router is being called from one of the MAIL or RCPT acl and
therefore invalidating the router?
I guess a fix would be to add to the router's condition a check for the
acl_m_use_ses variable only after the DATA acl gets processed?
