Developers of Sympa

[David Verdin <address@concealed>]

Hi Micah,

Micah Anderson a écrit :
> * David Verdin <address@concealed> [071211 04:39]:
>
>   
> > That message is a type of Feedback report unknown to Sympa. I'll 
> > investigate why.
> >     
>
> Interesting, if I look at that message and compare it to the ARF RFC
> examples[1], it appears to be valid ARF format[1] to me, unless I am
> missing something, Sympa is supposed to understand this format[2].
>
> Micah
>
> 1. http://wordtothewise.com/resources/annotatedsample.html
> 2. http://www.sympa.org/wiki/manual/bounces#arf
>   
Yes it is a valid ARF. I compared the RFC to what we did in Sympa and it 
appears we were wrong. We had a test in which we assumed the report was 
of known kind if:
- the "Feedback-Type" header value appeared in a list of known strings
- AND "Original-Rcpt-To" was set.
Which was wrong, because "Original-Rcpt-To" is an optional field in the RFC.

In your message file, "Feedback-Type" had the value "abuse", but 
Original-Rcpt-To was missing. Therefore, we discarded the message.

I fixed this in the development branch of Sympa. We now support normally 
the internet draft (version 0.3): we consider unknown the values 
undefined in the draft. However we notify the listmaster if, lacking an 
Original-Rcpt-To, we can't do anything.

Regards,

--
David Verdin
Comité réseau des universités