On Tuesday 20 October 2009 05:54:46 Murray S. Kucherawy wrote:
> Although the unit testing for libopendkim is fairly thorough, the unit
> testing for opendkim itself is almost non-existent.
>
> What we really need, and I think what the industry could really use, is a
> test tool that pretends to be an MTA but executes a script of milter
> actions sent to the filter with expected replies and then reports any
> deviations. It has to be able to do these in parallel so that
> multi-threaded loads can be simulated as well.
I'm thinking at least some of the tests should hookup a test version of the
MTA as the operation of the milter interfaces on the server side is key to the
acceptability of the product.
> I'll stick this on our proverbial task list someplace,
if you want to switch to the dark side of git(orious) there's a wiki too:
http://gitorious.org/opendkim/pages
> but I just wanted to
> mention that we need something like this in case anyone feels like taking
> on a project like that.
its got some real relevance to the async and policy protocol discussion. I'll
see how time goes.
> If there's interest, I'd be happy to flesh out
> the requirements.
defiantly interested. Not sure how to do great measurements but a list is a
good start.
> I might even start to tackle this myself since it may
> have applications to my day job and I might be able to talk them into
> donating the results.
good luck on behalf of all of us here.
Received on Tue Oct 20 2009 - 03:46:19 PST
This archive was generated by hypermail 2.2.0+W3C-0.50 : Tue Oct 20 2009 - 13:50:01 PST