Postfix and Mailscanner sitting in a tree k-iss-ing

[paddy]

On Thu, Dec 30, 2004 at 03:19:28PM +0000, Julian Field wrote:
> At this rate I might as well publish a book entitled something along the
> lines of "Jules' writings on all things MailScanner". I imagine it would
> be a cross between similar volumes by Chairman Mao and Tim O'Reilly.
> :-)

<big grin>

> Julian Field wrote:
>
> >MailScanner is built on the basis that what it does is take messages
> >from one queue, clean them up, and put them in another queue. That's the
> >fundamental design of MailScanner, and it's a very simple one. That's
> >the way it is.

Yep, got this.  Can't see a problem here: this is just email.

I would say that ...

MailScanner dequeues messages from an MTA
queues them for itself, (does arbitrary and unspeakable things to them ;)
and then eventually
queues some messages at an MTA

since MTA appears to be subclassed in the code.

The problem is at the interface to postfix ... ?

If I had to implement it for myself, I imagine I'd quite like to do the
initial dequeue and final queue by file ops: sounds simple and cheap to me.
But I can't yet see how the detail of how this is implemented get's its
tendrils deep enough into the code to get to ...

> >In Postfix's case, the "sanctioned interfaces" don't provide for any way
> >of implementing MailScanner's design, without changing the whole way
> >MailScanner works.

So, just for the sake of understanding, what would such a change look like?
just roughly.

> >Which I am not prepared to do. Quite reasonably, I think.

If that's really what its a case of, then you'll get no argument from me.
Indeed, 'quite reasonable' sums it up rather well :)

For that matter, not doing it, simply 'cos you're not _that_ interested
in it would be simply dandy by me. I don't use postfix and have no plans
to, so an actual implementation matters not one jot to me.

Understanding what the big deal is, I place value on, because, for me at
least, it can sometimes be a sign of something worth learning.

> >Wietse thought up a way for programs to interface with Postfix,
> >and stamps his feet saying that everyone must use the method *he*
> >thought of.

I've always thought (from the comfort of afar) of the postfix as more
evolutionary than revolutionary, perhaps I missed something.
I think I know what you mean though!

> >He didn't think of anyone doing it the way I did,

Ah! but did you file that patent on moving files in and out of queues yet!
;)

> >so he says I'm wrong.

> >It's an odd area of engineering where a novel design is wrong just
> >because someone else didn't think of it first.
> >
> >Yes, it is a technical problem. But not one I can/will solve the way
> >Wietse wants me to.

Okay, here's a guess, based on nothing more than idle speculation,
and a lifetime of cynicism:

The current mailscanner implementation is the best that anyone, who is
concerned with down-to-earth things like MailScanner performance, etc,
has come up with - it works and is supported by Julian (which is worth
at least his weight in gold, incidentally) - and users are happy.

Chances are it doesn't sit well with the postfix design, and so Wietse
is not willing to support it.  Naturally not supporting it means telling
people you don't support it, which entails explaining ... you can see
where this is going.

Add a dash of historical teething problems.

I still eager to be surprised by a description of why it would be
_technically_ difficult to implement a solution that could earn the
Wietse gold seal of approval.

My guess would be that noone is motivated to do so, and why should they be?

Tell me I'm wrong.

Regards,
Paddy
--
Perl 6 will give you the big knob. -- Larry Wall

------------------------ MailScanner list ------------------------
To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
'leave mailscanner' in the body of the email.
Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).

Support MailScanner development - buy the book off the website!