Aidas Kasparas <[email protected]> writes:

> And, I'm not against fair order for message sending. I just want to
> be fair per user, not per message.

Right, something would have to track how much each user has submitted
to the queue in the last [some amount of time], in number of bytes
multipled by number of recipients.  

I think to determine priority I'd take the square root of that figure
(multiplied by a configurable scaling factor, which could probably
just default to 1) and subtract it from the number of milliseconds the
thing's been in the queue.  The queued message with the greatest
difference goes next.  Thus, if you've sent a megabyte's worth of data
your priority starts around -1024, gets up to zero after 10.24 seconds
and climbs linearly from there (until it gets sent), but if you've
sent a gigabyte's worth it takes your priority almost five and a half
hours to get up to zero, and meanwhile all the short messages would be
going ahead of you unless there's spare bandwidth (which, in practice,
there frequently would be, but each time a new message came in it
would beat you to the next sending slot, if you've got multiple big
messages going out, or need to have your big message sent out multiple
times to different domains).  Just to be clear, messages with a
negative priority *would* be sent, but only *after* any messages with
higher priority.

But yeah, that sounds like fair amount of work to implement.

And it would be controversial (although, if the scaling factor
defaults to zero all messages would be sent in the order they are
received by default, unless they need to be requeued).

-- 
Nathan Eady
Galion Public Library

------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today
http://p.sf.net/sfu/msIE9-sfdev2dev
_______________________________________________
courier-users mailing list
[email protected]
Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users

Reply via email to