Jon Nelson <[EMAIL PROTECTED]> writes:

> If you want a unique (enough) id, why not take the md5sum or sha1sum
> of the entire message, headers included?  Why not it's inode?
> (Guaranteed to be unique over the lifetime of the file on a given
> partition.)

The inode is meaningless, because the message we see during global
filter processing is given to the filter by naming a file; the message
we see during local filter processing is simply readable from stdin, and
we are not given enough information to link these two sources of the
message.  If we could do that, our job would be done and no unique
ID would be needed.

md5sum or sha1sum would not be unique enough unless they were
100-percent unique ... and they are not.

But even if they were totally unique, any changes to the message as it
moves through the filtering system would change the value of these
checksums, thereby rendering them useless.


> It seems to me like all of this work could be avoided if there was a
> *reasonable* way to
>
> a) inject new messages in place of old ones (this implies that we "stop"
> delivery of a current message and "start" (better yet, start over with)
> delivery of a new message.
>
> b) replace *in-place* an existing message. Courier would somehow have to
> notice this and re-parse the mime structure, etc... of the message and
> perform whatever other checks it already performs.
>
> [ ... ]
>
> I think that what people really want here is the ability to do the same
> between the esmtpd and the final "queueing" of the message.  Sort of
> like this:
>
> esmtpd -> temporary queuing -> filters -> final queueing (or failure)
>
> and not just for esmtpd but all forms of delivery.

Well, after esmtpd, we have that exact mechanism today.  This is what is
called "local filtering".

The part that's missing from your picture is that _during_ esmtpd,
something called "global filtering" takes place, and it can cause the
esmptd dialog to fail.  Global filtering is written to be fast and easy
on resources, which is why (I presume) it doesn't offer as many
capabilities as local filtering.  For example, messages cannot be
altered during global filtering.

The problem is that certain things which are ideally done (IMO) during
global filtering require minor changes to messages ... such as addition
of headers.  SPF is one example of this.

This is the sole reason that I came up with my Modest Proposals.
The idea is to make as small a change as possible to the global
filtering semantics, so that this remains efficient.  All I really
want is a way to link a message that comes through global filtering
to its image that later comes through local filtering.  If I had
a guaranteed unique ID per message, I could then build a database,
using that ID as a key, which could pass state information between
the global and local processing steps.

The inclusion of a unique ID in Courier's "Received" header would
be a very inexpensive way to provide this linkage, and it would 
require no change in the semantics of global and local filtering.

The maintaining of the database would not be part of Courier.  Only
those users who want to implement such a thing would need to do so.


-- 
 Lloyd Zusman
 [EMAIL PROTECTED]



-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
courier-users mailing list
[EMAIL PROTECTED]
Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users

Reply via email to