Tim Legant <[EMAIL PROTECTED]> writes:

> When I wrote the previous message, I had given it no thought other
> than to wonder if a good, cryptographic fingerprint/hash of the
> message might be enough to identify it.  If so, then it's just a
> lookup in a file (DBM?) to find the name of a file (in the ~/.tmda/
> hierarchy) that contains a list of the delivery instructions for
> that message.  The file would be written the first time the message
> came through, if and only if one or more deliveries failed.

OK, that's all I was looking for, just at least one example
implementation that would theoretically work.  Thanks.

> I find this grotesque.

C'mon Tim, tell us what you really think <wink>.

> If I'm gone for a few days, I could have large numbers of messages
> delivered.  Assuming this could be a problem with a commonly used
> mailbox, every incoming message to that box would be delivered
> multiple times.  This "strategy" also plays hell with quotas,
> potentially rejecting important mail at the expense of multiple
> copies of something cute but useless that a friend forwarded.

It's clearly not ideal.  The rationale might have been something to
the effect of: Even though it's a half-assed implementation, the
condition that triggers it it not very likely.  That is, a block of
multiple deleveries where the first one suceeds, and a subsequent one
fails.  So, cater to the majority.

Of course, we've never had this mentality in TMDA, and have never
half-assed anything even though it would have been easier in many
cases.  :-)

I'm tempted to ask Sam what his thought process was when he chose to
do things this way in maildrop.  I'm pretty curious.

> I really think we need to explore something better.  

I'm game.  However, if we go with a more complex implementation, I'd
only like to turn it on when doing multiple deliveries.  Since most
users will only be using single delivery instructions, there isn't any
reason to subject them to this extra overhead.

> I'm curious about your thoughts on a good hash (MD5/SHA1) for
> message identification....

I think using a message digest (I'm partial to SHA1) would work fine
to identify the message.  We can discuss the actual implementation
details when you are ready, as I have some thoughts on that.
_________________________________________________
tmda-workers mailing list ([EMAIL PROTECTED])
http://tmda.net/lists/listinfo/tmda-workers

Reply via email to