On Wednesday 28 September 2005 14:23, Steve Lamb wrote:
> Why should I when the program has to handle failure anyway? The
> argument always seems to be "The MTA can handle failures." OK, and
> when the MTA fails what, the program sending mail is just supposed to
> route it to /dev/null? Even mutt, the current MUA de jour of the
> die-hard unixeistas has it's own primative queuing in place in case of
> MTA failure.
Every MUA I've seen has an "outbox" where mail sits until it can be SMTP'd
away. Is that what you mean by primitive queueing?
