On Wednesday 28 September 2005 14:18, Steve Lamb wrote:
> > I think a better question is: why should cron, at, mailx and all the
> > MUAs independently be coded and debugged to implement the
> > functionality that any one MTA can provide?
I forgot logcheck.
> Because the MTA could fail. Again, just because they would be able
> to handle a failure of the MTA doesn't translate into programming in an
> MTA. It's quite simple, like, uhm, not deleting the message and trying
> again later to the same interface. All of a dozen lines of Python the
> last time I did it.
Sure: the program could hold the message and try again later -- to
call /usr/lib/sendmail. Either the program has MTA functionality, or its
wrapper has that functionality, or the machine needs to provide an MTA.
Unless you just want to syslog all its output. But wait, that means you
have to rely on syslogd, which is almost as bad as requiring an MTA! We
could write all our programs to cat their output to their own logfiles, I
suppose.
