🚀 go-pugleaf

In article 25032@blilly.uucp, bruce@blilly.uucp (Bruce Lilly) writes:
>In article <1993Jan21.191902.5445@til.til.com>,
> posted to comp.mail.uucp,
> erik@til.com wrote:
>>
>>I, for one, wouldn't have *any* problem with having to send a new map entry
>>in *monthly*.  After all, it's just a trivial cron job, and it would solve
>>the obsolete entry problem.
>
>Would you like to be on the receiving end of 10,000+ such
>monthly cron jobs?

Not if I had to weed through them by hand.  To what extent can this entire
procedure be automated?  To what extent are the current map entries checked
by the rutgers team?  I suspect that they aren't checked at all - for
evidence of this, look at the latitude and longitude values.  If this is the
case, it's pretty easy to automate the job.  It's still a lot of email to
handle, but only a computer needs to look at it.

>If it's a cron job, and the person administering the machine
>leaves, but the obsolete entry continues to be submitted by
>cron (the assumption is that the replacement admninistrator is
>either ignorant about the cron job or the maps, or incompetent),
>how is the problem solved? 

Automatic submission by cron handles the problem of making the map entry for
machine <x> die when machine <x> leaves the uucp network - because either the
machine is no longer executing *at all*, or because it's disconnected and the
automatically emailed map entry no longer makes it off the machine.  It does
not *automatically* handle the job of updating the map entries list of sites
that machine <x> talks to.  I think that this could be done, at least partially,
by an sed script looking at the /etc/uucp/poll file (for a Sun, at least).