henry@utzoo.uucp (Henry Spencer) (08/06/89)
Geoff has figured out why C News's relaynews sometimes returns status 16 without producing any error message; it's a problem in stdio buffering. Fix is coming. At least some of the instances in which it is happening seem to be provoked by badly-malformed control messages (some people do not understand that posting an article to whatever.ctl makes it a control message and it must meet strict rules to be legal). -- 1961-1969: 8 years of Apollo. | Henry Spencer at U of Toronto Zoology 1969-1989: 20 years of nothing.| uunet!attcan!utzoo!henry henry@zoo.toronto.edu
henry@utzoo.uucp (Henry Spencer) (08/06/89)
A little PS from Geoff about this: note that error 16 is DROPPED, not DISKFULL as some have asserted. (It is admitted that the error codes are not as well documented as they ought to be...) -- 1961-1969: 8 years of Apollo. | Henry Spencer at U of Toronto Zoology 1969-1989: 20 years of nothing.| uunet!attcan!utzoo!henry henry@zoo.toronto.edu
richard@uhccux.uhcc.hawaii.edu (Richard Foulk) (08/06/89)
} A little PS from Geoff about this: note that error 16 is DROPPED, not } DISKFULL as some have asserted. (It is admitted that the error codes } are not as well documented as they ought to be...) Why use "codes" anyway?
geoff@utstat.uucp (Geoff Collyer) (08/06/89)
> Why are they "codes" at all?
Relaynews's exit status is a bit vector, each bit representing a
distinct kind of failure (RTFM: relaynews(8) DIAGNOSTICS). Relaynews
could instead just exit with status 1 in the event of an error, but
supplying the bit vector is sometimes useful to people who work on
relaynews. None of this really matters to most people, now that the
stdio flushing business has been fixed; you don't have to decipher the
exit status if you don't want to.
--
Geoff Collyer utzoo!utstat!geoff, geoff@utstat.toronto.edu