Xref: utzoo news.admin:10623 news.software.b:5740
Newsgroups: news.admin,news.software.b
Path: utzoo!henry
From: henry@zoo.toronto.edu (Henry Spencer)
Subject: Re: Bogus control -- need I do anything?
Message-ID: <1990Sep14.152216.28161@zoo.toronto.edu>
Organization: U of Toronto Zoology
References: <1990Sep12.180429.4305@barsoom.nhh.no> <1990Sep13.033355.1567@zoo.toronto.edu> <1990Sep14.014042.17586@wsrcc>
Date: Fri, 14 Sep 90 15:22:16 GMT

In article <1990Sep14.014042.17586@wsrcc> wolfgang@wsrcc (Wolfgang S. Rupprecht) writes:
>>... The only negative effect is the batch file cluttering up "bad",
>>and even that will go away in a few days...
>
>Does the batch really need to be junked?  I just fed it to newsrun
>again.  Was this superfluous or would the articles after the bad
>control msg in the batch have been lost?

It was superfluous; for this sort of error, everything gets processed,
including the offending article.  Some other classes of errors, like
decompression problems, can spoil a whole batch, so keeping the whole
thing around for investigation is generally a wise idea.  (It is also
complicated and difficult to keep only *part* of it around, and simple
and easy to save the whole thing.)

It is possible that some classes of problem, like bad control messages,
will be "decriminalized" :-) so that they produce a message in errlog
but don't get flagged as serious trouble justifying saving the batch.
-- 
TCP/IP: handling tomorrow's loads today| Henry Spencer at U of Toronto Zoology
OSI: handling yesterday's loads someday|  henry@zoo.toronto.edu   utzoo!henry