Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!uunet!mcsun!ukc!dcl-cs!gdt!ccsis
From: ccsis@gdt.bath.ac.uk (Icarus Sparry)
Newsgroups: news.software.b
Subject: Re: (C News) limit filesize and inews
Message-ID: <1991Jan5.181111.18941@gdt.bath.ac.uk>
Date: 5 Jan 91 18:11:11 GMT
References: <27839439.1746@ics.uci.edu> <1991Jan3.214226.9184@zoo.toronto.edu> <1991Jan4.115409.741@robobar.co.uk>
Organization: Bath University Computing Services, Bath, England
Lines: 43

In article <1991Jan4.115409.741@robobar.co.uk> ronald@robobar.co.uk
(Ronald S H Khoo) writes:
>One simple fix which I'd like for other reasons (but haven't actually
>bothered to implement ... sorry) would be to make inews call newsspool
>instead of relaynews wouldn't it ?  (configuration option, of course,
>but possibly the default one, since it's the more robust option)

We did that here quite some time ago. I happen to think that having
a file size limit which is reasonable, but can be simply changed is
a good thing. The problem is how to have it 'easy to change' without
everyone setting it to a very high number by default.

I added a line

( wc -c < $censart | sed 's/ */#! rnews /' ; cat $censart ) |\
	 /usr/lib/newsbin/input/newsspool

to the '$server' case near the end of inews.

You would need to do something similar in the case which ships the
news via 'rsh' as well. (I have a different problem, in that I have two
machines linked by NFS for news, but most people only have an account on
one machine. I therefore don't use the 'rsh' option, but do something
completely different).

>Of course there will be problems for sites that turn newsrunning off
>during the day.  Hmm...  It might be useful to have two separate
>in.coming queues anyway, for example, if I have organisation-wide
>news that needs to be run throughout the day, but usenet-wide stuff
>to be only run off-peak, then you need only put the locally posted
>stuff in the unaffected-by-newsrunning queue.

One very nice thing about Cnews is that it is driven by shell scripts.
If you want to only unspool small batches during the day, you could
make a copy of newsbin/input/newsrun, replace the 
	them=`ls | sed '/[^0-9]/d;50q'`
with
	them=`ls -l | awk '$5<10000 {print $9}' | sed '/[^0-9]/d;50q'`

(need $4 and $8 on BSD, or ls -lg) and run that from cron during the day.

This would catch almost all your local articles, as they would be
in a batch of 1 article.