Path: utzoo!yunexus!ists!shields
From: shields@ists.yorku.ca (Paul Shields)
Newsgroups: news.software.b
Subject: Re: /usr/spool/news on >1 fs?
Summary: hmmm... can you use sym-links?
Keywords: many small disks (<40M), hard links, trashed fs
Message-ID: <172@ists>
Date: 7 Jul 88 08:54:57 GMT
Article-I.D.: ists.172
References: <161@hawkmoon.MN.ORG>
Sender: news@ists
Lines: 35

In article <161@hawkmoon.MN.ORG>, det@hawkmoon.MN.ORG (Derek E. Terveer) writes:
> [...]
> In B2.11.14 News a "problem" that i have noticed that affects the running of my
> systems directly is that News seems to demand that /usr/spool/news, if mounted
> on a seperate file system, not have any seperate file systems mounted as sub-
> directories.
[example deleted]

>  Apparently [...] News forges links beween files 
> without checking that its on a seperate fs and somehow this is screwing up the
> overall fs tree (understandably).

The order is: open a temporary file in /usr/spool/news, link to the eventual
locations, write the file, then close and unlink the temp file.  Exactly
how this trashes the filesystem I don't know.

It would be easier if News did the links AFTER it wrote the files, so that
you could choose to copy-and-delete if the destination was on a different
filesystem.  Then, in the case of crossposting, you could use symbolic
links, if you have them.

(BTW, if you look at the source, you will find ifdef's for VMS, which 
doesn't have links, so it's kludged using fake symbolic links.  If you
have REAL symbolic links you're already 70% there.)

[...]
> Any ideas on how to configure News to get around this feature, or am i just
> stuck?  
> Or have i overlooked something real obvious??

Not that I can see.  It looks like a little code writing is necessary. 

Paul Shields, shields@ists.yorku.CA, shields@yunccn.UUCP
(...utzoo!yunexus!ists, ...mnetor!ontmoh!yunccn)!shields
It's amazing just how long it takes to get nothing done.