pure transit server

[Fabien Tassin]

According to Bill Davidsen:
> 
> On Thu, 8 Feb 2001, Fabien Tassin wrote:
> 
> > 
> > According to Fabien Tassin:
> > > I'm not satisfied by the current model. Too many indirections and conf files,
> > > in particular.
> > 
> > speaking of conf files, IMHO, we have too many. I'm starting to think of
> > the new generation. I'm tired to edit 3 or 4 different files to
> > just add a peer, one should be enough.
> 
>   I think combining innfeed and newsfeeds would be nice, and certainly
> there are some things in innfeed.conf which should be allowed as part of a
> higer (group) spec rather than required at each peer.
> 
>   But in general it's quite clear what things are doing, if slightly
> ponderous to edit several files. And I am dead against anything which
> would reduce the generality of newsfeeds bu building in an assumption that
> outgoing feeds go to innfeed, because many of mine don't.

I'm not doing such an assumption. It was just an example of what I wanted
to do.
 
>   You might try using m4 (or cpp) or some such to generalize generation of
> newsfeeds. I usually split my feeds into three sizes, having found better
> performance that way. That's easy to generate with a macro.

Sure but I want to cure the problem, not to patch it. See below.

>   I have played with a parl program to process a spec into entries for
> incoming, newsfeeds and innfeed.conf, which could then be added to the
> files, probably just by adding to the end. Something like the cycbuff
> config to generate the cycbuff.conf file and a shell script to create the
> cycbuffs. Let's not lose flexibility in the name of convenience, the three
> files do things which are really not always applied to each peer.

Everybody has its own tips n tricks to do that. What I want to do is to
rethink the model to make it easier. My proposal was for the next generation
of conf files, whatever it will be for INN or INNT. I agree with Russ on
the fact that if it can fit into both models, we should follow this path.
What I don't want is to limit the new model to make it compliant with the
old one that we are trying to improve. 

-- 
Fabien Tassin -+- fta at sofaraway.org