[ircd-ratbox] [patch] +S / sslonly support for ratbox-services (partial/draft)
lee at leeh.co.uk
Tue Sep 9 20:17:34 UTC 2008
On Mon, Sep 08, 2008 at 12:11:48AM +0200, Peter Schuller wrote:
> Does some trivial changes to make services understand the +S
> mode. This is based on ~1 hour of looking at the code so I would not
> be surprised if I missed things. It "seems to work" enough for
> chanserv to be able to enforce the +S mode.
> I does not address the issue of how to enable/disable this. If the
> ircd is built without +S support, I suppose one does not want the
> service to be trying to set such a mode, which this would allow.
> Earlier today there was some discussion on #ratbox on making the
> SSL-only stuff a config variable, so wanted to hold for
> feedback/opinions before I try to make it one (I am very weak on
> autoconf/configure so it would take some time for me).
Been busy with work, so haven't got round to looking at either of these
Generally speaking though, I tend to be a lot more lenient with what
features go into services, than we are with what goes into ircd. The ircd
still has an efnet direction -- whereas services isn't designed for efnet..
I just had a quick look at the ircd mode handler -- it is a lot more
permissive about unknown modes than services is. Services will error and
drop the link, and there's a very specific reason behind it.
When services gets a mode character you don't recognise, it doesn't know
whether that mode has an argument or not (ie, +snt vs +b x!y at z). Whilst
ircd presumes it doesn't take an argument, services avoids making a decision
by simply dropping the link.
I would wait a couple of days, we'll come to a decision on whether it should
only be sent to services with a suitable CAPAB -- this will impact how
services needs to deal with it..
- Lee H // anfl
- I code, therefore I break things.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 197 bytes
Desc: Digital signature
More information about the ircd-ratbox