[Standards] MUC Spam and MUC invites
dave at cridland.net
Wed Sep 5 16:05:46 UTC 2007
On Wed Sep 5 16:21:18 2007, Adam Nemeth wrote:
> Most of you may probably now that conference.jabber.org's rooms had
> spam-problems recently - https://stpeter.im/?p=2041 .
I had to read your comments there before this message made sufficient
sense to me. One thing I'd note is that, IMHO, the capability to
offer free-access conferencing services is a vital plus-point for
XMPP. Discarding that feature is a step backwards, and while I'm not
saying that jabber.org MUST reopen its chatroom creation, I do think
it's important that the technology allows it safely.
Open user registration is something else - you're effectively
allowing someone to create a new identity there. (I don't think
"CAPTCHA" is the way forward here either). I think that's the root
> You may also know that Google does not allow MUC invites because it
> does not come from somebody on your contact list, but from the room.
What you're saying here is that room invitations would benefit Google
users by coming from the actual sender, rather than relayed via the
room. This seems sensible, because blocking lists might otherwise
intervene as well for other users.
> A direct invitation would probably look like as follows (modified
> example 46 and 47 from XEP-0045):
This seems much more logical to me than the status-quo. You'd quite
possibly need to tell the room you're inviting someone, though.
> There's already a place for 'to' element in the xsd, it's optional
Both to and from are used, and to means something different to your
usage. I think there's a risk that there's a degree of overlap here,
in particular, a naïve client receiving your new message would
attempt to join the MUC room at "crone1 at shakespeare.lit/desktop", and
become horribly confused. Perhaps a different content might be better.
Dave Cridland - mailto:dave at cridland.net - xmpp:dwd at jabber.org
Infotrope Polymer - ACAP, IMAP, ESMTP, and Lemonade
More information about the Standards