[Standards-JIG] MUC Invitations, Jingle Relays, and Big Problems

Kevin Smith kevin at kismith.co.uk
Mon Nov 6 15:24:08 UTC 2006


Sorry, I'm replying to this mail despite the conversation having 
continued since, but there's a point here that I'd like to address.

Robert B Quattlebaum, Jr. wrote:
>> As you note, many use cases involve unknown contacts, and I can only 
>> see the use cases increasing rather than decreasing. I can only 
>> admire server admins which take a proactive approach to preventing 
>> unwanted messaging, but this seems to be a sub-optimal way to do it. 
>> Imagine if GMail only allowed you to receive mails from people in 
>> your address book, you'd never get notifications for  many things 
>> such as online purchases.
>
> Email is a bit different than instant messaging as you know, so this 
> comparison doesn't quite hold up. 
I don't think this is true. Certainly there are use cases for IM outside 
those for email, but the reverse isn't really true (I'm not about to 
suggest we try and replace email with xmpp, but there's no reason to not 
provide an alternative) For people I have a jid for, I almost 
universally contact them with xmpp rather than email.
> Sending unwanted and unsolicited instant messages to someone 
> immediately wastes their time and attention.
Well, I can't say I sit at my machine thinking "I'd love it if a Psi 
user contacted me right now to discuss X/Y/Z" but that doesn't mean I 
universally want them to not do so. Are these unwanted instant messages? 
I publish my Jid in various places -- if people contact me, is it 
unsolicited? Certainly I have no great desire to set up a presence 
subscription for a single conversation, but conversations without 
presence are not universally spam, or the suchlike.
> Good example. Perhaps we should start building a list of XEP's that 
> break with paranoid filtering.
Well, various others have said the same thing since this post was made, 
but I think it's better to just avoid this filtering behaviour.

/K

-- 
Kevin Smith
Psi project lead.



More information about the Standards mailing list