[Standards-JIG] Re: proto-JEP: User Nickname

Julian Missig julian at jabber.org
Mon Jun 7 06:24:03 UTC 2004

On 6 Jun 2004, at 8:20, Richard Dobson wrote:

>>>> On Fri, Jun 04, 2004 at 02:26:25PM -0600, Joe Hildebrand wrote:
>>>> Agree.  We have <NICKNAME> in vcard-temp now, and there should be
>>>> something
>>>> in the pub/sub vcard-temp replacement that does the same thing.
>>> I think the idea is not the nickname as we know it in vcard, a more 
>>> or
>>> less
>>> static value, but more like how many (young?) people use nicknames on
>>> MSN. That
>>> might be more appropriate outside vcard like things, and more
>>> comparable to
>>> user mood, user activity, etc.
>> No, I understand that, but I still think that it's a lot of network
>> activity just for that. Especially since many clients (mine) probably
>> won't actually be displaying these dynamic nicknames in lots of places
>> anyway..
> Thats why it uses pubsub rather than being a presence extension, so 
> that
> only people who want to receive the information will get it (i.e. the 
> people
> who subscribe), and it is not forced upon anyone like yourself if you 
> dont
> want it. If you think its a lot of network activity then I suggest you
> contact the pubsub JEP authors as it just uses pubsub as I understand 
> it and
> the level of traffic is a result of that JEP not this one, and if you 
> think
> it is such a serious problem for this nickname JEP then it will likely 
> be
> just as much of a problem for any future JEP's that use pubsub.

No, that's different. The difference is that this is one incredibly 
tiny piece of data. It's a single short string. It just feels to me to 
be a large waste of resources for that to be the *only* thing 
transmitted with all this pub/sub overhead. There are probably quite a 
few bits of data we would want to transmit just as much as a nickname 
that we could put together with it, I just don't know what that correct 
combination is right now.

My problem is not with pub/sub, it's with the fact that we're doing a 
*lot* for just a single string, when there are definitely other pieces 
of data which we will need..


More information about the Standards mailing list