I am testing out Tweet.IM and I definitely like the concept.

The trouble I am having is…  I am logged in on my laptop with .(JavaScript must be enabled to view this email address) (via .(JavaScript must be enabled to view this email address)/laptop which is a Google Apps for Domains controlled domain).  I have my federation DNS records on rempe.us setup correctly AFAIK.

However, when I connected on my desktop as .(JavaScript must be enabled to view this email address)/desktop I get the following error message:

.(JavaScript must be enabled to view this email address)
You are already logged in with another resource

Here is a screenshot:

http://img.skitch.com/20081222-d4kd9mun6ikb2sws7q66afa2qr.jpg

I am using iChat v 4.0.7 (616) on OS X 10.5.6

Any ideas?  Is this a limitation of the tweet.im service, or is Google Apps for Domains interfering?  I have never seen this message before and I really want to be able to keep both my laptop and my desktop logged in at all times since I switch between the two frequently and I thought that is exactly what the XMPP resources can be used for.

Thanks.

Following up.  I get the same behavior in PSI on OS X.  Here is the relevant stanzas from the PSI XML console:

I also pasted the error here in case of formatting errors in the forum: http://gist.github.com/39154

Nudge Nudge :-)

Glenn Rempe - 29 December 2008 11:32 PM

Nudge Nudge :-)

Sorry, but most of the team is in holidays break.

Hi.

Perhaps now that the holidays are over someone could comment on whether this is:

- intended behavior
- a known bug thats being worked on
- a complete mystery?

The current situation seems to prevent use of this service from more than one location/computer (in my case, desktop and laptop) which kind of makes it a non-starter for me and I’m sure many others.

What I would hope for would be any tweets from people I follow find me on each of the computers I leave logged in.

Thanks.

Glenn Rempe - 17 January 2009 01:53 AM

Hi.

Perhaps now that the holidays are over someone could comment on whether this is:

- intended behavior
- a known bug thats being worked on
- a complete mystery?

This is the intended behaviour: We did not wanted to managed a list of all resources connected and status in our gateway.

However, an improvement to implement this feature is being worked on.

Actually I get 409 Conflict even when I am connceted just from Jabber account at the time to my own ejabberd (2.0.3) server. Currently it makes tweet.im for me unusable, which is a pitty—I hate additional protocols. According to /var/log/ejabberd/ejabberd.log I have no problems to get s2s connection to twitter.tweet.im but the protocol never gets active for me.

Hi,

having the same troubles - would like to use tweet.im (in Pidgin) on the workplace as well as @home. Often forget to logout @home so cannot tweet.im @work. What are the plans concerning this?

Would it be possible to ignore resources, and simply send messages to the JID generically, letting the user’s server do the routing? Since tweets are stateless, it seems sensible to send without a resource, since you’re not in a conversation with a particular resource.