On Dec 2, 2008, at 14:50, Kevin Smith wrote:
Mineing solves most things fine, apart from the one issue of when something should be un-mined. I had thought that what was being suggested was that when a user leaves one machine to go to another, he should hit a button on the machine labeled "release Mines", which could either change presence, or send <gone/> or any number of other notifications. Roughly equivalent to this is that when the user reaches the new machine, he can do some Remote Controlled "release Mines" on the other client. If this was the case (and I'm assuming it's not, and hoping someone will explain to me where I went wrong), then the user could just as easily hit the "lower my priority" button (either locally or remotely) as the "release Mines" button, which would in turn leave the new client with a higher priority and therefore get stuff routed to it without a Mine.
Ok, this is the "I leave chat windows open forever" scenario. I was concentrating on the (arguably more common) "I leave chat windows open for a little while" scenario[1]. Hopefully, the following words aren't too long (-:
For "your" scenario, I don't know if there's a way to avoid the user doing something to change state (locally or remotely), or clients being aggressive on their recipient address resets (e.g. fairly short timeouts, which I don't think are as wrong and/or evil as others).
For "my" scenario, mine-ing gets closer to the ideal than presence/ priority changes. This is even more true when client's use the <gone/ > chat state (which I'm seeing more and more frequently).
I hope this helps, at least explain my enthusiasm (-: -- Matthew A. Miller [EMAIL PROTECTED][1] I argue it is more common because that's the behavior I see with most of my colleagues, customers, family, and friends. If I didn't work for an organization that lives and breathes this, I wouldn't try to argue the point too vigorously (-:
smime.p7s
Description: S/MIME cryptographic signature
