6 messages in com.googlegroups.google-talk-openRe: Voice-communication with other XM...| Subject: | Re: Voice-communication with other XMPP clients |
|---|---|
| From: | sbogus (huli...@googlemail.com) |
| Date: | 05/10/2007 03:39:46 AM |
| List: | com.googlegroups.google-talk-open |
Thank you for the very valuable information! It seems I must operate in one single field - I must use jingle- clients from one and the same implementation in order to acomplish my tasks...
Anyway, clarifying the current situation helps me greatly, thanks again! Kind regards,
sbogus.
stpeter написа:
On May 7, 2:57 pm, sinanimam <sina...@gmail.com> wrote:
Maybe this is not obvious from the above message, so to clarify I want to add these: - Jingle standard is currently in "experimental" phase. It is planned to pass to "draft" status very soon. I'm waiting for this too, with great impatience. You can get the news from jabber sites and blogs, for example:http://blog.xmpp.org/
Yes I write most of the entries there. ;-)
- Many clients started to implement Jingle features but nearly all of them are stalled and outdated now. For example in Psi it is very experimental and usually doesn't work. All client authors are waiting for the standard to pass to draft phase, because it would be a loss of time to implement jingle in its current form.
We are trying to get the STUN server discovery processes settled and we're also waiting for the IETF to finalize the ICE specs for NAT traversal. But those are moving along quite well.
- First of all the current jingle spec and libjingle are currently different. I think Google Talk's jingle is also different from all these. And Spark has another different jingle code. All or most of them are incompatible with each other, but that's not their fault. Jingle needs a bit more time.
Once Google updates Google Talk and libjingle to follow the specs, I think we'll see a lot more implementations. And I know that a lot of Jabber clients are adding support for Jingle right now. Plus it works today in the Nokia 800 and in the One Laptop Per Child machines. But we'll see a lot more soon.
- stpeter, is there any estimate on when it will happen? For example when could it happen earliest or latest? I'm expecting just a guess.
As I say we're held up by (1) STUN server discovery and (2) ICE. I am pushing on the STUN stuff right now. I need to ping IETF people about ICE. We should know more in a few days. Earliest would probably be about six weeks from now.
And I wonder how fast client authors will implement this feature after that.
Those coding efforts are going on right now.
Am I right that clients (especially Psi) will implement jingle as soon as it gets Draft status, or maybe they'll wait until Active status?
Draft is the next phase, after that is Final. "Active" is for informational specs. Since the specs seem to be quite stable, many clients are coding to the specs as they are today. I expect very few changes to the core specs -- the only thing that's really changing right now is the STUN server discovery process.
And yes I want support in Psi soon too, since that's the client I use every day. :)
--stpeter