[Standards] Flash binding for Jingle

Matt Tucker matt at jivesoftware.com
Thu Feb 1 02:15:03 UTC 2007


Daniel,

> >From where I sit, Flash and Java took the same amount of time to get
> working on all three of the computers I use.  On Linux, 
> neither were installed out of the box.  On Mac, both were 
> installed out of the box.  On Windows, Flash was installed 
> but it was a matter of days before some web site forced me to 
> manually upgrade it.

Java can't be used to deliver VoIP through a web browser, unfortunately.
The main reason is that audio capture and RTP requires the Java Media
Framework, which can't really be delivered via an applet. We (Jive)
probably believe in Java more than just about anyone, but it's just not
suitable for VoIP in a web page. I don't think it was hyperbole from Ian
that Flash is the *only* choice here.
 
> I have nothing against this sort of gateway of course, I'm 
> just skeptical of the need for standardisation of how it 
> should be done.  I mean, we don't standardise how an MSN 
> transport should be implemented.  What would make a lot more 
> sense is a standard plugin or library which a developer could 
> just drop in, to make it "just work." 

We're not talking about a gateway so I'm not sure how the MSN analogy
applies. In fact, the XEP doesn't strictly have anything to do with
Flash. It simply defines a way to use the RTMP protocol through Jingle.

Flash has built-in support for RTMP, which is why we should think about
supporting it through Jingle. There's also now an Open Source RTMP
server. So, no, it's not a fully open protocol, but supporting it
through Jingle is practical.

> That makes even more 
> sense when you consider that this sort of thing is only a 
> stopgap measure, which may not even be worth keeping around 
> once Flash start providing some decent level of native socket 
> networking.

I'm not sure what you're talking about. Flash XMPP support is a totally
different issue. What we're talking about here is streaming media and
Jingle.

> If the XSF publishes specs for interacting with Flash, then 
> the next step would be publishing specs for interacting with 
> all the proprietary video protocols for the other IM systems. 
>  Should we really need to bother?

Again, I think you might be talking about something totally different.
:)

Regards,
Matt



More information about the Standards mailing list