gnash-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnash-dev] Getting the IOChannel working for RTMP


From: strk
Subject: Re: [Gnash-dev] Getting the IOChannel working for RTMP
Date: Mon, 29 Mar 2010 16:20:43 +0200

On Mon, Mar 29, 2010 at 01:53:25PM +0200, Benjamin Wolsey wrote:
> > The RTMPIOChannel doesn't necessarely need to be in NetConnection_as.cpp,
> > but that might be a good place to start experimenting with it.
> > You'll want to implement ::read, ::seek ::eof ::tell ...
> > well the more you do the better it is. ::size I guess will be hard,
> > but hopefully not needed.
> 
> Bear in mind that RTMP streaming controls are instructions to the server
> sent using AMF. This is different from the way http streaming is
> currently implemented. I've posted the following URL a number of times
> already:
> 
> http://markmail.org/download.xqy?id=d4dpnvqxs5nnmbzs&number=1
> 
> Note that play, pause, seek etc are all called using the same native
> function (2101, 202), which we know sends an AMF message when there is
> an rtmp connection. What else the function does is unclear, but it would
> be interesting if anyone has an idea how to find out. Does it behave
> differently for play, seek, pause, and stop, or does its behaviour
> depend on what the server does? In that case, how does http streaming
> work?
> 
> Possibly an IOChannel isn't the best abstraction for this case.

Seek requests to MediaParser still result in a ::seek request to IOChannel
so nothing prevents us from still using that interface.
But yes, the pareser aren't necessarely ready for that as there's an overuse
of ::seek for things that don't really require any server communication.

I won't insist though, if it isn't possible ... 
Was just an idea to avoid touching lots more places.

--strk;

  ()   Free GIS & Flash consultant/developer
  /\   http://strk.keybit.net/services.html




reply via email to

[Prev in Thread] Current Thread [Next in Thread]