Re: [Gnash-dev] Fw: rtmpget - usable?

[Rob Savoye]

Alexander Huemer wrote:

> unfortunately i did not get a reply to my message from 2009/03/19.

  Hey, I've been at a conference, which put me behind more than usual on
email...

> i noticed, that there were no changes in the rtmp branch since then.
> could please someone take the time to describe the current state of the 
> code?

  If you'll noticed I made a big checkin a week ago, and more yesterday.

> i chose gnash-dev, because the rtmp branch is not part of the mainline
> distribution.

  Correct, it's my experimental branch for reverse engineering RTMP. At
intervals I migrate it to trunk, which I'll be doing again in about a
week or two. Lately I've been working in an even more experimental
branch, which I migrated to 'rtmp' yesterday.

> is this somehow my failure?
> or is rtmpget just in a unusable state?

  No, it's not, but it's only a little broken cause the client side API
has been changing as I worked on Cygnal, and I haven't updated rtmpget
yet. Instead I recently added RTMPT support to Gnash itself, with real
RTMP support following in about a week... As Gnash optionally stores all
the files it loads to disk, it'll accomplish the same thing as rtmpget.
I plan to fix rtmpget as soon as I finish debugging this current RTMP
hacking. (which involves handling chunked data for *really* big messages)

  As the client side API was changing, I realized rtmpget was a bit too
different from how Gnash needed it to work, so it made sense to tweak
the API for Gnash while I was working on it, and then let rtmpget catch
up after it's stabilized. Up till now I'd primarily worked on the server
side of RTMPT/RTMP/HTTP for Cygnal, so the client side needed some love...

  The plan is to have support for RTMP based video conferencing and
streaming in the next release supported by both Gnash and Cygnal. Course
it'd help if somebody could implement the Camera and Microphone classes
so I don't have too... :-)

        - rob -