From Gnash Project Wiki
This list is not a schedule, but a rough guide as to what I think we should be focusing on for the next time period, whatever that turns out to be. A big generic priority is we need more people to work on these projects, so if you have suggestions, send them my way.
I think Gnash running on fat embedded systems (Linux based, framebuffer backend, hard drive enabled) is very important. To me it appears there are more people wanting to use Gnash on an embedded device than as a browser plugin. I also think this is the future for Gnash, to be a ubiqutous client in an embedded world of streaming audio/video/flash content. As Gnash now supports a framebuffer backend, and will run without a GUI, I can run Gnash just fine on my Sharp Zaurus, Nokia 770, and Pepper Pad 3.
The current priority is getting the video streaming infrastructure fully working. By using ffmpeg and/or Gstreamer, we can utilize any codec for use by Gnash. We primarily need to handle streaming by RTMP over HTTP, (RTMPT) as well from a Flash server using RTMP.
I think to a less degree Flash compatability is important. While using Gnash as a Flash plugin is a popular way that people will be using Gnash, it's the embedded space that's a huge potential space for Gnash to dominate sooner than later as there are no good embedded Flash alternatives.
Figuring out a rough feature set and prototype is the main priority for the streaming server. Playing disk files turns out to be easy, and we can use libshout to experiment with transcoding a live stream from one codec to another, ie mpg to Theora. While doing so we'll want full logging of connection, file download, and performance statistics.
We want to continue the support of Ming, so we can use it for creating test cases, plus the free software world needs free Flash creation tools anyway. Our main priority for Ming is adding whatever support it needs for audio and video, and to extend this to support using Ogg, Flac or Theora.
We also want to continue to use Ming, so we'd like to see the development of a Flash ActionScript test suite that can be used to test Ming and Gnash. A longer term priority is adding whatever it is Ming needs to be ActionScript 3 and swf v9 compatible.