yea.. It could also potentially handle upload progress / resuming for non video files (maybe even multi-file select :) For video, the biggest advantage is the that it will support server side settings for transcoding and contributers can select the highest quality source they have available locally. This will avoids the variable setting down-sampling prior to re-encoding. (ie people first compress their DV or MPEG2 video to lower res with random mpeg4 or other codec settings (so that they can upload it). Whatever they upload gets re-encoded again to flash for web visitors. This type of extension will avoid that extra re-encode as the theora file they upload will be directly playable from the server this should greatly improve quality. 2) The other big benefit ofcourse is the autoupdate feature of firefox extensions. So once dirac "is ready" we can more easialy push out that encoder. (ie ffmpeg2dirac (which is quickly 'nearing readiness' and entering the debian repositories: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=505739 Then the server could request two qualities from the client one theora copy low bandwidth and "accessible" web streaming and another for archival at ~near original quality~. Its on my todo list to take a look at integrating firefogg to the mediaWiki upload interface :) peace, michael Gregory Maxwell wrote: