Why couldnt a pseudostreaming plugin be created to work with any web server that supports resuming downloads/progressive downloads. What I mean is, cant the plugin keep the array of frame or seconds and bit locations so that when you skip past the buffered video, it just starts a new download from that location, still keeping the original file header in memory since the header would not be downloaded again? It seems this would allow any web server to be used.
If I understand correctly, the way it works now is the same except the streaming script slaps a copy of the header onto this "new" download, mp4 moov atom, or flv header, so that this new header information is collected again? Maybe im wrong, but I dont understand why the header should be sent again if the plugin can hold on to that and just request a new file without the server having to do anything special.
I know you may want to request the file for the first time maybe half way through, but the plugin could start a download at the front of the file to grab the header, then stop that download, start again at a new location, and it though it may take a second for that to happen, it should be the same in the end.
Am I overlooking something simple here?
If I understand correctly, the way it works now is the same except the streaming script slaps a copy of the header onto this "new" download, mp4 moov atom, or flv header, so that this new header information is collected again? Maybe im wrong, but I dont understand why the header should be sent again if the plugin can hold on to that and just request a new file without the server having to do anything special.
I know you may want to request the file for the first time maybe half way through, but the plugin could start a download at the front of the file to grab the header, then stop that download, start again at a new location, and it though it may take a second for that to happen, it should be the same in the end.
Am I overlooking something simple here?