From: David S. <ds...@sc...> - 2003-07-29 18:36:00
|
On Tue, Jul 29, 2003 at 11:13:43AM +0200, Ronald Bultje wrote: > Interestingly, I've been working on something similar, but for filesink > (would work the same for filesrc). The idea is that filesink can write > to FIFOs and sockets, which aren't seekable. I'm doing this kind of > differently, I'm creating an event mask based on seekability. Code: ... > This looks more "gstreamerish" than properties, imo. I want to get rid > of as many "unneeded" properties as possible - the idea here is that a > pipeline solves all problems internally, and the application doesn't > need to do anything. Properties cannot do that. A good idea. "seekable" is definitely more a stream property that elements enable or modify. However, does testing seekability only work for the peer, or does it look at the entire pipeline? dave... |