GNOME Bugzilla – Bug 158406
[strings] GStreamer told me to file a bug
Last modified: 2005-08-26 00:02:59 UTC
lvs@lvs:~> gst-launch-0.8 filesrc location=../Temp/017.part ! avidemux ! queue ! ffdec_msmpeg4 ! xvimagesink avidemux0. ! queue ! mad ! osssink RUNNING pipeline ... (process:4523): GStreamer-WARNING **: queue queue0 does not connect different schedulers ERROR: from element /pipeline0/queue0: Internal GStreamer error: state change failed. File a bug. (process:4523): GStreamer-WARNING **: queue queue1 does not connect different schedulers ERROR: from element /pipeline0/queue1: Internal GStreamer error: state change failed. File a bug. Execution ended after 1 iterations (sum 2087438000 ns, average 2087438000 ns, min 2087438000 ns, max 2087438000 ns). lvs@lvs:~> I think the message should not advise filing bugs.
It looks like 2 more or less independent error messages are interacting in a way that sends out wrong message... If state change fails, it usually is error condition signalling a bug, except for things like queue. Maybe amend queue's message to read: "queue queue1 does not connect different schedulers. This is likely a bug in application"? It would change the meaning of following "file a bug" enough to give right idea I think.
It's quite definately a programming error, so you don't even need a GST_ELEMENT_ERROR. I can't find the reference to this error anywhere in the sources, though, isn't this local to your version? The only STATE_CHANGE error is in fakesink.
See also bug #305291. Asking the user to file bugs is standard practice for grave core errors, so I don't mind it too much. We are a bit impolite though. Will close this bug shortly if there are no objections.
The ERROR is a bit bogus, it's not a GStreamer error (rather an application error). But since it doesn't occur any more in HEAD I'm closing as FIXED.