From: Brad M. <bmi...@xm...> - 2006-03-29 00:29:47
|
Hi I see the docs* talking about registering a separate function for accepting events, but in going through an encoder for gsm, I see event handling in the chain function using GST_IS_EVENT and GST_IS_BUFFER. Is one or the other of these preferred? (*) http://gstreamer.freedesktop.org/data/doc/gstreamer/head/pwg/html/chapter-building-chainfn.html Brad |
From: Jan S. <th...@no...> - 2006-03-29 04:28:31
|
On Tue, 2006-03-28 at 17:29 -0700, Brad Midgley wrote: > Hi > > I see the docs* talking about registering a separate function for > accepting events, but in going through an encoder for gsm, I see event > handling in the chain function using GST_IS_EVENT and GST_IS_BUFFER. Is > one or the other of these preferred? > That means that gsm element has been properly to 0.10 - event handling and data passing used to be mixed in 0.8, but they're separate paths now, so the element has a bug. Cheers, Jan. -- Jan Schmidt th...@no... The woman you buy -- and she is the least expensive -- takes a great deal of money. The woman who gives herself takes all your time. - Balzac |
From: Michael S. <ms...@fl...> - 2006-03-29 09:11:35
|
On Tue, 2006-03-28 at 17:29 -0700, Brad Midgley wrote: > Hi > > I see the docs* talking about registering a separate function for > accepting events, but in going through an encoder for gsm, I see event > handling in the chain function using GST_IS_EVENT and GST_IS_BUFFER. Is > one or the other of these preferred? That's a sign of a poor port from 0.8 - where they were both handled in the chain function. The event case will now never be triggered in that chain function. The docs, in this case, are correct. Mike |
From: Thomas V. S. <th...@ap...> - 2006-03-29 09:20:40
|
On Tue, 2006-03-28 at 17:29 -0700, Brad Midgley wrote: > Hi > > I see the docs* talking about registering a separate function for > accepting events, but in going through an encoder for gsm, I see event > handling in the chain function using GST_IS_EVENT and GST_IS_BUFFER. Is > one or the other of these preferred? As a general guideline, you *should* *not* look at *any* element in -bad for examples and how to write elements. gsm is in -bad, and this is a good reason why :) If you want to compare docs against code, look for elements in core first, then base, then good, then ugly. Thomas |