[gst-devel] addition of a buffer flag

[Ronald Bultje]

Hi,

On Wed, 19 May 2004, Thomas Vander Stichele wrote:
> > Both are a way to solve the problem. None is invalid w.r.t the other.
> > It's about preference. So let's continue with that as a starting point.
> >
> > Ogg fails to nicely prepare data for AVI. AVI fails to nicely prepare data
> > for Ogg.
>
> Not sure I follow.  Why would you want AVI to prepare data for Ogg ?

We're not OggStreamer, we're GStreamer. avi -> ogg transmuxing is supposed
to work, so our vorbis data protocol must allow for this. The other way
around should also work.

This is all unrelated since the problem you're facing appears to be
outside the mux/codec area, which is what my reply assumed.

> Anyway, I'll take up the exact problem up from the top in my next mail,
> since on IRC you seemed to indicate that you at first didn't understand
> my specific problem.

Your problem seems to be between oggmux and the pipeline sink element, not
between muxer/encoder. I don't have a better fix for it then what you
proposed, although I don't like your proposed fix. That's why I said: go
ahead. I also explained why I didn't particularly like the proposed
solution.

I'll restate what I meant, just so it's clear once and for all:

* for header data between encoder/muxer or demuxer/decoder, we use
GstCaps.

* for header data between muxer-output and outputs (filesink,
fdsink, ...), we have no good solution as for now. Therefore, go ahead
with your solution. However, just be aware: some of us are not convinced
that it is "the right"(tm) solution.

The discussion seems to be about this last tiny thing, so feel free to
continue the discussion on that bit in any follow-up email. The other
parts are done, as far as I'm concerned, so let's stop that discussion and
go back to work. ;).

Ronald