[gst-devel] Re: [gst-cvs] omegahacker gstreamer: gstreamer/gst/

[Erik Walthinsen]

On Fri, 31 Jan 2003, David I. Lehn wrote:

> The g_print is the whole point of this default handler.  Why comment it
> out?  Since this is an exported API function it should probably actually
> function.  I assume apps will adopt similar options that gst-launch has
> (--silent and --exclude) if they use this handler so there seems to be
> no reason to nullify it.  If you think it's a bad default for gst-launch
> then maybe gst-launch should be changed to have a --verbose option
> instead of --silent.

It is debugging information, so no matter what tool it is it should be
only turned on when requested.  I will revert that patch and instead
change the default so that it is not used in -launch.  Or perhaps the
definition of this information is fuzzy.  I don't consider any of the
information I'm seeing from this printout while doing a gst-launch much
use unless doing significant debugging of either the scheduler or the caps
interactions of the elements, and even then it isn't entirely clear how
useful it would be.

JZW did have a valid point about programs spewing all kinds of both errors
and "status" information that's totally irrelevant.  In that sense we
should consider whether even the initializer INFO lines should ever be
displayed.  I thought once upon a time I had tweaked the debug system to
show nothing unless it was requested, apparently that didn't stick.
Things like '0 live buffer(s)' are certainly not things users need or want
to see.

      Erik Walthinsen <omega at temple-baptist.com> - System Administrator
        __
       /  \                GStreamer - The only way to stream!
      |    | M E G A        ***** http://gstreamer.net/ *****
      _\  /_