Pared down build via cerbero?

Michael McNamara mmcnamara at
Wed Jun 11 15:40:11 PDT 2014

Hi Chuck,

Thanks for the reply.  I've successfully cut down the number of plugins in use inside our apps via #define statements in iOS and Android while maintaining the functionality that's needed.  (I agree that it is tricky to get the right combination sometimes!)  Presumably those plugins that aren't #defined aren't making it into the final app binary.  The overall binary size in fact grows or shrinks based on those #defines,  so I feel pretty good about what's going on there.

My question is more about keeping the unneeded plugins from ever being built via cerbero as they're not needed - or if it's even reasonably possible to do something along those lines.


From: Chuck Crisler <ccrisler at<mailto:ccrisler at>>
Reply-To: Discussion of the development of and with GStreamer <gstreamer-devel at<mailto:gstreamer-devel at>>
Date: Wednesday, June 11, 2014 at 1:43 PM
To: Discussion of the development of and with GStreamer <gstreamer-devel at<mailto:gstreamer-devel at>>
Subject: Re: Pared down build via cerbero?

I don't know anything about using cerbero, but on linux you can remove features with configure. You just have to be careful that you don't remove something that you really do need. Sometimes it isn't immediately obvious.

On Wed, Jun 11, 2014 at 2:47 PM, Michael McNamara <mmcnamara at<mailto:mmcnamara at>> wrote:

Here's a build question...  If you know that you're only in need of a small subset of plugins, is there a relatively simple way to only include those plugins that are needed in the build process?  What would the best approach to take in such a scenario?  How fragile will things get if I start removing libs from the get-plugins-good recipe for example?  Or is that the wrong direction to head?

Cheers and thank you!

gstreamer-devel mailing list
gstreamer-devel at<mailto:gstreamer-devel at>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the gstreamer-devel mailing list