Bug#765766: Dirac package unloved and appears to be nearly obsolete, maybe consider removal after jessie.

[peter green]

Package: dirac

The dirac package seems pretty unloved. It's only ever had one upstream 
version in debian and all of the recent uploads have been "team 
uploads", mostly tweaking the packaging with a handful of FTBFS fixes.

The package fails to build on armhf with gcc-4.9 with a testsuite 
segfault and has been that way for months (see bug 756777). It doesn't 
appear there is anyone with the time/inclination to properly debug this 
and work out whether it's a true bug in the compiler or some reliance on 
undefined behaviour in the dirac code.

Furthermore the package seems to have mostly been replaced by 
Schrödinger . The decoder seems to have no remaining reverse 
dependencies and the encoder seems to have only one reverse dependency 
remaining. Unfortunately that one remaining dependency is a fairly 
imporant package, namely gstreamer0.10-plugins-bad .

Upstream also seems to be trying to push people towards Schrödinger. 
Thier website says

"The Dirac project maintains two encoder implementations: 
dirac-research, a research encoder, and Schrödinger, which is meant for 
user applications. Schrödinger outperforms dirac-research in most 
encoding situations, both in terms of encoding speed and visual quality."

And there do not appear to have been any new releases of the plain dirac 
package since 2009.

As a band aid for jessie. I just uploaded a NMU to use 4.8 on armhf but 
that is not a long term soloution.