GStreamer backend on kdebase

Thomas Zander zander at
Sat Dec 15 11:07:02 CET 2007

On Friday 14 December 2007 20:48:04 Albert Astals Cid wrote:
> Vir says it does not work and it's not expected to work.

That sounds odd...

> So i ask myself why is it sitting in the place of the code we are going
> to release in less than one month.
> Can we please move it to somewhere were it should belong like playgound
> now that all the marketing buzz has been done and not bring it back
> until it sort of works?
> Albert
> P.S: Obviously i'm inmensely grateful to TT for giving us working
> Phonon backends on all platforms and for making it in a more open way
> and all that stuff, but that does not makes them able to not follow
> guides, and we are on a "release mode" and importing non-working code
> to the release branch is just plain wrong on my scale of things.

This code has been tested by various people (including myself) over the 
last weeks and it works great. On contrast, I never got anything sane out 
of the xine backend.
I'm not sure what the problems you are referring to, but it surely is not 
a black and white thing. Multimedia is hard.
Has there been a discussion with the trolls that maintain this code? 
One of them, Jens, was still waiting for his svn account when I talked to 
him last night, as soon as that process is completed I'm positive he will 
fix any issues you might find.

The code has been tested and has seen peer review from quite some people 
inside TT, so I don't think its fair to equate it to new code and the 
rules we have for that kind of code inside KDE.

On top of that; this is a piece of code that gives people choice. If it 
doesn't work it doesn't take anything away. The xine backend is known to 
have problems, what is wrong with giving people a second choice if the 
xine one doesn't work for them.

I think that is the bottom line here is; do we want phonon to work for a 
bigger chunk of people or not? Having more backends seems like a good 
thing to me...

Oh, One thing with your mail. While I value Virs opinion immensely; he 
stated on his blog that he never looked at this multi-thousand LOC 
codebase before last week.  So any statement that it can't work might be 
a bit quick, don't you think so?
Thomas Zander
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : 

More information about the release-team mailing list