multimedia infrastructure

Christian Müller cmueller at gmx.de
Tue Dec 31 12:40:25 GMT 2002


Am Dienstag, 31. Dezember 2002 04:02 schrieb Florin Andrei:
> On Mon, 2002-12-30 at 14:54, James Richard Tyrer wrote:
> > Florin Andrei wrote:
> > > Just imagine you have that installed on your system, and link different
> > > applications to it: all of a sudden, Konqueror can play DivX files
> > > "natively" (well, through xine-lib, of course); or any other
> > > application you choose.
> >
> > Yes, this appears to be the KDE way of doing other things.
> > So, we would have a library, a stand alone player, and a Kpart for
> > Konqueror. Sounds good to me.
> > Did you think that KDE should just use Xine (directly) or that we should
> > make our own library?
>
> I'm glad that at least someone understood the point of my initial
> message (having a multimedia library as a foundation for everything) and
> didn't took it as a "yet another My Player Is Better" lame type of
> thing. After all, i'm not advocating for a particular player, but for
> the idea of a multimedia foundation.
>
> Well, there are advantages and disadvantages to both ways.
> Use the original xine-lib, you don't have to worry much about
> maintaining it, but the API might change.
> Fork it, but then you have to sort of take care of it.
> I'm slightly in favour of using the original library. After all, you're
> not forced to upgrade the version you're providing, but instead KDE can
> declare some xine version as being the "official" KDE version. Or
> something.
>
> And hey, while we are at it, how about Gstreamer?
>
> http://www.gstreamer.net/
>
> Over xine-lib, it has the advantage of being specifically designed to be
> a multimedia foundation for a complex environment (after staring for a
> long time at the specifications and API, i had the impression that xine
> was designed to be just that: a multimedia library to build players on
> top of it, while Gstreamer seems somewhat closer to what i'm trying to
> suggest here).
> It has the disadvantage of being in an earlier phase of implementation,
> and less tested.
>
> Long term, i'd feel inclined in favour of Gstreamer: start implementing
> hooks to it now, so that after a while all the functionality will be
> there, and quite powerful and adequate for a desktop environment.
> Short term, xine makes more sense: one could make it basically right now
> a multimedia foundation for any desktop environment, but it wasn't
> exactly designed with desktop environments in mind.


I think something like this has been discussed before. 
A quick search in the list archives on lists.kde.org turned 
up the following threads: 

http://lists.kde.org/?l=kde-multimedia&m=101834192115997&w=2
http://lists.kde.org/?l=kde-multimedia&m=102960144902005&w=2
http://lists.kde.org/?l=kde-multimedia&m=100543760008814&w=2
http://lists.kde.org/?l=kde-multimedia&m=103538808617233&w=2


Best regards, 
Christian Müller.



More information about the kde-multimedia mailing list