Mp3 support patch
tim at ngus.net
Wed Oct 4 15:13:04 PDT 2006
On Wed, 04 Oct 2006 14:24:23 -0700 Zachary wrote:
> However, it must be made clear that this isn't a democracy. We must
> be discriminatory in order to keep the code manageable. Some of my
> other projects got "support" from developers who ended up leaving and
> not maintaining their additions (autotools on OES, for instance). I
> will not stand for that on ioquake3.
Just for the record.
My basic opposition is that we already have Ogg Vorbis support. This is free, already works and is not /potentially/ patent encumbered. The legal details aren't really important, at best MP3 support is always (well for the next few years) going to be a legal grey area. We can argue about the intricacies of this for days, but this only goes to prove my point.
We already have a compressed audio solution in Ogg Vorbis, what is the sense in adding another (inferior) one? Having MP3 support for EF obviously makes absolute sense as the data is unavoidably in MP3 format (and already licensed). In the case of ioq3 however, there is no prior media in this format. Any potential use of MP3 support is in the future only. In my opinion it would be better to encourage future use of Ogg Vorbis instead as it does not come with any baggage and is likely the superior codec for most use cases anyway.
As a general statement it's a bad thing to be adding features with no real world gain. It introduces more code which means more bugs and more maintenance effort. It doesn't matter who takes responsibility for this maintenance, it's just a fact. If adding a feature brings little tangible benefit yet increases work load, we don't need it. Feature creep is something to be avoided.
More information about the ioquake3