Re: New package, review needed: icecast

[Andreas Thienemann <andreas bawue net>]

On Fri, 29 Jul 2005, Thomas Vander Stichele wrote:

> I didn't say that :) I know who Karl is.
I just wanted to make sure. ;)

> But there's a reason hisB branch isn't mainline - it's precisely because
> he is moving fast on features a lot that his branch is not the mainline.  
> It's a deliberate choice, and hence it doesn't make sense to package the
> branch and have people believe it's the official stable icecast.
Hmm. I thought that would have been made clear by the version number. But 
hey, I can just whip up a stable icecast release as well.

> > Thus the codebase is _very_ similar to icecast-trunk.
> It's not - it's a bit more unstable, and it's precisely so because he's
> moving fast on features.
The features I'v been interested in in the past have all been folded into 
trunk. That's what I was referring to with kh being similar to trunk.
And yes, his features are a bit more experimental. But the last year I 
have been running several icecast-kh daemons in productive use and not one 
of them did crash, something which I can unfortunately not say about the 
ices tools . ;)

> This allows you to install stuff with a slightly different name.
Ohh. Neat!

> If you use program transform name, they can be installed side by side
> with a little work, and that would probably be the nicest to do.
> If you can't be bothered, then a conflicts makes the most sense,
Okay, I'll see what I can manage there.

> but I wouldn't have it provide an actual icecast package.  People would
> be duped into thinking they get the actual icecast release, and might be
> turned off if it crashes for whatever reason.
Hmmm. WE should think about offering some virtual package such as 
webserver is being provided by httpd.
That way we could have clients depend on streaming-server and it doesn't 
matter if the user is installing icecast, icecast-kh or shoutcast or 
whatever....

bye,
 andreas