On Tue, 2011-11-15 at 07:40 +0100, David Henningsson wrote: > On 11/14/2011 09:37 PM, Tanu Kaskinen wrote: > > Ok. I'd still like to keep the clear phases of configuration parsing > > (structure creation) and probing (removing unavailable stuff). So what > > do you think? Your comment didn't really tell me whether you agree or > > not. > > Well, for me I haven't really seen it as if it were two specific phases > like that. So I haven't tried not to mix them. > I'd personally prefer whatever is fastest (we all strive to get a faster > boot!). However I haven't made any specific charts, so I can't tell for > sure that it is significant. > > What would the advantage of having "clear phases" be (except possibly a > fuzzy feeling)? Well, hopefully easier to understand code. But actually my mental model of the old version was wrong anyway - pa_alsa_path_sets were not created at configuration reading time before either, they were created by the sinks and sources. I'm fine with your approach. -- Tanu
