Re: [Jack-Devel] The Situation(s) With JACK
Le 13 déc. 2011 à 16:37, Paul Davis a écrit :
> On Tue, Dec 13, 2011 at 8:22 AM, Stéphane Letz <[hidden]> wrote:
>
> Ok, lets be realistic. We have 3 different categories of problems:
>
> 1) things that make it harder or impossible for multiple
> implementations of JACK from being as fully compatible with each other
> as they could be
> 2) things that users (reasonably) want that one or other
> implementations of JACK currently does not do. This is somewhat
> affected by
> the split in developer effort, but its also somewhat
> independent of it.
> 3) the issues that occur when distributions and users are forced to
> choose between implementations
>
> What I'm hearing from you is that there is no way you can imagine
> fixing (3). I don't know if I agree or not, but lets just assume that
> its true.
>
> Could we agree to do what is needed to solve (1) and also provide
> proper guidance and information regarding (3) ? This would then allow
> the current fractured state of development, which looks unlikely to go
> away, to at least continue with a defined set of headers and a defined
> ABI/soname?
>
> If so, then the resolution of (2) is just going to come down to who
> does the work and for which implementation of JACK. Not a good
> situation but at least better than we have now.
>
So basically Paul you still don't want to choose...
Grame position is quite clear: we are committed to work on JACK2 codebase, which basically implement a superset of JACK1 features (http://trac.jackaudio.org/wiki/Q_differenc_jack1_jack2). If critical lacks in JACK2 (present in JACK1) are identified and are considered mandatory, please list them first. We consider that this basic choice has to be done first.
Stephane
1323798096.5838_0.ltw:2,a <4A07D073-23E3-4E42-B507-6D29BD997F6E at grame dot fr>