<div dir="ltr"><br><div class="gmail_extra">On 28 January 2013 15:53, Peter Nelson <span dir="ltr"><<a href="mailto:peter@fuzzle.org" target="_blank">peter@fuzzle.org</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div class="im"><br>
On Mon, 2013-01-28 at 14:27 +0000, Nathan Hulse wrote:<br>
<br>
> Yes, sorry about that. One could also say that it is not PA's<br>
> responsibility to map connections to JACK at all, and yet it does. It would<br>
> actually be cool if module-jackdbus-detect were fired again after loss of<br>
> the connections to JACK, in order to detect potential JACK backend changes<br>
> and remap again, although I know it only works on a one-shot basis at<br>
> present.<br>
<br>
</div>But pulseaudio isn't losing its connection to JACK, it is the ports is<br>
it connected to that disappear and come back. pulseaudio should not try<br>
to maintain knowledge of these connections as the user can change them<br>
at will to all sorts of other apps. Keeping track of that is delving<br>
into patchbay-territory.<br>
<div class=""><div class="h5"><br></div></div></blockquote><div><br></div><div>The purpose of module-jackdbus-autodetect is automated patching: We're already into patchbay territory, so you're arguing after-the-fact here. In actuality, one could say you're arguing against delving further, but this seems pointless to me. module-jackdubs-autodetect appears to exist for the puropose of desktop integration; if it is to detect ports and automatically map connections at startup, why should it not also keep track?<br>
<br></div><div>I'm not saying that the module is deficient because it doesn't track changes, only that it'd be nice if it did, should David wish to work on it further. Its certainly a corner-case, afterall. In my position, yes, the problem is not PulseAudio at all: That JACK still, after all this time, cannot handle suspend/resume is obviously not an issue for PA to solve. Furthermore, I only use a stereo pair and the FFADO driver forces to presentation of of every channel available in hardware, to which module-jackdbus-autodetect then automagically maps. I'm still learning here, but I've come to relaise that I really should be scripting the patching entirely and not using module-jackdbus-autodetect!</div>
</div><br></div></div>