Re: [LAU] A survey if you don't mind - results

On Sat, 25 Oct 2014, Will Godfrey wrote: I wonder how many choices are based on the SW in use rather than choice of audio driver. Some one using an application like audacity with a broken jack implementation or one of the midi apps that only does ALSA midi, have less choices. Someone developing embedded HW has to develop ALSA drivers first then get jack to work on that platform as well. There are lots of jack client walk throughs, but no guide at all on writing jack backends that might allow such people to go direct to jack. The info is probably right in front of my face but not in a way I see it... In any case I think any backend is either jack1 or jack2. a2jmidi (at least with jack2) does not clearly make the HW midi ports look like they belong in jack midi, giving the user the idea that alsa midi makes more sense. While it is true that HW midi ports are becoming rare, USB midi still shows up like hw ports. Anything using long sysex midi uses alsa midi. SO this choice is made by jack itself. (I know work is being done to allow jack to do non-rt midi transfers, but it will take some time for developers to make use of it) The midi jack/alsa thing is getting more complex with applications choosing one or the other and forcing the user to use both. a2jmidi is really a way of using both, but allowing the connections to be made in one place. These are meant to observations, I am personally glad to have all these things working as well as they do now. But it is obvious that people's choices of audio/midi backends is based on more than dis/likes. That would be interesting. -- Len Ovens www.ovenwerks.net