News for hackers. Stuff that matters.
Re:Paul Davis, lead developer of Ardour on fixing big Linux ... by PaulDavisThe1st on Sunday July 05, 2020 @02:17PM (#23740676)
About 10 years ago, Digidesign bought a nice protocol that was really tightly engineered to allow DAWs to build interfaces with control surfaces in an incredibly deep way. The DAW and the surface could negotiate based on their capabilities and needs. It looked really promising.

It has gone nowhere.

There are a number of reasons why, but I would wager that the most important is that people simply do not need this kind of complexity. It is the kind of thing that OEMs can use to build cool devices, rather than the kind of things the overwhelming majority of users ever wants to deal with.

Note that the MIDNAM specification also does a significant part of what you describe above, and although it's not true to say that it has gone nowhere, the industry essentially abandoned it rather than push it forward.

I apologize for sound cynical or skeptical about MIDI 2.0. MIDI 1.0 was a technical and social engineering miracle. But having been around the audio tech industry for 20+ years, I'm not optimistic that the MMA's processes are capable of replicating what was so good about MIDI 1.0 or avoiding the screw-ups they've been involved with since then.

Re:Paul Davis, lead developer of Ardour on fixing big Linux ... by WhatIsDukkha on Monday July 06, 2020 @01:11AM (#23744905)
Some things come too early and/or are underspecified.

Midi 2.0 seems like its about a sweet spot that will solve real problems.

I'm looking at the midnam files that come with Ardour...

It seems like these are focused on just named patches? I see the spec can do more like parameter names but it doesn't seem like that's been the focus? Is there a bigger pile of work somewhere?

Thanks, I might poke at this a bit and see what I can get working in Ardour.