[LAD] OPEN CONTROL ARCHITECTURE

[Len Ovens]

On Sat, 28 Feb 2015, Frederick Gleason wrote:

> Right.  It’s the ‘mechanism, but no policy’ conundrum all over again.

This is a big one all right.

> I have no particular beef against OSC or OCA.  However, there is already 
> a small multitude of AoIP control protocols out there (Dante, JetNet, 
> LiveWire, Q-Lan, WheatNet, …).  I don’t think that adding Yet Another 
> Incompatible Protocol is fundamentally going to improve this situation. 
> What would be far more helpful would be a decent FOSS library for 
> supporting one of the quasi-dominant AoIP systems; something that could 
> help push that system over the top to make it the defacto industry 
> standard.

I completely agree.... except, they don't seem very open. A lot of them 
are not very good either. There seems to be an inverse trend of openness 
to goodness. The more open things are the worse they are too.

> Today, in the pro audio/broadcasting space, there are really 
> only two realistic contenders for this role: LiveWire or Ravenna.  Pick 
> one.

These two are basically no standard. Zeroconf for discovery and then web 
browser for everything else. This is ok for something that will be set up 
once during install and then will be used with that one set up there 
after. (I have been reading device manuals) For a lot of Broadcasting set 
ups this is probably fine. For live sound it is not... probably Dante is 
bigger there though, but completely closed as it is the only thing that 
feeds money to Audinate (who are now a partner with OCA BTW). Every audio 
console maker that has split controler/mixer or digital snake seems to 
have their own control system as well, mostly based on MIDI as far as I 
can tell (they don't tell, just let some things slip :)

For the uses I see in Linux, the browser approach is always available, it 
is the device setup standard pretty much across everything. Every embedded 
OS supports some sort of web server. But this stuff is slow for live 
control. MIDI is the thing people use right now for RT control, but it is 
not at all standard or discoverable.

I see your point about defacto standard. The MCP is a good example, not 
the best protocol, quite limited with only 16 faders. The Yamaha MIDI 
mixer control is much better (Allen & Heath allows more controls but at 
lower rez too and I am sure there are others).

If I make some OCA kinds of things, it will be because I see a use for it. 
It will help me in my setup or others with a similar small setup. It has 
been a long time since I was in the broadcast world and if I get back in 
it will be in a small setup as I am not willing to move back to "the big 
city". In that case I will be working with whatever they have anyway.

A quick note on Audinate. They are a partner with OCA as they were one of 
the players in AES67. They are the only player from AES67 (that I can see) 
that have not yet released anything that allows Dante devices/sw to work 
with AES67 which they promised for the end of 2014 (the promises are still 
on the website :) Their involvement in OCA does not in my mind mean they 
will support OCA in any way and may be only making sure none of their 
patents are disturbed. Audinate has the most to loose with any new 
standard as their business is their protocol. Other network standards seem 
to be linked with hardware. So the company has less to loose by supporting 
another protocol so long as they include it with their hardware. It may in 
fact allow them to streamline their product line by dropping some bits and 
concentrate on the most profitable bits. If OCA is released in the next 
few months as a standard (as it is supposed to be), another year will tell 
if it is going to be used. Ravenna and Livewire (not to be confused with 
LiveWire(tm) which is a motorcycle, or Livewire 
http://static.comicvine.com/uploads/original/1/19859/846008-livewire__animated__picture_100.jpg 
) both could bennefit from OCA as far as I can see though as you say they 
may not feel the need to.

--
Len Ovens
www.ovenwerks.net