Why are there no OSC Sequencers?

This was a question posed via Twitter by @artgillespie and retweeted by @CDMBlogs. I sat around and thought about it a bit, and came up with the following.

Quite simply, I think that OSC is in a development phase whereby it is being used to plug the weakness in MIDI - e.g. resolution of realtime parameter control being the primary failing of MIDI, and the main use for OSC, as opposed to creating a unified approach for control. There are sequencers for OSC, however they centered quite squarely on sequencing the control of specific parameters, or eccentric approaches to composition. It's a case of fixing the most broke stuff first.

MIDI has many issues, however the timing accuracy of simple note data, as controlled by a sequencer is not a major one of those, there have been many people who have criticized the quality of MIDI timing (BT comes to mind) but the reality is that the system can, on a good day, with the wind behind it, delivery note position accuracy of around 1/1500th of a beat @ 120BPM, when compared to the significant issues surrounding the control of parameters, and the 127 steps, this seems like much less of an issue.

The second reason there is less of a demand for OSC sequencing is that where one wants to create sets of data which require that level of accuracy, it is unlikely (but not impossible) that they are going to be recorded live, and quite often there is not only the timing consideration, but a timbral one, variations in timbre of that level of detail are very difficult, if not impossible to edit effectively, and therefore it makes more sense to use an audio recording system (which has a higher resolution anyway!)

There is of course the definition issue - MIDI is nice and easy and pre-defined, you press a note, and the synth knows what it is and plays along nicely - OSC has no such consistent definitions, you would need to write the definitions for yourself, and I personally don't fancy having to specify definitions for every single parameter I want to control, then have to start on each note, and the constituent parts of those notes - at both ends of the system, just as part of the set up of my system.

OSC works really well at the moment in it's current applications - I really enjoy Touch OSC and the other Hexler Apps on my Ipod Touch, but I don't think it will be hogging much light in the mainstream until a more user friendly set of definitions is established that manufacturers can conform to, and yet, at the same time I think that would be very sad, principally because the extensibility of OSC is one of it's real strengths.