[gmpi] Re: Topic 1: Audience for and users of plugins

At 01:40 PM 2/20/03 +0100, you wrote: 
All is negotiable per definition since we are talking about communication
protocole.
The problem is not to know first if we process float 32bits or integer 16
bits. The problem is first to define the communication protocol which will
help the host to implement this plug-in and the third part to program it .

This is a fine approach, but it presupposes that the plug-in will dynamically negotiate formats and capabilities with the host. I think there are others who would advocate using a simpler, static interface, where the formats and capabilities are defined in the spec.

The current goal is to choose the audience first, then debate features. It's hard to do this without making the obvious extrapolations from audience to target architecture to feature requirements. For example, external box developers -> fixed point DSP -> support for integer data types. I agree that a dynamically negotiated approach will tend to be far more inclusive of various target audiences, yet it may end up getting complicated too, like DirectX, and this in itself may lose an audience that prefers something really simple.

Regarding the fixed-point, floating point discussion, fixed point is always going to be cheaper than floating point, so it will never go away, rather it will migrate to cost sensitive applications. Witness the persistence of the 56K. If GMPI ends up supporting fixed-point DSP implementations, it makes sense to directly support fixed point data types. It would be really silly to pass 32-bit floats to a 56K.

Bill


----------------------------------------------------------------------
Generalized Music Plugin Interface (GMPI) public discussion list
Participation in this list is contingent upon your abiding by the
following rules:  Please stay on topic.  You are responsible for your own
words.  Please respect your fellow subscribers.  Please do not
redistribute anyone else's words without their permission.

Archive: http://www.freelists.org/archives/gmpi
Email gmpi-request@xxxxxxxxxxxxx w/ subject "unsubscribe" to unsubscribe

Other related posts: