Re: [mif] Adoption of API document as the MIF WG document

[Ted Lemon <[e-mail]>]

On Dec 16, 2011, at 11:25 AM, Simon Perreault wrote:
I feel that the API draft tries to standardize the various incompatible APIs we have on various platforms, the very APIs that are currently used by apps for their hackish behaviour. I think it will make it easier for applications to implement hackish behaviour. It will perpetuate the mess we're in.

This is the same point you raised at the meeting in QC, and it's wrong for the same reason: this API is not intended to be used by applications.   It is intended to describe all of the functionality that's required to implement a high-level API, which would then be used by applications.

There is pretty broad agreement that no single high-level API will address every use case: happy eyeballs is nice for certain use cases, but not for all, for example.   So we could try to enumerate every possible sort of high-level API and then document each of these, or we can try to tease out the functionality that any such API would require, and document that.   This spec is intended to serve the latter purpose.

The question of whether applications will program to this API is then orthogonal to the question of whether the API itself is useful.   You propose that if we don't define an API, applications will not use it, but in fact each operating system already has an incomplete API that looks something like what we are trying to document here.   So right now, applications can already misbehave in the way that you fear.

However, right now, if they do so, they will do so in a different way on each operating system, and indeed may simply not work on some operating systems because those systems do not provide any means to accomplish with a MIF-capable application needs to accomplish.   So by not documenting a low-level MIF API, what we do is make the scenario you are concerned about worse, not better.