Re: Smoothing the edges by example (was pilot case?)

On Tuesday, 7 May 2013 at 20:55, Brian Kardell wrote:

> Was having a brief conversation offline and it became evident that
> perhaps I did not explain my intentions well enough here, so let me
> try again.
>  
> The intent is not to formalize a process or anything like that!!!
> Instead, it is to find think about how we get 'from here to there' and
> figure out which kinds of things make the process easier.
>  
> The starting point can be anything: Currently some things start out
> even in existing WGs as nothing more than an email to a relevant list.
> In terms of prollyfills, what we need is a working implementation and
> some kinda doc that describes its proposal. We've had a few attempts
> at this already and they generally start out pretty informal. That's
> really fine. The goal of a working implementation is to get it out
> there, usable/useful to authors, potentially forkable, hopefully
> forward compatible, etc... That is where the evolution happens - it's
> a really critical part that has been missing in standards.
>  
> Assuming the 'slang' begins to pick up, and the rough corners get smoothed out
> (or forks and some alternative matures), eventually we'd expect to
> discover use cases, edge cases, etc - as it matures a decent
> expectation is that it will pick up some tests and better docs and so
> on. Eventually, as Brendan said, you'll need to work with
> implementors or people on WGs to cross the finish line.
>  
> Ideally, I think, things would get more "formal" as it goes... As
> implementers or WG folks get involved they have to cross a different
> bar and actually do have a formal process. How this could work well
> is really an open question. I don't pretend to have answers, I'm just
> opening the conversation.
>  
> My suggestion was simply that we pick something that foregoes a bunch
> of that and figure out what would be a convenient way to make a
> transition... Let's try some things and see what works and what
> doesn't. To that end, I chose something that is already
> specified/picked up by a WG and we just happen to have a fill for.
>  
> If implementors/WG folks and the community can begin to discuss, maybe
> try to work through a few, we could start to figure some of these
> things out. I was thinking that tests might be a really good thing to
> consider - if prollyfillers know what it is gonna take and have an
> easy way to hand off in a way that helps the WG or standards body that
> will ultimately take it up, perhaps they will choose to do so. If we
> can agree to some basic sanity things and set some good examples that
> make it easier to review these things, maybe people will follow the
> lead.
>  
> Thoughts?
We need to add value - lots of it. Prototyping and testing helps, but without being part of the target WG (or directly involved in the work, it's hard). Documenting the experiences would be of great value to the community, we just need to do it (once we get enough experience). I have experience from prollyfilling a few APIs with the WebIDL project… and I plan to do a few more in the near future (for SysApps).  

I'm happy brand that work as having come from this CG.

Received on Wednesday, 8 May 2013 11:00:52 UTC