Re: Repo: should we take it up?

On Jul 1, 2013 1:25 PM, "Angelina Fabbro" <afabbro@mozilla.com> wrote:
>
> I'm of the opinion that the sooner we can get developers using these
things the sooner we can fail fast, hard, and early and then iterate to
refine our specifications and implementations. So yes, let's get a repo
going and see how that experiment goes. I'm tired of
waterfall-design-by-committee, it's not the most efficient way to get at
the best answer.
>
> - Angelina
>
>
> On 2013-06-29 12:54 PM, Brian Kardell wrote:
>>
>> This has been discussed in a lot of contexts, but I'd like to propose
that maybe we are far enough a long that we should: a) decide if we want to
take this up as a group b) if so, what exactly should it's goals be.
>>
>> My own personal view is that we should, and here are my thoughts about
goals.
>>
>> minimally:
>>
>> - provide a repo, under prollyfill.org maybe
>> - give devs an easy way to submit and a place to discover them
>>
>> Ideally though, i think it has a lot of potential uses we have discussed
on the list too.
>>
>> It could provide metadata that tracks useful information from both the
dev point of view and implementer/standards body points of view.
>>
>> It could create a sort of simple, more grassroots/public, shadow process
toward standardization of "slang".  By defining some hurdles and milestones
it would be possible for developers to see something like a caniuse and
manage their risk tolerance based on how far along something is.  Simple
things like whether it has been peer reviewed, whether it has tests,
whether (and how many) members of a relevant working group endorse the
proposal, how many people "like" it, etc would help.  We would have to
discuss details, but this might be a way to reduce the noise enough to
allow standards folks to follow things that reach a certain level.  It
might also collect information about whether it is a plugin for a framework
vs something standalone.  This could be useful because it gives a sense at
how useful the framework is, what sorts of use-cases it enables, etc.
>>
>> Likewise, since this group is vendor neutral, we might be able to think
about how to set up things (at least those that meet certain criteria) as a
trusted source CDN and see if vendors can give it some special treatments -
yoav's special-cache thing, for example.
>>
>> Others?  This is intended to start a discussion more than be a very
specific proposal.
>>
>> Brian
>
>

I've gotten offlist responses in the affirmative as well, but not enough
that I am sure the group wants to take this up.

The easiest thing to do seems to be to create a github repo under the
extensible web org with a json file containing the metadata we want and
just let folks send pulls.