Re: Reading some polyfills

On Sat, Nov 24, 2012 at 10:30 PM, Marcos Caceres <w3c@marcosc.com> wrote:

>
>
> On 25 Nov 2012, at 00:27, François REMY <francois.remy.dev@outlook.com>
> wrote:
>
>  I propose Brian look at some CSS polyfills/preprocessors he didn't wrote
> himself and I'll be looking at some HTML polyfills (I'm specifically
> looking for HTML5 input tags and also x-tags which emulate web components).
>
>
> Feel free to report on what you're going to work to avoid duplicates ;-)
>
>
> Ok, cool. I'll stick with DOM/HTML also: DOMTokenList, @srcset,
> picturefill, and the like.
>
>
>  ------------------------------
> De : Marcos Caceres
> Envoyé : 24/11/2012 23:55
> À : François REMY
> Cc : <public-nextweb@w3.org>
> Objet : Re: Reading some polyfills
>
>
>
> On 24 Nov 2012, at 20:47, François REMY <francois.remy.dev@outlook.com>
> wrote:
>
> > Okay guys, I’m back from Dublin. I’m not going to have a lot of time to
> spend on this but I plan to read some popular polyfills this week. Feel
> free to do the same and share with the others when you find something
> interesting ;-)
> >
> > François
>
> If we do 5-10 each, we should get a good sample of best practice (let's
> aim for 30 initially). But let's please coordinate in the wiki so we don't
> end up reviewing the same ones.
>
> What are we looking for specifically?
>
>
Guys, am I confused - I thought that we agreed last week that we were
focusing on prollyfills, yet the wiki currently doesn't event mention
"prollyfill" - but uses "polyfill"... Did I miss something?  Let me
clarify:  @srcset seems to me like a prollyfill because I see no native
implementations at all for that, but <details> for example has significant
native implementations already - that seems like a polyfill to me and
without question the input placeholder is as literally every new/major
browser has support.

Polyfills have been around a long time and there are lots of sites and
libraries dedicated to them and powerful advocates -- anything that we
could add in that area would be merely filling "short strokes".
 Prollyfills on the other hand are basically a blank canvas/pretty new
concept/entirely unsolved problem and I'd like to suggest that we try to
focus on things which are not virtually solved problems lest we burn our
limited resources and risk getting the message watered down as well.

Mutation Summary could be a prollyfill -- if it had a draft proposal.
 Daniel's selector listeners could be a prollyfill -- if it had a draft
proposals.  Adobe's recent Regions changes are a prollyfill.  Everything in
Hitch is a prollyfill.  Some x-tags elements could be prollyfills and
x-tags itself is a prollyfill.  There are new things on the lists every
week that could be as well.  I'd rather be analyzing things like these and
finding overlap, discussing, developing best practices and common bits that
make these things easier - like the CSS Parse model I mentioned and Marco's
generator from WEB-IDL.

Just my opinion.