Re: W3C Polyfill service?

On Wed, Oct 19, 2016 at 9:51 AM, Travis Leithead <
travis.leithead@microsoft.com> wrote:

> To some extent, the W3C’s investment in web-platform-tests is a similar
> thing—focused on getting interoperability through testing. What would be
> the goal of a polyfil repository? Would they count as an implementation of
> a feature? Could they also be used to help the testing effort? If so, then
> I would love to see them integrated into web-platform-tests.
>
>
>
> (On a side note, I resorted to a polyfill to help test the DOM Parsing and
> Serialization spec—it proved very helpful for “debugging” the spec’s
> algorithms.)
>

​On that note, I have also developed a few polyfills for EME features,
which enable us to develop and test the tests, without full browser support
of the feature or in the presence of browser bugs (provided they can be
worked around in JS). This is particularly useful for developing tests for
features which are only available on a platform which is not your main
development platform: you can develop and debug the test quickly and then
run the test on the other platform only to get the results.

It's not clear to me why an implementation of a feature purely in C++
should "count" any more as an "implementation" than one that is implemented
partly in C++ and partly in JS. In fact IIUC, some parts of some browsers
are already implemented in JS. The only difference between a full browser
implementation and one which includes "polyfill" code is that the browser
implementor has "blessed" the former.

...Mark



>
>
> *From:* Andrew Betts [mailto:andrew.betts@gmail.com]
> *Sent:* Wednesday, October 19, 2016 1:47 AM
> *To:* www-tag@w3.org
> *Subject:* W3C Polyfill service?
>
>
>
> At a W3C Japan meeting in Tokyo yesterday, which I attended at Keio
> University, polyfills were mentioned frequently (I noticed this
> particularly because it was one of the few words I could understand when
> presentations were in Japanese).
>
>
>
> Satoru Takagi from KDDI (a large mobile operator) subsequently suggested
> to me that W3C could provide an 'official' source of spec-compliant
> polyfills for new features (something like "polyfills.w3.org"), and
> adopting polyfill.io as a starting point for this would seem to be a
> sensible approach.
>
>
>
> His slides:
>
>
>
> http://www.slideshare.net/totipalmate/svg2-candidate-
> recommendation-in-english
>
>
>
> As the maintainer of polyfill.io, I'm open to this idea.  There's
> precedent for this in things like the validator services W3C operates, and
> there would likely be huge developer interest and adoption. W3C could bring
> governance and administration support to the project and the existing
> participants could continue to perform their existing roles.
>
>
>
> I mentioned it to Mike Smith at the event, and he was skeptical, but I
> thought it worth asking if anyone thinks this is worth pursuing.
>
>
>
> Cheers,
>
>
>
> Andrew
>