- From: Brian Di Palma <offler@gmail.com>
- Date: Sat, 7 Dec 2013 23:48:07 +0000
- To: Domenic Denicola <domenic@domenicdenicola.com>
- Cc: Brendan Eich <brendan@secure.meer.net>, Ryosuke Niwa <rniwa@apple.com>, Erik Arvidsson <arv@chromium.org>, Maciej Stachowiak <mjs@apple.com>, Elliott Sprehn <esprehn@chromium.org>, Webapps WG <public-webapps@w3.org>
On Sat, Dec 7, 2013 at 11:07 PM, Domenic Denicola <domenic@domenicdenicola.com> wrote: > From: Brian Di Palma [mailto:offler@gmail.com] > >> From your email it seems you can still achieve everything you can with custom elements when not using them, it would just involve more code/boilerplate. > > I don't see how you can draw that conclusion from my email, except perhaps the trivial "everything can be implemented in terms of everything else in a Turing complete platform" sense. > > Indeed, you could say that shadow DOM gives you nothing but syntactic sugar for a DOM access library that doesn't allow you to access things marked with a ".shadowed" class, and templates give you nothing more than <script type="text/x-template">. > I guess you'd have to see how good the polyfill for each spec is. Can the shadow dom polyfill that Polymer uses give you the exact same power as the fully implemented spec in the browser? Much like a WeakMap can't be fully shimmed in ES5 there are issues shimming shadow dom, it's not just sugar, it's deeper. What about the custom elements polyfill? The shim for that spec doesn't seem to have any "known limitations". It looks more like sugar. > Everything is syntactic sugar at one level. But custom elements (by themselves) seem to add significant power by giving you parser hooks and allowing you to create, well, custom elements, which can be returned from standard browser methods like querySelector. > >> Splitting it up into separate specs is good, but it's clear they combine with each other. >> Making it easy and painless to combine them seems like common sense. > > I think that there are many ways to combine the specs, and most of them are pretty darn easy already. Making one of them even easier by introducing the proposed coupling doesn't seem like a win. I'm not sure we will have an answer to whether this parameter/proposal is useful or not until people actually start making use of these specs in routine development. It's not a huge cost to waiting and seeing, it's a shimable and minor change after all.
Received on Saturday, 7 December 2013 23:48:35 UTC