- From: Ian Hickson <ian@hixie.ch>
- Date: Thu, 7 Aug 2014 21:47:53 +0000 (UTC)
- To: Domenic Denicola <domenic@domenicdenicola.com>
- cc: "public-script-coord@w3.org" <public-script-coord@w3.org>
On Thu, 7 Aug 2014, Domenic Denicola wrote: > > In [2], Anne brought up that it would be nice to standardize [Reflect] > in WebIDL. I agree, except that it is weird for WebIDL to define > something that only applies to HTML element interfaces. Not that weird. The "Web" part of "WebIDL" is centered around HTML to a large extent. IMHO when we lose sight of that we risk doing things like designing DOM APIs around Java constraints, and we all know where that leads... Personally, the main reason I haven't pushed for [Reflect] is that it's not that simple. For example, you have to distinguish between reflection of content attributes that contain regular strings from those that contain URLs. You have to map attribute names like "httpEquiv" to "http-equiv" and "useMap" to "usemap". You have to note which reflections are "normal" and which are "limited to only known values". And so on. At the spec level, it's just simpler to define it in prose. I don't know how implementations of [Reflect] handle all these edge cases. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 7 August 2014 21:48:17 UTC