- From: Ian Hickson <ian@hixie.ch>
- Date: Thu, 3 Sep 2009 07:31:20 +0000 (UTC)
- To: Mark Baker <distobj@acm.org>
- Cc: "public-html@w3.org WG" <public-html@w3.org>
On Thu, 3 Sep 2009, Mark Baker wrote: > > How it's handled/processed depends upon the type of user agent. > Browsers might handle things one way, screen readers another, and > spiders yet another. Sure, there might be *some* common behaviour > between *some* classes of agent, but there's only *two* things which are > common to *all* agents, past, present, and future; the shared > understanding of the meaning of the document, and the recognition of the > text/html media type. No, this isn't true. You can define processing requirements -- indeed, HTML5 does define processing requirements -- without being so wishy washy as to rely on a "shared understanding of the meaning". For example, there are very specific rules for how to process sections and headings, and how they relate to outlines. These requirements apply to a huge number of conformance classes -- editors, screen readers, search engines, validators, outliners, etc. What obsolete features are we lacking suitable "shared understanding" for? > > what features are lacking these requirements? > > I will look for more, but for now, let's just get those two attributes > defined. Which two attributes? I'm not aware of anything that's lacking suitable conformance criteria. Since I fixed <head profile> and <meta scheme>, I've not heard anyone list for any other features that need work. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 3 September 2009 07:28:52 UTC