- From: Matt Morgan-May <mattmay@adobe.com>
- Date: Thu, 15 May 2008 10:46:06 -0700
- To: Henri Sivonen <hsivonen@iki.fi>, Gez Lemon <gez.lemon@gmail.com>
- CC: Dave Singer <singer@apple.com>, HTML Working Group <public-html@w3.org>, W3C WAI-XTECH <wai-xtech@w3.org>
On 5/15/08 12:43 AM, "Henri Sivonen" <hsivonen@iki.fi> wrote: > The notion that a syntax specification should require software > conforming to the specification to produce syntactically non- > conforming output under some circumstances is patently bizarre. We > shouldn't require something that is bizarre in a way that it doesn't > fit the software developer mindset, because then we don't get the > reactions we want from software developers. So what should an authoring tool do when someone creates an <img> element with no @src? (Not a 404, a missing @src.) What about a missing </table>? An unclosed attribute value? An unterminated entity? Guessing at any of these things can have unintended negative consequences. Authoring tools can prompt and warn for broken or missing content (as ATAG specifies, and as Dreamweaver does), but if the author goes about breaking the code in ways the tool can't automatically repair and then says "now, save it," the tool has no choice but to save. Even Amaya saves invalid HTML content when it can't repair it authoritatively. (Though it purposely refuses to parse malformed XHTML -- a feature when you're editing a TR document, but one that would put it at a severe disadvantage if it were to compete for users.) Authoring tools _must not_ guess in order to make invalid content valid. That includes neglecting @alt. If the resulting content is invalid, especially when a tool has provided the author with every necessary facility to make it valid, there's nothing more for the tool to do. - m
Received on Thursday, 15 May 2008 17:46:56 UTC