W3C home > Mailing lists > Public > public-html@w3.org > April 2007

RE: Versioning and html[5]

From: Chris Wilson <Chris.Wilson@microsoft.com>
Date: Fri, 13 Apr 2007 12:20:40 -0700
To: Lachlan Hunt <lachlan.hunt@lachy.id.au>, Mihai Sucan <mihai.sucan@gmail.com>
CC: "public-html@w3.org" <public-html@w3.org>
Message-ID: <5C276AFCCD083E4F94BD5C2DA883F05A27D7192747@tk5-exmlt-w600.wingroup.windeploy.ntdev.microsoft.com>
Lachlan Hunt [mailto:lachlan.hunt@lachy.id.au] wrote:
>> Using something like the IE conditional comments sounds good for me.
>
>No, absolutely not!  Do not force authors to use UA-specific opt-ins
>like conditional comments just to get the most standards compliant
>behaviour.

Great!  Then leave a version number on the identifier for HTML, because I've clearly stated long-term, that is how we will get away from UA-specific opt-ins for "best standards compliance".

><!DOCTYPE html> should trigger the same standards mode as existing
>DOCTYPEs, and that should always be the most standards compliant mode.

No.  Web developers don't write to some abstract understanding of the standard- they write to an implementation.  If "the standard" (and remember the current standard in place is probably IE6/7 behavior) changes with no version switch, their pages break.  If an implementation suddenly changes behavior, they're also broken.

>Authors that use such DOCTYPEs have already implicitly opted in to the
>latest standard, don't force them to keep changing that opt-in method ad
>infinitum.

At a high level, for IE there can be no content that just gets automatically opted in to the "new behavior", if we are changing anything but crashes, hangs and security problems.

>Chris Wilson seems to have the fallacious notion Firefox, Opera, Safari
>and other browsers can implement the HTML5 spec and be compatible with
>the web, yet IE can't implement the *same* spec without significantly
>breaking it.  That is just completely illogical.

I'm trying hard not to resent your implication that I am crazy and/or stupid.  I would highlight again what I started this thread with - I have a different view of the world.  One that is rooted in being the most-adopted browser, the most-relied-on-to-be-compatible browser.  Remember that damn near every complex page on the web today DOES use UA switching or CSS hacks or some other form of opt-in, for IE the most but for other browsers too, despite the running sentiment from a contingent here that it just shouldn't be so.

They expect US to be compatible with IE6/7 behavior.  They may or may not expect Safari to be compatible with it; Safari may or may not care, as Maciej eloquently stated, as their compatibility philosophy is different from ours.  I don't claim theirs is wrong, I do claim Microsoft's is valid.

>As as been stated many times in this thread (though, it clearly bears
>repeating since Chris doesn't seem to understand):
>
>   *if something in the spec can't be implemented without
>    breaking compatibility, we should fix the spec!*

Then your spec is a binary copy of IE7.  Or IE6, if you prefer.  Start with all the reasons that's a bad idea.  Remember that WE (Microsoft) get sued if we disrupt businesses.

Look, I'm not trying to force you all to live in my point of view.  It IS necessary that you understand it, and it is necessary that the standard take it in to account.

-Chris

Received on Friday, 13 April 2007 19:20:55 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:38:42 UTC