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

Re: A Compromise to the Versioning Debate

From: Lachlan Hunt <lachlan.hunt@lachy.id.au>
Date: Mon, 16 Apr 2007 01:07:15 +1000
Message-ID: <46223FA3.5090906@lachy.id.au>
To: Mihai Sucan <mihai.sucan@gmail.com>
CC: public-html@w3.org

Mihai Sucan wrote:
> Le Sun, 15 Apr 2007 13:18:15 +0300, Lachlan Hunt 
> <lachlan.hunt@lachy.id.au> a écrit:
>> At this point, it seems all we can do is negotiate for a slightly 
>> less painful and disastrous outcome.
> 
> Your suggestion is not any inch better. It might be even worse: it gives 
> the *false* impression we can have an "always standards mode" (latest, 
> greatest standards mode) rendering.

Well, an "always standards mode" is exactly what we want and need, and 
MS has to provide that for us in one way or another.  The only other 
alternative is to give up and accept that we will never an "always 
standards mode" switch, but I'm not ready to give up yet.

> We all agree that an opt-in is *required* for improved rendering and 
> standards support in IE.next. 

Ah, no, we certainly don't all agree.  Many of us strongly disagree with 
that notion.  Adding any additional versioning will be a huge mistake 
with disastrous consequences.  But, as I said, MS has made it clear that 
we have absolutely no choice in the matter.

> I don't want the other UAs to "care" about which rendering mode I want
> in IE. However, that will probably prove to be inevitable as well: 
> Opera, Firefox, and Safari will probably "catch"/parse those
> conditional comments if they ever make it into any IE version.

That is the situation really must try to avoid.

> If new versions of IE will be released only on a 5-years basis, then, 
> yes, we will have a new opt-in for the rendering improvements, with each 
> release of IE.

That is a serious problem.  Every single one of those new modes will be 
undocumented, and other browsers will have no choice but to attempt to 
reverse engineer each one of them.

Given that MS has given us no choice in whether or not we have modes, we 
need a way to make those modes irrelevant to both authors who don't need 
them, and browser vendors so they do not have to reverse engineer them all.

My solution works slightly better because it assumes by default that 
pages don't rely on bugs that will be removed from the next browser 
version in incompatible ways, and doesn't force authors to learn a new 
opt-in every for every single browser upgrade.  But it provides a quick, 
efficient and cost effective temporary work around for web sites in the 
event that they were built to rely on such bugs.

There will be many sites (an increasing number of them built for 
standards compliance) that will not break with the next browser upgrade. 
   If we succeed in our goal of defining HTML5 in a way that is 
compatible with the web, it will be significantly more than than those 
that will break.

Microsoft's proposed solution is backwards.  It assumes by default that 
every site will break with the next upgrade.  That is significantly more 
harmful because it makes the existence of those modes much more relevant 
to other browser vendors attempting to be interoperable.

-- 
Lachlan Hunt
http://lachy.id.au/
Received on Sunday, 15 April 2007 15:07:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:15:53 GMT