Re: The argument for |bugmode| (was Re: If we have versioning, it should be in an attribute, not the doctype)

Alfonso Martínez de Lizarrondo wrote:
> Let's analyze the situation:
> You've built a web page, it's rendered correctly and works properly in
> all the current browsers.
> IE.next is released with a new standards mode, there are some possibilities:
> 
> 1.Your page did relied on a bug and it's fixed in the new IE.
> 1.a If the opt-in is automatic for every new release then your page
> now is broken and you must check and fix it.
> 1.b If you must set the opt-in for the new standards mode then your
> page will remain fine until you add the opt-in and fix the problems.
> 
> 2. Your page didn't relied on any bug fixed in IE.next. No matter if
> there's an opt-in or not, no matter if there's a new Standards mode or
> not, everything works fine and everybody is happy.

3. Your page works fine, but IE.current doesn't support some parts of it
yet / doesn't do things as well as it could. You accept that (as the
page remains legible), and hope IE.next will be getting to the same
level as all other browsers. Then IE.next is released, and indeed is as
capable as all other browsers... but it still won't render your page as
intended, and the only way to get it to do so is by adding some new
attribute. (And worse, if this IE.next only added _some_ capabilities,
then for IE.next.next you'll need to do it _again_.)

This is what's not acceptable in having to opt-in to the best IE has to
offer - we'd be punishing those developers who do things Right (TM),
thus removing any incentive whatsoever to follow the standards.


In the past, with 95%+ marketshare for IE, authors might have been
unaware that they were relying on IE-specific bugs, and fixing those
bugs might indeed have been breaking them. But I submit that in this day
and age (and hopefully for all time to come), there's a diverse enough
browser ecosystem that no one can develop websites and not test in other
rendering engines, thus becoming aware of browser bugs they might be
relying on. Being aware of the existence of the bug, they can work
around it or not as they please, but either way they can right then
opt-in to "bugmode" or opt-out of "most standards mode", and won't
(shouldn't?) be surprised if the bug gets fixed in IE.next. (And if they
turn out surprised anyway, they'll have a very simple way to "fix" their
site and make it look as intended again.)

Cheers,
Sander

-- 
Sander van Lambalgen   * Have Skill Webdevelopment
http://have-skill.com/ * "Have Skill, Will Travel"

Received on Thursday, 19 April 2007 19:59:39 UTC