Microsoft Hatred FAQ

[Mike Meyer]

"Mike Schilling" <mscottschilling at hotmail.com> writes:
> "Mike Meyer" <mwm at mired.org> wrote in message 
> news:86y84rxryr.fsf at bhuda.mired.org...
>> "Mike Schilling" <mscottschilling at hotmail.com> writes:
>>> "Mike Meyer" <mwm at mired.org> wrote in message
>>> news:86fyqzzskt.fsf at bhuda.mired.org...
>>>> "Mike Schilling" <mscottschilling at hotmail.com> writes:
>>>>> What matters in generating HTML is which browsers you want to support 
>>>>> and
>>>>> what they understand.  Standards and recommendations are both 
>>>>> irrelevant.
>>>> Unless, of course, you want to support any compliant browser.
>>> Since no browser I know of is perfectly compliant (e.g. bug-free), that's
>>> not a feasible goal.
>> I guess you'd say developing any software isn't a feasible goal,
>> because it'll never be bug-free, will never have bug-free compilers to
>> compile it, bug-free linkers to link it, bug-free GUI/db/etc libraries
>> to link with it, bug-free servers to communicate with, and bug-free
>> operating systems to run it on. Fortunately, most developers aren't
>> quite that anal, and realize that you can get useful work done in a
>> less-than-perfect environment.
> I'm not speaking theroetically. My company (though not me personally) 
> creates browser-based UIs, and one of the biggest expenses has been dealing 
> with IE rendering bugs   Given the market share of IE, the fact that 
> something should work, and even does work in Firefox, Opera, etc, is 
> irrelevant.  If it breaks IE, we can't use it.

Been there, done that, threw out the T-shirt as to ugly to wear.

Yes, you have to work around bugs in the popular browsers. That hasn't
changed since the first published specs showed up. That doesn't mean
you throw out the standards and only support a trivial set of
browsers. That means you restrict yourself to a subset of the
standard, or - better - detect the deficiency and fail soft, the same
as you would do when you get a visit from someone who's disabled some
feature you want to use. In extreme cases, you wind up implementing
something twice: once for busted-but-popular browsers, and once for
people using browsers written by developers who read specifications.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.