Who wants a forgiving parser anyways? You write something with semi-broken syntax and test it in forgiving browser X and then it doesn't work in less forgiving browser Y. I fail to see what's supposed to be good about that.
It only wastes time.
Take this piece of css for example:
#something{
background:transparent url(image.png)4px center no-repeat;
}Works as expected with Firefox and Opera, but it doesn't work with IE6. Why? There is a space missing after the closing bracket! Took me hours until I finally found it (it was a fairly complex layout with a 3 level side navigation with lots of special classes). I actually gave up on that issue after about an hour and only found it by accident.
No one really needs a zillion ways to write the same thing, of which 90% work in browser X, 80% in browser Y and 5% in browser Z. (Well, we both know that I mean IE with browser Z.)
Imagine this - how would webdev be like if non vaild css or xhtml would display an error message and nothing else? It would be easier and quicker. And even IE5 would have been damn standards compliant, because the programmers wouldn't have wasted time with figuring out how to handle broken markup.
(Yes, I know that you get parser error messages for non well formed xhtml, that is... with the right Content-Type, but often that isn't an option.)
I validate my markup regularly with different validators. By the looks of it I also need to do the same for CSS now... it wouldn't have helped in this case tho.
Comments
Post new comment