[UPHPU] Validation

[Mac Newbold]

Today at 5:07pm, Jonathan Duncan said:

>> I have a question.  I am referring to the recent "CSS and IE" thread.  I
>> know I will probably be laughed off the list or flamed or something, but why
>> is validation so important?  Is is just to be able to display the validation
>> image?  I know it's important to have standards but sometimes it seems we
>> give up what we want so that our code will validate.  I usually ignore
>> validation.  If I can get my code to work properly and it's (somewhat)
>> useful then I feel that's 90 percent of the battle.  IMHO.
>
> Good question:
>
> http://validator.w3.org/docs/why.html

They present some reasonable arguments, but they're not very practical 
about it. They say:

"Well, firstly there is the very practical issue that non-valid pages are 
(by definition) relying on error-correction by a browser. This error 
correction can and does vary radically across different browsers and 
versions"

If all browsers were standards compliant, that would be perfectly true and 
reasonable and practical. But they aren't. If you make standards compliant 
pages that validate, but they use things from the standards that aren't 
implemented by MSIE (or any other browser), or are impelmented wrong by 
that browser, they won't work properly in that browser, despite their 
correctness and validation.

Most people who have tried to do anything remotely complex or out of the 
ordinary and have it work on MSIE+Firefox+Safari know how small the set of 
instructions is that works the same in every browser. That is the subset 
that I out of necessity must code do, and I guess you could call it a 
"standard" of sorts.

This ties into the "table-less" issue too. I don't agree with it, and find 
that it produces much simpler code in a much shorter time to use tables 
where they help, since it is one of the things that the browsers do 
reasonably compatibly.

Mac

--
Mac Newbold		MNE - Mac Newbold Enterprises, LLC
mac at macnewbold.com	http://www.macnewbold.com/