[thelist] Validation > pressure on software companies??

[Arlen.P.Walker at jci.com]

>I've seen valid code that breaks on NN 4.5 Mac. So the validator seems to
me
>no measure to assure correct page display.

You're right, it isn't. The best analogy I can come up with for this is the
following program:

main()
{
   int x, y, z;

   x = 5;
   y = 6;

   z = x / y;

}

This program will compile without errors. Yet z will be 0, not 0.833333
(the real answer for 5/6). It is perfectly valid C code, it just gives the
wrong answer because the programmer used the wrong data types. It
validates, but is wrong nonetheless.

It's quite possible a valid HTML file will display badly. It's also
possible for a valid HTML file to display properly. Validation does not
replace testing. Validation only assures you that the code you've written
makes syntactic sense, which means that another browser that you haven't
tested, and which might implement the standard in a more strict manner,
won't have a problem with the code. It's up to you as the designber to make
sure that the valid code you've written actually displays the page the way
you want it displayed. Browsers are keyboard readers, not mind readers.
They do what they're told; it's up to you to make sure you actually mean
them to do what you're telling them to do.

>This leads me to the conclusion that the only reason validation makes
sense,
>is to discipline software companies to stick to some standard. I am really
>not understanding this otherwhise.

Pretty much correct. If you write valid code, then any user agent which
follows the standard will be able to handle your code. That includes voice
browsers, PDA's, and refrigerators.

And that's what's important about writing valid code. People who create new
user agents have a standard in front of them, and they'll code their user
agent to expect markup that follows that standard. They may also code it to
accept previous browser foibles, and perhaps even invent some of their own,
but I wouldn't be too sure that they would support other browsers faults,
and we can't predict what their new foibles will be. Invalid code which
works in a browser today may quite possibly break in a future browser,
especially one expecting the more rigidly defined XHTML.

I look at writing valid code like the auto mechanic in the (US-centric
reference alert) Fram oil filters commercial. I can either take the time
*now* to write valid code, or take the time later to rewrite what I've
written as valid code. It's a "pay me now, or pay me later" kind of thing.
I think the smart choice is to do it now.

Have fun,
Arlen
Chief Managing Director In Charge, Department of Redundancy Department
DNRC 224

Arlen.P.Walker at JCI.Com
----------------------------------------------
In God we trust; all others must provide data.
----------------------------------------------
Opinions expressed are mine and mine alone.
If JCI had an opinion on this, they'd hire someone else to deliver it.