Why won't C9 validate?

Posted By: turrican | Aug 11th @ 3:52 PM
page 1 of 1
Comments: 19 | Views: 915
#Aug 11th @ 3:52 PM
turrican
turrican
Condemnation without investigation is the height of ignorance! - Albert Einstein

View source C9 :

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en-US">

Nice... but :

111 Errors »

C9 @ validator.w3.org »

Why won't C9 validate? ...if I can do it, why can't you? I'm just very curious. Is it Microsoft policy or do the developers of C9 just don't care?

#Aug 11th @ 5:03 PM
mVPstar
mVPstar
I'm white because I smelt an onion.

Because it's not absolutely necessary.

turrican said:
if I can do it, why can't you?


Most corporations care about their site working first for a wider audience and don't want to spend time (and money) trying to design sites that validate AND look good and function well.

Anyway, that's 111 errors with only 3-4 distinct. Not bad actually. I had expected at least 50 distinct errors.

Is the page displaying poorly for you or something?

#Aug 12th @ 11:33 AM
HumanCompiler
HumanCompiler
Compiling humans...and code
That's pretty much what I would've said.  We care, but it's low priority.

Try looking at other major web sites out there.  We're doing better than a lot of them.
#Aug 12th @ 4:53 PM
mVPstar
mVPstar
I'm white because I smelt an onion.
Most of the errors have to do with div class=button being nested in anchors. If you changed that to a span, you'll be down to like 10 errors instantly.
#Aug 12th @ 9:45 PM
turrican
turrican
Condemnation without investigation is the height of ignorance! - Albert Einstein

Ok I undrestand.

But don't you guys write the code in Visual Studio or something? I mean is it your tool generating non-valid code or is this writen by "hand"?

The thing is I'm currently using Visual Web Dev and I'm just worried how much I really should use this tool if it maybe stops creating valid stuff or maybe doesn't always creates valid code or something.

I had the same way a couple of months ago though, "As long as it works, doesn't need validation" but I think one will hit a brick wall at some point. I don't think that making a site validated takes all that much of a time, specially for big corps... but I do see your point. It's basically company policy not caring about validation.

I'm not even sure why the validation gotten to my head so much lately *lol*

#Aug 13th @ 8:24 AM
mVPstar
mVPstar
I'm white because I smelt an onion.
turrican said:
But don't you guys write the code in Visual Studio or something? I mean is it your tool generating non-valid code or is this writen by "hand"?


Seeing as how I'm representing HumanCompiler now...
Custom web controls, which are written by hand. 

Tongue Out

 
turrican said:
I'm not even sure why the validation gotten to my head so much lately *lol*


I was the same way, not too long ago. W3bbo's pedantry put me in "standards standards standards" mode. After a while when you hit a few brick walls and find yourself going through complicated scenarios just to get something simple to work (while spending a lot of time doing it), you'll realize that it's better to find a balance between proper validation and functionality that works best within your time constraints, as well as level of motivation.
#Aug 13th @ 5:56 PM
turrican
turrican
Condemnation without investigation is the height of ignorance! - Albert Einstein

I blame W3bbo too! Tongue Out his fault... however, I have now managed to find a good balance on the stuff I need so I'm 100% validated now and will continue keeping it that way. I think the secret is simplicity. I think just using basic DIV positioning and some very basic async ajax stuff, you can do wonders. I will share with you guys the results in a couple of weeks.

Smiley

thanks.

#Aug 13th @ 10:41 PM
SlackmasterK
SlackmasterK
I write my OWN blogging engines
Someone 'splain this to me... If a page renders and behaves right in both major browsers, why does validation matter?
#Aug 14th @ 12:02 AM
turrican
turrican
Condemnation without investigation is the height of ignorance! - Albert Einstein

If a bike rolls fine on the street, does it matter if the screws of it sits tight enough? ...Perhaps it will break somewhere in the line. It's like building an IRC client which isn't RFC compatible or something. The whole idéa av WWW is that it is accessible to people... "Major browsers" doesn't mean "All browsers"... but that's just me. I'm new to this thing too, so I don't have many arguments actually.

I just like to make things "right", you know... "Good quality code".

#Aug 14th @ 1:43 AM
blowdart
blowdart
Peek-a-boo
The thing is you're quoting an ideal. IE breaks with some validating code. As does Firefox. That's what ACID2 and ACID3 set out to prove.

Your analogy is a bad one, if the bike rolls fine on the streets does it matter what colour it is might be more accurate.
#Aug 14th @ 3:10 AM
stevo_
stevo_
Casablanca != Manchester

Validating websites against the w3c automated validator.. serious business!

#Aug 14th @ 3:51 PM
turrican
turrican
Condemnation without investigation is the height of ignorance! - Albert Einstein
!?
#Aug 15th @ 4:06 PM
matthews
matthews

In a lot of cases, perfectly valid code doesn't work on all web browsers. If I'm developing a site, I test it against IE7, Firefox, Safari, and Opera, and if it breaks in something else I really don't care (I know I'm leaving out IE6, but then most websites I make are for personal stuff not serious business). I'll worry about valid code when IE8 becomes very popular and until then they'll be a little content="IE=7" in my <meta> tags.

#Aug 15th @ 7:17 PM
turrican
turrican
Condemnation without investigation is the height of ignorance! - Albert Einstein
One of the sites of one of my customers has around hm... 5-6000 customers a day, I seen IE6 going down lately to around 21% ... hopefully, this continues to go down to perhaps less than 4% so we can "finally" move on to a validated website.
#Aug 15th @ 9:19 PM
CannotResolveSymbol
CannotResolveSymbol
{insert caption here}
Validated != Renders properly everywhere.

Although the Acid 2 test has helped to get some consistency between browsers, there's still a lot of variation between web browsers (because of ambiguity in the specs combined with the fact that there's no reference renderer to check yourself against).  And when you get to Javascript/rich content, it gets even worse (take a look at your rich text editor problem in Opera:  Opera has had problems with their rich text editor for a while now:  it simply doesn't behave like it does in Firefox and IE (Safari's used to work differently too, but that's been fixed in the last version or two).  No amount of code validation is going to fix that:  you're going to have to work around different browsers' quirks whether you validate or not.
#Aug 15th @ 9:48 PM
turrican
turrican
Condemnation without investigation is the height of ignorance! - Albert Einstein

I'm not sure if I agree with you. So far, building this new website, with ajax and all I have managed to keep it validated and it actually looks 100% the same in IE7, O, S, FF.

I would like to see an editor which Is validated And breaks in Opera. Do you have an example?

#Aug 15th @ 10:32 PM
CannotResolveSymbol
CannotResolveSymbol
{insert caption here}
http://www.kevinroth.com/rte/demo.htm validates as XHTML 1.0 Transitional and behaves the same way the C9 editor does, generating double line breaks when enter is pressed.

Of course, there's a good reason it does that:  Opera's RTE generates <p></p> when you press enter rather than <br /> like IE or Firefox does.  That's the kind of cross-browser inconsistency that no degree of validation is going to fix (unless the spec is revised to unequivocally say what browsers are supposed to do in that case).
#Aug 15th @ 10:39 PM
turrican
turrican
Condemnation without investigation is the height of ignorance! - Albert Einstein
Sorry, I meant "1.0 strict", I can undrestand this in a Transitional document.
#Aug 15th @ 10:54 PM
CannotResolveSymbol
CannotResolveSymbol
{insert caption here}
The two validation errors presented if you change the doctype of the above page to XHTML 1.0 Strict do not affect the rendering of the page (language attribute on a script tag, and name attribute on a form tag).  I'd put up an example that validates as XHTML 1.0 Strict (as it requires changing two lines), but I don't have any web space to put it on.  Feel free to do it yourself if you feel so inclined.
#Aug 27th @ 1:21 PM
Duncanma
Duncanma
Just Coding for Fun...
We did spend time on this issue, by the way... but I only see two real errors that we can control ... one is that we have a <div> inside a <a> ... which does cause problems on Opera, so we render it differently for them... but our styling is dependent on the <div> at the moment so we leave it alone on other browsers. The other issues is that we have <img> elements with no source. That is an interesting one (at least to me) ... we need the <img> element there to put images into as you page through the site... so we started out with <img src="" /> (which validates) but running traces we noticed that syntax was causing FireFox to load the entire page (whatever page the <img> element was on) in as the image and therefore causing the user to retrieve twice as much data as was necessary. So we overrode the default asp.net image control to omit the src parameter if it was blank. That solved the double-retrieve issue, but added the validation error you are seeing.

From what I can see, the editor is causing the rest of the validation errors. We could bug Telerik about it, but it is a low priority for us so we probably won't worry about it for right now. They release frequent updates though, so some future update may end up resolving these issues.