1  Tomas Jogin  24 July 2004 23:57

I agree with you, however, I also think that if you find yourself in the position of asking your peers for some kind of help on your code, which does not validate, then you should not be asking them for help, until your code does.

I seriously cannot count the times I’ve helped a colleague or a friend with their code (html, php, asp, c, etc), by correcting a “minor” error, or warning, or even just a not-so-bullet-proof method of solving a problem, which ultimately was the entire problem. If those people had coded more thoroughly, instead of burying their code in inefficiencies and uncertainties, then they wouldn’t have wasted my time. I would even say that if they had, then they would also have learned something, and wouldn’t have to ask me for help all the time.

Again, this zero-tolerance with perfect code is only necessary when you’re asking for help, not when you’re publishing a website (although perfect code under those circumstances never hurt anyone, either).

2  Justin Goodlett  25 July 2004 00:01

Right on D! I have to agree with you on this discussion. I believe that markup-validation matters to a certain extent, but there seems to be many people out there at the moment that look at valid markup as the written word. To not validate would be on the verge of mortal sin. Really! I don’t think so. Does your site validate? Great. If not, who cares? Does your site function and work the way it was designed to? Even better…

3  Morgan Aldridge  25 July 2004 01:12

I definitely agree that validation itself is not an end-all solution, nor is failure to validate proof of an inferior site or designer.

Thomas his right on with his comment. It’s an excellent way of knowing you’re on the right track (but does not imply you’ve got it right, as you said) and that can be infinitely helpful when asking for help or sharing code. I personally use validation in that way, but get lazy and ignore it too long at times. It’s a great tool when used.

I feel that another point to validation is to look at what it takes to have correct markup. The goal is to eventually get away from the lazy handling of poor markup that browsers such as Internet Explorer do, and hence browsers like Safari must emulate, for a web mostly comprised of xml. Of course, we’ll probably never see the day that we can completely disregard browser-specific hacks, but we’re that much better off the closer we do get.

4  Faruk Ates  25 July 2004 01:17

Didier said: “So, keep those small buttons and please care about standards, just do not bust my ass about an ampersand that is not encoded. Deal?”

Like your IMPOSSIBLE-TO-READ pixelhump buttons, eh? ;)

(hey, you told us today to comment on those, we’re just complying. That’s what being standards-compliant is aaaaaaaaall about! ;))

5  Mike D.  25 July 2004 07:01

Agreed on all points. Validation, right now, is most useful to developers and not users. There may eventually come a day when we live in a strict XHTML world and the user will benefit from error-free code, but right now, the content/display/usability/interactivity is a lot more important than whether or not all rules were followed. That’s not to say you shouldn’t try to follow rules… only that your degree of compliance to these rules (good or bad) does not eclipse all of the other things which go into publishing a web site.

As you say, when factors outside your control conspire to create invalid code, it’s neither your fault nor anything to lose sleep over. Joe Clark and I will disagree with each other about this probably until the day we both die, but that’s just fine with me. I respect him for the work he does in encouraging accessible web publishing, but I just don’t think minor validation errors have much to do with that goal at this point in time. Will it eventually? Maybe. And by then we’ll all be writing better code anyway.

6  Joe Clark  26 July 2004 00:24

Yes, well, all this would be fine if I were aware of a clear and demonstrable trend of, for example, bloggers naming and shaming other bloggers for their invalid code.

We’re not Slashdotters. We’re not sitting here on the lookout for a single deviation from catechism that’s somehow gonna invalidate your entire project. As I have written many times before, minor validation errors (typically involving ampersands) are just that: Minor. They’re also fixable.

Now, there are many beautiful, easy-to-use sites with crappy markup, but increasingly, the beautiful, easy-to-use sites have great markup. Much of the time, they can have perfectly valid markup. And when they don’t, pointing it out neutrally or politely (as I did just last week) is not the same as what Didier implies— some kind of pretext to dismiss and indeed *invalidate* the entirety of the work presented.

So I really don’t know what Didier is talking about. I don’t see masses of standardistas bitching out other would-be standardistas because they failed to close an img element with a slash. I just don’t see that happening.

Nonetheless, whoever wants to present himself or herself as a designer committed to standards compliance *must comply with the standards*. At root, that means valid code. We can imagine times when ugly valid code is necessary (as in the use of the i or b elements when nothing else works), and there are other times when, apart from valid code, we do something nonstandard (like blank alt text when the subsequent plain text suffices as text equivalent), but really, you can’t claim to support Web standards when your own site does not comply with them.

I see the entire argument as a political statement based on a lie. The only parallels I can think of come from anti-gay fundamentalists, so I won’t spell them out. Nonetheless, someone utters the lie and then people automatically think it’s actually happening. I mean, it isn’t. There is no trend to bash people for minor standards failures.

And we know very well that large sites with many content contributors are difficult to keep valid. We know that. But that too is largely an abstraction. We don’t have a lot of commercial sites with many content contributors that ever *pretended* to have valid code in the first place. The objection is vapourware.

Should anyone doubt that Web standards have bearing on accessibility, though, note that the two leading screen-reader manufacturers say it does.

I see this entire discussion as a spirited and impassioned defense against an accusation nobody made. I view it as an overreaction to a nonexistent stimulus. I see it, essentially, as false.

7  Eris  26 July 2004 01:03

Hehe, Standardista, I love it. That’s one for the business card.

</irrelevent to discussion>

8  AkaXakA  26 July 2004 04:24

I agree with Didier, the validation is a tool, not a goal in it’s own right.

It can help you find stupid mistakes, but it shouldn’t hold you for ransom over some & in an advertisment url.

Also, people tend to mix up xhtml vs html and css vs inline styling, but that’s a whole nother debate.

9  Didier Hilhorst  26 July 2004 11:29

Ok, so I exagerated a little. There are not masses of standardistas bitching (this is *the* new trendy word, mark my words.) Maybe the vendetta does not get personal as such, but we all know that some have a hard time putting validation in perspective. Besides, I still occasionally get an email explaining me what’s wrong (not a bad thing as such, but the tone is at times misplaced.)

Joe, I still think you ignore my most important point (atleast the most important to me personally): “my job does not involve coding on a daily basis.” I agree with you that if one claims to support standards, he or she must comply to those standards. Maybe I should not talk about this subject, maybe it’s inexistent or false. But if it is, why am I talking about it? My plan is not to create an atmosphere wherein a false debate is going on. Everyone is entitled their personal opinion on the matter.

We are all professionals (or atleast that’s the way I view Joe, others and myself in this debate.) If I were to work on a large web site project I would call Joe and hire him to share his expertise on accessibility (and if for some reason I can not, I still have his book.) The same goes for coding. Again, each adds his own expertise. Joe may point out some accessibility problems in my design (even before the coding stage, but most probably after.) I believe in hiring the best of the best with each their own professional expertise. I also know that is not always possible. Maybe I am less of a generalist than most, my background is different too.

Anyway, I think I should stop talking. I will let others discuss this matter with more sense. My final words are that I care about standards and do my best to both support them and comply with them. If I did not, we would not have this discussion.

10  Stefan Hayden  26 July 2004 14:44

I’m glad every time this topic is covered. I’d still debating my own position on standards and I need all the help I can get.

11  Tomas Jogin  26 July 2004 14:56

Didier: I think you missed some of Joe Clark’s point as well, which would be that regardless of what one’s job is, if one purports to be a standards advocate, one should comply to standards.

12  Didier Hilhorst  26 July 2004 15:18

See? I knew I should stop talking about this subject. It had to bite me in the ass. I agree with both Tomas and Joe, please do not get me wrong. I do not necessarily see myself as a standards advocate per se, but I do care, and would rather see standards implemented and used more widely. Does that make sense? Oh well, it does to me anyway.

Somehow I think validation comes for free. The same argument has been used for accessibility. Although some exogenous factors may screw up things, in theory it is possible to make any web site both comply with standards and perfectly accessible, regardless of size, design or scope. I guess education and willingness are the things that are missing.

Maybe I am indeed half wrong on this subject. But that is not a problem, please educate me. It has become clear to me that it is better to keep some seconds thoughts about validation to myself and just support a cause I think deserves my vote, as opposed to bitch about details. I still think validation is just a tool and that other factors may eventually be more important to end users. But a well balanced mix will always win, even though code is something users do not care about, that is as long as it works.

13  Arikawa  26 July 2004 16:04

“… even though code is something users do not care about, that is as long as it works.”

Generally, I agree with you Didier. However, I think that simplifying the ‘coding with standards’ and ‘validation’ debates to “as long as it works, the user doesn’t care” is a bit misguided. [ Maybe I am over-simplying your own comments. If so, I apologize in advance… ]

Part of the reason for standards (and for validation tools to check for adherence to those standards) is not for the end-user.

It is to provide browser manufacturers with guidelines to (mostly) adhere to, or at least, to strive toward.

Standards (and validation tools) are also to make the designer and developer’s life easier…

So I agree that the work involved should be transparent to the end-user, but I don’t agree that the end-user is the only one who benefits/matters.

14  Mike D.  26 July 2004 17:29

I personally prefer the term “validatorian”, as it implies someone who is really smart and graduated at the top of their class but no one can stand. :)

My problem with the standards debate is that I simply view standards differently than validatorians. They are guidelines to follow for safe web publishing, just as speed limits are guidelines to follow for safe driving. I try to follow them as closely as possible, but I don’t mind speeding a little here and there when necessary.

Some validatorians get on my case for saying things which smack of hyperbole, but that sort of hyperbole is really no different than Joe saying “If you’re not writing valid HTML, you are not making actual Web sites. You may be creating something else, and Web browsers may be able to read your document, but you’re not really engaged in Web development.”

I take Joe’s statement for what it is: hyperbole to get a point across. I get the point, and it is a good one. Valid code is important and you should strive to achieve it wherever possible. The only difference for me is where my efforts end and why they sometimes fall short. I am okay with introducing interactive elements into pages which may technically break validity. I am also okay with “progressive improvement”, which means that if you redesign a site and you can’t achieve 100% validity yet due to extenuating factors such as a CMS, third-party code, workers still learning, etc, then it’s perfectly okay to relaunch anyway and be vocal about your standards-support considering you are 90% closer to where you need to be. Maybe this isn’t a good excuse for a smallish site where one person has control of every single byte of the code, but for a major site, it’s certainly understandable.

15  wim  27 July 2004 11:56

Goed gedaan, jochie!

16  Fernando Dunn II  27 July 2004 20:04

A validator is a tool that determine’s whether or not you follow a certain standard, and what you need to fix to get there. If your site declares a certain standard and that standard’s requirements say “Thou shalt not forget to code the ampersand” yet you don’t comply, is your site really following that standard. Or is it following its own standard?

Sometimes you can’t have your cake and eat it too. But when you can, there is no good reason not to.

17  Josh  29 July 2004 05:43

I send my documents as application/xhtml+xml to browsers that can handle it.

If I screw up, Mozilla informs me in big, ugly, red-on-light-yellow text. *

Perhaps I’m insane, perhaps I’m a standardista (I’m putting that on a t-shirt :), perhaps it’s because I was dropped on my head as a child. But I’m emitting XML, and I’m going to keep my end of the bargain.

Glory pedantica.

* Mozilla only checks for wellformedness, not validity, but non-wellformedness is likely to be the problem.

18  Gabriel Mihalache  31 July 2004 13:20

Validation is such an issue because it’s the exception and not the norm. In the far-too-distant-future, people will author documents and content which will validate by default, because of the tools used. You won’t have to think about validating XHTML much in the same way you don’t think about validating a PNG’s headers/format.

That being said, and my sincere hope, much of the fault stands with browser authors, which made them tolerant to syntactic FUBARs. If you rip a few lines out of a .java file, chances are your compiler is going to complain and fail. IE will tolerate A LOT, so people got used to being sloppy and coding sloppy tools.

19  Ali Owen  31 July 2004 16:46

To a certain extent I agree with you Didier in that people should not get too wrapped up with validation, especially if it interferes with the design process. Instead the validators offered by the W3C should mainly be used as a guide (and often are quite useful for debugging bits of code).

Acheiving validation is a lot different from standards though. It is quite possible to use extraneous table tags to layout a site, or to use CSS incorrectly – but still have it validate. Validation is basically a notice to tell you that you can close xHTML tags and use the alt property (often with pointless values); therefore Validatorians are not the same as Standardistas. Standardistas are infuriated by IE not parsing the CSS Box Model correctly and use the full range of xHTML tags such as abbr and thead, we don’t just stop at validation.

20  Mike D.  01 August 2004 02:30

Ali,

Very interesting. Validatorians are not the same as standardistas… I like it! One thing I’m not clear on from your post though: can you be one without being the other? In other words, can you be a standardista without being a validatorian, and vice versa?

If so, I might actually consider myself a standardista — someone who cares about standards-based layouts and best practices but not to the point where I see validation as a must. Perhaps “rogue standardista” is a third category…

21  Ali Owen  01 August 2004 10:18

Well, I think that you can definitely be Validatorian without being a Standardista, as non-standard code can still validate which often makes some Validatorians believe that they are Standardistas. But yes I think it’s split for Standardistas who are also Validatorians, it’s always reasonably satisfying (well the first few times) for the validator to say “Your Page is Valid XHTML Strict” or whatever, but if it doesn’t validate it’s not always the end of the world.

A lot of times it can be best to leave your code not to validate if you are a Standardista. For example with embedding Flash, IMO it’s much more sensible to just leave the default Macromedia embedding rather than sataying it just to please the validator.

I think that we could have three categories: The Validatorian, The Standardatorian and the Standardista. Mind you, I’m not quite sure which category I fall into…yet.

22  Nathan K  03 August 2004 12:05

“Mind you, I’m not quite sure which category I fall into yet.”

This seems to be a very common aphorism, I must admit I am only just getting acquainted with XHTML and Validation, but I am doing it for two simple reasons and that is essentially to make my sites cleaner and more accessible (though I still have far to go). Whether validation truly makes for more accessible websites is still unbeknownst to me, however it does make me feel better about the end product one way or another.

23  Caleb  05 August 2004 23:03

I agree with you. Validation can be extremely difficult. I have pretty much decided to give up trying, especially with my weblog. Whenever I ad posts with images or pop up links without accesability options validation goes by by. Sure I could take the time to add the right code into every post so it would validate, but what’s the fun in having your own weblog if you have to go though the pain of making sure it validates. your personal weblog should be fun,. not a chore.

24  alex  09 August 2004 05:56

I’m with Tomas. Validating shouldn’t be a design technique, but getting an XHTML template to validate is 95% of the way to cross-browser bugfixing it - a no-brainer blast off the tee guaranteed to land me as close to the flag as possible before I putt in.

I don’t need to be anal with my code, because I don’t ‘think’ invalid code anymore. I’d actually have to concentrate to write bad markup. No big deal

25  Jina  11 August 2004 18:46

Just a question - am I correct in that XHTML transitions people to XML? I thought I had read that somewhere - I know I focus a lot on valid code - but honestly, it is to “check myself” - I want to learn XML, because I have read of the many wonderful things you can do with it. So I figured if I learned XHTML I would be one step closer. If I am way off-base, please tell me.