What's up with the W3C?

I'm all for standards compliance, but then again I'm not a designer. My reasons are pragmatic rather than ideological though.

I have worked with a few designers though and have managed to get them all seeing things the valid semantic standards based way in the end - even to the point where I think they are going overboard :)

Those two articles indicate to me that standards compliance isn't the problem - rather that the W3C is losing the plot and its new recommendations are going astray. It can either be fixed or replaced by something else IMO, but we still need standards.

But I do have some counter points for you....

The main thing that I think is just silly about this whole standardized code discussion is that it keeps changing... what is standard today is not acceptible tomorrow. So, it's really not very standardized for the long run.

...

I've been designing for about 10 years now. And one thing that everyone else in my shoes should agree with (or have noticed) is the fact that this stuff just keeps on changing. Standards based design is an evolving monster

I disagree. Things haven't effectively moved anywhere since 2001 and the release of IE 6.0 (lowest common denominator etc). The most current final release CSS standard is now 8 years old - that's half the entire lifetime of the web itself! The only reason this stuff seems like news now is the demise of NN4 (and more recently IE 5.x) and the fact that a majority of designers have finally caught on to what this stuff is, what the benefits are and how to use it. The standards aren't evolving - the industry is. It has been a slow process but the industry has finally come around to seeing things from a standards based point of view.

I'm still wary of using XHTML because IE doesn't support it correctly though - but you don't have to use XHTML. Older versions of HTML are still as valid now as they were 10yrs ago. Granted Drupal makes it hard to avoid XHTML though - probably best to not fight it :)

And what I understood about xhtml when it first came out was that it was just a "transitional" language to get everybody in the mode of coding pure xml in the future... closing tags etc. Just more strict. So, at some point we're supposed to ditch xhtml and just be doing xml, from what I understand.

XHTML is XML - there is no transition :)

XML isn't a markup language, it is for creating markup languages. HTML is a hypertext markup language created with SGML, and XHTML is a port of that language to XML (XML is a stricter clearer subset of SGML). CSS can be applied to any XML language, it isn't (X)HTML specific.

So, I was very interested to hear about a few articles recently that share some of the concerns I've just expressed. I think the idea or the gist of what they are saying is that... crossing every "T" and dotting every "i" is not (or should not) be the goal. Instead, we should be focusing on creating great websites, great user experiences, and sure, sites that are accessible to the largest audience if possible. But let's face it, there are some things that you can do with the design that are just not going to fit within these specs, whether it's because it uses a sliced up image in a table, or it's not accessible to dogs, or the code doesn't validate.

Note there is nothing in any standard that says that slicing up images and putting them in tables is 'invalid'. And as for canine accesibility, that isn't a problem - after all "noone knows you're a dog on the internet" ;)

I see compliance with accessibility standards as a different kettle of fish as complying with the coding standards. Accessibility is largely a subjective human focussed issue - one that can be handled by guidelines and common sense. It is a human that is the target and makes the final decision whether something is accesible to themselves or not (and those criteria vary from person to person). But with the coding it is software that is the target - and we all know how badly computers do with handling subjective input and making sense of it. Computers need precision when being told what to do, which is why markup should be validated.

The point of validation isn't as some sort of mutual back slapping exercise - it is the fundamental first step in avoiding problems. You wouldn't ignore syntax errors in a programming language just because a few popular implementations ignore them and try to guess what you were meaning anyway would you? If your markup isn't valid, you are relying on browsers to guess what your intentions are. It makes it impossible to tell whether a problem is your code or a browser bug.

Writing valid code isn't hard, and I have yet to come across anything that can't be made to work with valid code and requires breaking of standards. I've seen many designers come around to seeing it that way. And creating well marked up (ie using the right elements for the right job) and valid code makes future maintenance and extension so much easier as well as making it much more flexible in accommodating changes or being viewed on different devices or media etc. Not only that, but future browsers are much more likely to render it properly - it is more futureproof.

Sometimes with all the new hype about an 8 year old standard, it might seem that people are pushing it on religious grounds. But it would be a mistake to dismiss this approach because of that. There are very good pragmatic reasons to prefer doing stuff this way.

I see it as being like novice programmers growing out of their initial spaghetti coding habits. Some of them might complain at first about learning to do things properly - after all their spaghetti code still produces the correct result. But doing it properly makes future bugs much easier to fix, as well as making it much easier for future people to pick up and add new stuff to. I reckon being a good web designer is more than just being an artist - it is also about being a craftsman (or women). You are building more than just something to look at, it has to work and be robust under unforeseen circumstances.

Thanks for the discussion. Anyway, I've probably spent far too long typing this... :)

--
Cheers
Anton

I didn't think that this had enough Drupal content to stay in the listing. Sorry, just trying to follow some of our own guidelines.