Invalid code does not conform to standards — that’s the way it is, wise or not.

“Not mention standards” is very different from saying that invalid code conforms to standards.
It does not matter if your blog is 1% invalid — it is still better than 1000 valid ones, including this. And you should absolutely talk about standards — you have best web-oriented blog and thousands are listening (reading, I mean).

And that’s why you should be careful with what you say. You say “It is invalid now but I am working on it” — it is OK. But when someone says “It is not valid, and will never be, so f*** the W3C” — that’s bad. It makes developers comfortable with a thought that invalid code is OK. “Keith said it, Mike said it, … said it” — so let’s move on.
There is the limit of invalidity? How bad code should be when you will say, “no, that’s not acceptable”? Are majority of the web is based on web standards now, cause their code looks much like HTML?
How are we going to have CMS that produce valid code if nobody cares?

Do you agree, that getting valid code is the easiest part of developing the site (I am not talking about maintenance here)? Design, IA or usability is much more difficult to get done right.

And not every site will be a blog with comments system. So if you done it right from the start you shouldn’t have much problems later.

If one leaves code cleaning for the better times, that are the chances those times will ever come?

Call me whatever you want. Calling names will not change anything. It will not improve anything. And we are on the same side anyway. Only you choose to emphasize what user sees and interacts, I am trying to say that code is important too.

Or is it? Is Anne right: there is no point?

Is that wise? I mean my site is 99% valid. I’ve got a problem with my comments, which are essential to my site and I’ve actually spent quite a bit of time trying to sort that out. It’s not possible without sacrificing functionality. Period. So I’ve got to choose between validity and functionality. I chose functionality.

Now, back to your favor. Should I now not mention standards on my site because I have a few valid pages? Fine by me, but I don’t see how this HELPS the cause of Web standards. If anything by alientating folks like me and Mike (and many, many more) all you’re doing is hurting the “cause.”

As of now, I’ve removed my validation links. I’ve got absolutley no problem not supporting Web standards. I’ll leave that to the Zealots.

do not say that invalid code conforms to standards

Exactly! You made a much better job of summing this up than I managed to do. I’m curious, though, about your statement that invalid code often is better than valid code. If you had said ’sometimes’ I might have nodded, but ‘often’?
My main focus is accessibility. With invalid code, although it may work very well in most user agents, you can never guarantee that it will work with assistive technologies. A left-out </table> tag may cause the DOM tree to be broken, and most screen readers use the DOM tree to access the content of a web document.