What's better for search engines: xhtml/css or traditional html?
 

Opinions please.

xhtml/css of course. less html clutter, the spider can find the content easier.

Good point.

xhtml/css is way cleaner code so you get higher word to code density

But do the SE's care about word/code density? It doesn't seem unlikely that they'd try and stop design style from having influence on rankings, because design doesn't necessarily say something about relevance of the content.

the xhtml/css route is better unless you can code real clean HTML.

Yes I agree. Use css to style your pages instead of html. It's been the standard for quite some time, besides it's so much more capable once you learn a little.

CSS is industry standard. Porn sites are just retarded. This shouldn't even be a question.

P.S. I've read somewhere that the quality of the coding influence on the rankings ... not sure thought

Huh. Farewell to tables then, from now on it's divs :(

Css takes a bit to get used to, but its worth it. I've actually just converted from crazy html with lots of graphics to css. But i haven't really noticed any change to be honest. Anyone switch and notice a difference or does it take a while?

tables/nested tables are crap for more reasons than seo

10 times smaller or more using x/css
Bots can get through it easier than going through a load of badly coded nested tables that are probably invalid
Text density (relevance isnt massive but still part of the principle)
Easier to hand code
Far far faster to update and update cross sites not by editing attributes

That would make common sense, so I would tend to believe it with no further proof.
I think all thats needed to remember is that SE's want relevant results displayed for the search term entered. The more people who can reach your page (more widely portable/readable the page), the better the page by default.

So yeah, run your code through a validatormatrixolator.

and it's extremely quick + easy to give your site a "holiday" theme change with css.

What makes you think crawlers go looking for tables or any other code for that matter? if I were a crawler I would go looking for <div>contents</div> <p>contents</p> etc. I highly doubt they take into account whether you have nested nested nested tables or not.

A html document is going to look more like a big ascii soup to a crawler, tags don't mean anything code wise.. you could probably nest a <a href> in a heap of dody / unclosed tags and they would still pick it up. They just know particular (simple) tags may contain text.. and some may be more relevant than others.. <h1> <h2> etc

One thing that could possibly matter is whether your document is valid xhtml / html on the whole.. which applies to both anyway.. it is easy to put an invalid parameter / tag into either, so that isn't really relevant to this discussion.

I always validate my code, as well as check that it works on all platforms and in all (important) browsers, so that's not a problem. Everybody should do that, always.

My personal standards are ... IE5 + and Firefox in 800*600.

I never said they look for code or tables, for that matter. On the contrary they dont want to see a load of shit. It is far more logical to parse nice light code and pick up the relevent text and possibly give extra consideration to tags that do make a bit of difference than. Nobody said any css/xhtml tags themselves bring benefits to your page they simply facilitate a superior style of coding.

I like the xhtml, more extensible