Does it ever give a clean bill of health?
I know this question is going to be hamstrung by the inability to post URLs in this forum (something that never ceases to amaze me), but I am getting real curious. Even the most straighforward and robust HTML I produce fails the W3C validator in every possible way. And even when I find some immaculate examples of new-gen coding - such as the pages of Automatic Labs (I am in no way affiliated with these people - I was just impressed with their hardcore standards based approach to pagesmithing), the W3C validator dumps all over them like a donkey with dysentry.
Can anyone point me to some pages (cryptic keywords to google will be fine and shouldn't upset the moderators) that actually get a clean bill of health? I just want to see if it ever happens, and what this immaculate code looks like.
Lots of pages validate - it's fairly easy if you hand code. I've never used HTML generators, but from what I understand their output can be tough to validate. So, as an example, my profile site validates on most pages (though occasionally an error slips through).
One way to look for examples is to search the Google image database for small "valid HTML" [images.google.com] icons. That will show you lots of pages displaying the W3C "Valid HTML" logo, and most of them will validate (just click the icon).
Nice - thanks.
It also occured to me to try the W3C pages themselves - and indeed they validate.
And yes, I use Dreamweaver lots.
There is a deep sense of satisfaction the first time one gets a page to validate.
Handcoding is the answer. It gets easier with practice. Last one I ran through the validator had 3 errors, which was a snap to clean up
Keep at it.
Almost every page I put out validates zero error when I deliver it to the client, if I maintain it it stays that way, if I don't, it doesn't. XHTML 1 strict, transitional, html 4.0 strict or transitional, it's pretty much as easy to do any of these as any other once you learn the rules for each, but 4 transitional is the easiest. The problem is dreamweaver, frontpage, etc, handcoding is the best way to learn how create error free html since you are creating the html, dreamweaver can be made to do this I've read but not tested, you need an extension 4eyes always recommends, can't remember its name.
Also of course any page that calls itself xhtml and is not delivered with xhtml mimetype is in fact an error filled page even though the validator says its valid, but that's another issue.
Usually most validation issues come down to fixing a very small number of errors, closing tags, alt=, and so on, and getting rid of the illegal characters that jump in if you copy and paste from word documents. I run all my client word documents through a ms word macro that replaces all illegal characters with real html character entities like —, “ and so on.
It takes a few days to get your first site error free, you just start at the first error on the validator, fix every occurance on the site using search and replace etc, revalidate, fix again, you'll find the 100 errors quickly vanish since many of them aren't actually errors once the first error is fixed.
Once you have your site error free you get the real payoff, which is the ability to actually be able to do code debugging using the validators, that can be a huge time saver on complex html/css.
With the newest version of DW you should be able to produce validating pages.
Some errors the validator generates come from errors earlier in the document. So it does make sense to correct the first error in the document first and see what happens, and so on.
I do handcode too, so I have no clue what needs to be done in DW to produce validating pages. Make sure you're using the right doctypes. There are plenty of articles around, googling for "doctype" or "doctypes" will help.