Skip to: Navigation | Content | Sidebar | Footer

Weblog Entry

Validation Woes

July 15, 2003
This page is not Valid XHTML 1.0 Transitional!

Designers transitioning to XHTML often get tripped up by the cryptic errors when validating. The W3C’s validator has long been lamented as being un–user–friendly.

Take the source of this very page. When everything is kosher, my XHTML 1.0 Transitional gets a thumbs up. But if I make a small, very understandable error like missing a </p> for example, look what happens. I missed one closing tag, and received 52 validation errors.

The enthusiasm of a beginning web designer getting her feet wet with XHTML rapidly wanes when this happens to her. Seeing so many errors on one page is a blow to one’s psyche, and when they don’t realize that all 52 can be eliminated with the addition of four bytes to their code, it can make the difference between sticking it out until validation is achieved, and turning their back on valid markup because it’s just too hard.

update: Zeldman has pointed out that the validation services are maintained by a small, unpaid and under–appreciated staff of volunteers who are looking for help to improve the service. It’s not fair of me to point out the shortcomings of a group that is doing so much with so little, especially out of my own ignorance, so I’ve toned down the original post. Hopefully my criticism comes across as more constructive than otherwise…

If you have the time and technical expertise necessary, this would be a great project to get involved with. My personal gratitude to those working on this necessary tool.

Reader Comments

MikeyC says:
July 15, 01h

“I can’t validate that page. I have to actually go to the page in a browser, save the source as a file and then validate that file.”

Why would you have to do that? I am not following…what do server-side tools that assemble your page (like php) have to do with preventing you from validating the end-result page?

I just checked your page against the validator and there are some very real errors. So I’m not sure how you can claim that your site is valid by first saving it locally and then validating?

Jim Dabell says:
July 15, 01h


I think he’s talking about the results of POST operations etc.

scH says:
July 15, 01h

MikeyC, no I in no way think that my current site is valid. I’m working on a much more valid site at

I went to the validator at and it actually validated my site this time. Before I’ve always gotten errors that looked like they were related to not being able to pull data in my results. Maybe I’m totally wrong and it’s just another example of one simple error causing many error messages, like Dave is talking about.

zeldman says:
July 15, 09h

“Better Living Through XHTML” (ALA, 2002) discusses these common validation service problems, which are partly due to the limits of software (software can’t parse your intentions), but in some cases could be made less frustrating via clearer, simpler, more coherent error messages.

Have spoken with some of the folks who build these tools as a labor of love. They are all volunteers and have no budget. They are aware of the language problem and would like to fix it but face certain technical obstacles. They’ve asked for help from the community:

Until they get that help, it’s true that poor experiences with the validation services could frustrate designers and developers who want to do the right thing.

scH says:
July 15, 09h

Another issue I have with validators is that if I have content pulled in via PHP and mySQL, which I almost always do. I can’t validate that page. I have to actually go to the page in a browser, save the source as a file and then validate that file. While not that big of a deal, and one that I know the cause of, it makes it hard for me to, in good conscience, put up a “valid xhtml” banner when anyone simply validating the page will find out that it isn’t (technically).

I’m fairly new to design, and very new to caring whether my sites are valid or not, does anyone have a work around for this issue besides what I’ve already stated?

Jim Dabell says:
July 15, 10h

I’ll take this opportunity to point out that one of the people working on these tools, Nick Kew, also sells an accessibility checker which, in my (admittedly limited) experience is far better than Bobby or that other one that I can’t even remember the name of (Miss Manners?).

You can trial it or buy it (hint) from the Accessibility Valet website.

July 15, 11h

scH: Once you load the page in the browser, you can use a favelet to validate your HTML and/or CSS. Combine Internet Explorer’s Toolbar Favorites feature and mini applications (applets) written in jscript, stamped with a “javascript:” URI scheme, and you have Favelets - a way to add features built from DHTML to your browser.

Jim Dabell says:
July 15, 11h


All the validating favelets, to my knowledge, operate by looking at the URL and sending that to the validator, not by uploading a copy of the HTML.

July 16, 09h

Man, that has to be my pet peave. You forget to close or accidently delete a tag and wam you site doesn’t validate. Recently I drove myself nuts when I deleted a line of my CSS file by mistake. It caused the half of it to not work. The blessing was once I tried to validate it, I found the obvious error. Validators aren’t perfect, but they sure are useful!

July 16, 12h

All the validating favelets, to my knowledge, operate by looking at the URL and sending that to the validator, not by uploading a copy of the HTML.

When you go to the URL, the server renders the PHP page as HTML, and when you click the favelet, it validates the HTML of the rendered page. Once you’ve gone to the page, it’s a single HTML page, and that’s what the validators validate.