Validation Woes

grumpy old grandpa

[grandpa]

I just assumed the duties of maintaining a friends commercial web site - he runs a home based business. So I got busy and fixed a lot of "little" things to improve the appearance of his pages. Now I realize that not one of these pages will pass any validation test - heck there's not even a doctype declaration to be found.

My questions to this are:
1) just how important is doctype and validation to the spiders and bots that we want to impress?

2) is there a really old doctype I can declare - or do I have to get back up to speed on my html?

HTML was a hobby of mine 15 yrs ago. I understand the basics, and those pages look pretty good to me. But I want to "get it right" if I can... friends are good thing to have.

[Birdman]

I think validation is getting more and more important, but it probably will not hurt the site too much. I still suggest you get your doctypes and such in place. Make sure the <title>s are relevant to the pages content, as well...for search engine's sake.

Check out w3c.org for all the doctypes...you can go pretty far back!

[mattur]

The most important thing is that browsers and bots can access your content. For this, there's no substitute for testing. Validation may help with finding markup errors, but it certainly isn't required for an effective site.

Current browsers and bots are designed to handle non-valid html. As long as it is structured properly eg no nesting errors, the absence of a dtd is not a problem. HTH.

[g1smd]

Run the code through [validator.w3.org...] and fix as many errors as possible. You never know if one of those is important to a spider or to a particular brand of browser.

[grandpa]

I appreciate the response. Yesterday I started looking into CSS as a solution... oh my! Still, I suppose if I want to claim a victory here I'll do whatever it takes. I have already started fixing what I can with the validation feedback. One step at a time.

[Mohamed_E]

> One step at a time.

I completely agree.

When I first started validating I was willing to accept a lot of "invalid" code which worked but violated the standards in one way or the other. Over time I became stricter, my code now validates completely to 4.01 Transitional and almost to strict.

[RammsteinNicCage]

**deposits $0.02**

After inserting a doctype and validating my site, my search engine position on google jumped about 30 spots. I don't remember making any other changes at the time, so I can only attribute the better placement to the validating.

Jennifer

[grandpa]

I'm coming back to thread just to say that I have begun validating those pages. I tinkered a bit with external CSS, but right now I'm simply using STYLE to accomplish my validation. To date, maybe 15% of my site is now valid.

Other problems are popping up as I validate these pages, mostly nesting errors and missing tags, so at least for me validation is a positive step.

Doing all this also gives me a chance to look at these pages with an eye to total site redesign, something I've approached in another thread, and something I'll do on paper first - not online.

The responses here have been beneficial to me as a newbie administrator, Thanks!

[g1smd]

I fix all instances of nesting errors, unclosed tags, tags closed in the wrong order, block inside inline, tag typos (like <talbe> or </p and so on).

I occasionally leave in the odd IE-specific piece of code like body bgproperties="fixed" to come back to at some later time.

I am gradually moving to CSS. You really do get lean HTML code when all the font, i, b, u, etc, tags are gone. Make sure you use headings <hx> and paragraphs <p> as they were intended. That can help too.

[jbinbpt]

I found that making sure the code is valid, cleaned up a lot of what I thought were cross browsers issues.
As long as you are in there tweaking the code, it is a good time to rethink your Meta tags. Just make sure that they make sense. I know that the current thinking is that meta tags are a waste of time, but who knows what the future holds. They help me keep organized.
jb