13 replies to this topic

[thetraveler]

My site has 21 errors on the validator. Can there be legitimate reasons why my site designer would make code that would appear invalid to the W3C software? Or, should all errors be fixed, no exceptions?

Thank You!

[1dmf]

Well, hmm, it depends on which angle you come from....

As a coder and beleiver in quality, ethical , professional work then there is no excuse whatsoever for invalid code (and that's without then talking semantic code!).

For SEO purposes, the majority here will tell you it's irrelivent and does not help or hinder indexing or ranking, though I would argue some of those points, missing title tag, no <h#> tags or other invalid stuff surely would affect ranking, but I know from previous discussion, this argument can spiral with many on both sides of the fence so let's not go there on this thread.

Personally , why do you need invalid code, if you really must for example use the target attribute on an anchor, then either use JavaScript to change it with the onclick event (that way desired result and validates) OR , use a different doc type, don't tell the browser you're using X/HTML 1.0 Strict and then not code using the rules for that standard!

Change it to transitional , or if you can't be assed to end ALL tags as the XML specs require, use HTML 4.01 , if you need frames use the framset doc type, etc. etc..

Curiosity makes me ask what are the errors, could you give a URL or post results, it could be something simple as not using the correct escape codes for an ampersand (&) = &amp;

[Randy]

Errors or warnings? And do the errors actually cause things not to work as expected? Does the page display the way it should across multiple browser and OS combinations?

If things work as expected what's to worry about? We can all have our own opinions, often depending upon just how much of a perfectionist bent we have. But do you wanna know just how much stock Google puts in pages validating?

Even for a very simple page construct like Google's home page, before there are any search results in it, they have 59 Errors as things stand today. It's been even worse.

[thetraveler]

Thank you both very much for your input.

It says errors and warnings. Although Google certainly is part of the question, the other part is just wanting to make sure the site works on everyone's computer.



Here is a sampling if that helps.

Validation Output: 21 Errors

1. Warning Line 10, Column 78: NET-enabling start-tag requires SHORTTAG YES.

…Tjwxe4SYrNQtynOmFSoBrpxI6/08fBVGItfs=" />

✉


2. Error Line 11, Column 51: document type does not allow element "META" here.

<meta http-equiv="Content-Language" content="en-us">
✉



3. Error Line 12, Column 66: document type does not allow element "META" here.

…v="Content-Type" content="text/html; charset=utf-8">
✉


4. Error Line 13, Column 55: document type does not allow element "LINK" here.

<link rel="stylesheet" href="style.css" type="text/css">
✉

5. Error Line 15, Column 6: end tag for element "HEAD" which is not open.

[1dmf]

lol - that's always made me chuckle , G! has always had errors, even when it was nothing than an image and input field and a submit button!

I begin to wonder if it's deliberate ?

At the end of the day isn't it always better to advise to validate and correct if at all possible, if I had a mechanic work on my car and make a right pigs ear of it, would him retorting "It goes doesn't it" suffice?

Or what about when you get it back and it's got oily hand prints everywhere, the engine might run like a dream and they have done a 1st class job, it still sucks getting it back covered in oil from carelessness!

But enough of my insightful anicdotes, you're right Randy, it is funny

[Randy]

Those all look to be errors/warnings related to the declared doctype ambtmt. Meaning changing the page's document type to support the code actually being used would probably cure most if not all of them. Dunno what doctype is being declared, but I often see such warnings/errors when someone is declaring a Strict doctype, when they should really be using a Loose or Transitional doctype.

I rather doubt any of those are going to cause either the search engines or real users any grief.

[Randy]

At the end of the day isn't it always better to advise to validate and correct if at all possible

Yes, of course. Especially if one has to concern themselves with meeting the 508 Accessibility standards.

What's necessary for users and what's necessary for bots are two entirely different things. However it's never a bad thing to at least attempt to fix validation errors. Even if the errors or fixing them isn't going to have any net effect.

Now, that said I'll be the first to admit the pot shouldn't be calling the kettle black. My own blog home page has had in excess of 100 validation errors/warnings for years. And I've yet to get around to looking into them or fixing them.

[1dmf]

agreed Randy, it could be doc-type issue.

But it also might be genuine typo errors, as in they have just forgotten the forward slash tag ending off tags that have no actual end tag such as <input /> , <img /> etc.. (if using XHTML doc type)

Also if in XHTML , you MUST use lower case for all tags and attributes, there is no <HEAD> ,it's <head> so no </HEAD> it's </head>.

old HTML right up to 4.01 didn't care on upper/lower mish mash of syntax , but as X/HTML is based on the XML syntax, case becomes paramount! , also makes your code easier to read!

[Nueromancer]

My site has 21 errors on the validator. Can there be legitimate reasons why my site designer would make code that would appear invalid to the W3C software? Or, should all errors be fixed, no exceptions?

Thank You!

depends what your validateing against often if ime writing html by hand i dont bother with a doc type and becuse of my history in html (1994) i just tend to write some what old school html - abeit sticking to the xhtml stndard of lowercase tags.

it depends on where you want to put the effort being anal about validation or doing a professional cost benefit analysis and puting the effort in where it will have an effect.

I used to work in OSI networking which which had a lot more standards that this kiddy TCP/IP stuff and stndards organisations are riddled with faction fighting - and you do get things like ICL not being able to count (counter MUST start from one does not mee start from 0) or Sprint saying yes we know its not to X.400 1984 cant you just chnage the standard to make it work sprints way!!!! Sigh******

[1dmf]

lol - i'm having a mare with MS Office Outlook not being compatible or able to import it's own files, so if MS can't work with itself what chance has anyone else!

But when you say

it depends on where you want to put the effort being anal about validation or doing a professional cost benefit analysis and puting the effort in where it will have an effect.

What cost? every page I write I write to X/HTML 1.0 strict, doesn't mean I get it right evertime, but it takes seconds to run it through the validator and correct the stupid errors or typos, it cost nothing, and using the excuse my page doesn't validate because I can't afford to correct it, well sorry that's hogwash. It takes seconds!

now, if it's also legacy code written in the old days by sloppy coders, well, there lies your problem, not the anal validators! , but the assholes who wrote it in the first place!

Sometimes it's quicker, cheaper and easier to re-code the page, than trying to rectify someones tag soup!

[Randy]

That's me. The a hole who wrote the non-compliant code, even by legacy standards in the first place.

[Ron Carnell]

You know, to the best of my knowledge, I've never had a validator click an ad or buy one of my products?

I'm only interested in pages that work. Back when I was still learning how to make pages that worked, a validator might have been a useful tool. It could have helped shorten the learning curve. It's been at least three years, however, since last I validated a page. And that was only because it was for someone else. My pages are already perfect.

[1dmf]

If they have errors then they aren't perfect, well unless you have poor standards of course.

But that said I used to be one of them a holes too, till I found www.tek-tips.com and learned better!

Making money from your web page doesn't make it a good page or a perfect page or a valid page, the chinese made a bucket load of money from selling weapons to Sudan which enabled the genaside in Darfur, I guess they have a perfect arms trade then!! regardless of the treaties (standards) they breach!

Ethical SEO my a hole!

Edited by 1dmf, 18 July 2008 - 08:25 PM.

[ArtphotoasiA]

My 2 cents......

search engine should strongly support W3C and start to consider much more in SERP the value of a W3C website... personally I make all my web following W3C