Webmaster forum - A good reason to validate your HTML

Author

A good reason to validate your HTML

Scott Bryce

2006-09-24, 6:58 pm

Last night I uploaded a bunch of updates to my site. This morning income
is WAY down. I get 5 emails telling me that people can't sign up.
Everything looks OK to me. But the HTML wasn't valid (I'm working on it
now) and people using IE didn't see the submit button. I lost a bunch
of money becaue I didn't take the time to validate my HTML.

Brian Wakem

2006-09-24, 6:58 pm

Scott Bryce wrote:

> Last night I uploaded a bunch of updates to my site. This morning income
> is WAY down. I get 5 emails telling me that people can't sign up.
> Everything looks OK to me. But the HTML wasn't valid (I'm working on it
> now) and people using IE didn't see the submit button. I lost a bunch
> of money becaue I didn't take the time to validate my HTML.

No. You lost a load of money because you didn't test it in multiple
browsers (and not testing in IE a sin).

Also, validating your code is by no means a guarantee it will work in all
browsers.

--
Brian Wakem
Email: http://homepage.ntlworld.com/b.wakem/myemail.png

Scott Bryce

2006-09-24, 6:58 pm

Brian Wakem wrote:

> No. You lost a load of money because you didn't test it in multiple
> browsers.

OK, that too.

> Also, validating your code is by no means a guarantee it will work in
> all browsers.

I understand that, but in this case when I fixed the HTML errors, it
worked in IE. Had I validated, it would have worked in IE whether I
tested in IE or not. Testing in IE would have shown me that I should
have validated. Validating showed me what needed to be fixed.

Matt Probert

2006-09-24, 6:58 pm

On Tue, 19 Sep 2006 13:42:43 -0600, Scott Bryce
<sbryce@scottbryce.com> wrote:

>Last night I uploaded a bunch of updates to my site. This morning income
>is WAY down. I get 5 emails telling me that people can't sign up.
>Everything looks OK to me. But the HTML wasn't valid (I'm working on it
>now) and people using IE didn't see the submit button. I lost a bunch
>of money becaue I didn't take the time to validate my HTML.

Do you mean validate or debug?

There *is* a difference. Failing to close tags can cause problems!
Forgetting an alt text in an image tag is very unlikely to bring a
site to its knees!

And we shant mention http://scottbryce.com <g>

Matt

--
Veritas Vincti
http://www.probertencyclopaedia.com

Scott Bryce

2006-09-24, 6:58 pm

Matt Probert wrote:

> Do you mean validate or debug?

I didn't run it through a validator. Since HTML isn't a language, I
don't think of it as debugging. But there were errors. Lots of them.

> There *is* a difference. Failing to close tags can cause problems!

Bingo!

> Forgetting an alt text in an image tag is very unlikely to bring a
> site to its knees!
>
> And we shant mention http://scottbryce.com <g>

No, we shan't. I did that a LONG time ago.

Charles Sweeney

2006-09-24, 6:58 pm

Scott Bryce wrote

> Brian Wakem wrote:
>
>
> OK, that too.
>
>
> I understand that, but in this case when I fixed the HTML errors, it
> worked in IE. Had I validated, it would have worked in IE whether I
> tested in IE or not. Testing in IE would have shown me that I should
> have validated. Validating showed me what needed to be fixed.

Yep, a simple HTML-parse check would have done the same thing.

http://www.doctor-html.com/RxHTML/

--
Charles Sweeney
http://CharlesSweeney.com

David Dorward

2006-09-24, 6:58 pm

Scott Bryce wrote:

> I didn't run it through a validator. Since HTML isn't a language

HTML is a language, not a programming language, but still a language.

--
David Dorward <http://blog.dorward.me.uk/> <http://dorward.me.uk/>
Home is where the ~/.bashrc is

David Dorward

2006-09-24, 6:58 pm

Matt Probert wrote:

> Do you mean validate or debug?
>
> There *is* a difference. Failing to close tags can cause problems!

And validation will pick that up

> Forgetting an alt text in an image tag is very unlikely to bring a
> site to its knees!

But could fail to convey essential information to someone who needs it, thus
resulting in a lost sale.

--
David Dorward <http://blog.dorward.me.uk/> <http://dorward.me.uk/>
Home is where the ~/.bashrc is

Nikita the Spider

2006-09-24, 6:58 pm

In article <Xns9843E5A5C28A9mecharlessweeneycom@130.133.1.4>,
Charles Sweeney <me@charlessweeney.com> wrote:

> Scott Bryce wrote
>
>
> Yep, a simple HTML-parse check would have done the same thing.
>
> http://www.doctor-html.com/RxHTML/

No offense, Charles, as I realize you were trying to be helpful, but I
think people need to be very wary of syntax checkers and not mistake
them for proper validation. Syntax checkers are capable of spotting
unclosed tags and the like but unless they're true SGML validators then
they'll come up with some false positives and miss some real problems.

Real world example: I'm currently helping a group that's fond of Cold
Fusion Studio as their editor, and apparently it has a built-in syntax
checker. Sadly, it doesn't know that 'single quotes' are just as valid
as "double quotes" around HTML attribute values. They were checking some
of my HTML and thought they'd found a problem because CF Studio
complained about the single quotes. What's more frustrating for me is
that they also checked the file against the W3C Validator which
pronounced it valid HTML 4.01 Strict, but they didn't know whether to
believe the W3C Validator or CF Studio. The W3C Validator isn't perfect,
but it isn't often wrong either.

To be fair, syntax checkers can look at things which a validator
ignores, like whether or not an attribute contains a valid data type (an
integer versus text, for instance). And the W3C validator's messages are
technically correct but often opaque; there's a bit of an art to
interpreting them. Nevertheless, I'd much rather spend my time
interpreting a validator's opaque messages than getting clear-but-wrong
messages from a syntax checker. To put it another way, if I'm going to
take the time to programmatically check my HTML, a syntax checker isn't
the tool I'd use.

As for Doctor-HTML:
- Their pages don't validate and they don't even contain a <!DOCTYPE>
declaration.
- When I go to the single-page analysis page
(http://www.doctor-html.com/RxHTMLpro/cgi-bin/single.cgi) and click on
the advanced options, I get a popup that says "The program is unable to
load /home/drhtml/www/RxHTMLpro/html/RxHTMLpro/advanced_template.html."
It was this way when I checked it about a year ago, too.
- In my browser (Firefox) that same page has a huge blank space below
the main text and has four unmarked radio buttons at the bottom of the
page.

Nobody's perfect, and I'm sure the folks behind this tool have good
intentions. But their own site could use some help which doesn't speak
well for the strength of their tools. Anyone wanting to check his or her
HTML would be better off (IMO) using the W3C Validator, the one at
htmlhelp.com (up to 100 pages free) or my own which can validate an
entire site and do link checking as well.

Disclaimer: my own validator is right now 100% free but I'll eventually
charge money for it.

Cheers

--
Philip
http://NikitaTheSpider.com/
Whole-site HTML validation, link checking and more

Toby Inkster

2006-09-24, 6:58 pm

Brian Wakem wrote:

> Also, validating your code is by no means a guarantee it will work in all
> browsers.

But it's a guarantee that it **should** work in all browsers. (i.e. that
if the browser gets it wrong, its the browser's fault!)

Of course, that doesn't help the bottom line, so browser testing is still
neccessary -- but validation will quickly help spot simple errors like the
one mentioned.

--
Toby A Inkster BSc (Hons) ARCS
Contact Me ~ http://tobyinkster.co.uk/contact

Toby Inkster

2006-09-24, 6:58 pm

Nikita the Spider wrote:

> The W3C Validator isn't perfect, but it isn't often wrong either.

If talking about their HTML validator, then it is actually pretty close to
perfect.

OTOH, if talking about their CSS "validator", I lost faith in it a long
time ago. I haven't yet found a CSS syntax checker that gives
useful/reliable results. :-(

--
Toby A Inkster BSc (Hons) ARCS
Contact Me ~ http://tobyinkster.co.uk/contact

Nikita the Spider

2006-09-24, 6:59 pm

In article <kla6u3-ese.ln1@ophelia.g5n.co.uk>,
Toby Inkster <usenet200608@tobyinkster.co.uk> wrote:

> Nikita the Spider wrote:
>
>
> If talking about their HTML validator, then it is actually pretty close to
> perfect.

Agreed, but I was afraid that stating that too strongly would invite a
litany of imperfectly-handled edge cases that most users are unlikely to
encounter.

> OTOH, if talking about their CSS "validator", I lost faith in it a long
> time ago.

I haven't used the CSS validator myself, but nothing I've heard
encourages me to try it.

> I haven't yet found a CSS syntax checker that gives
> useful/reliable results. :-(

Sounds like an Open Source project waiting to happen!

Cheers

--
Philip
http://NikitaTheSpider.com/
Whole-site HTML validation, link checking and more

William Tasso

2006-09-24, 6:59 pm

Fleeing from the madness of the jungle
Toby Inkster <usenet200608@tobyinkster.co.uk> stumbled into
news:alt.www.webmaster
and said:

> Brian Wakem wrote:
>
>
> But it's a guarantee that it **should** work in all browsers. (i.e. that
> if the browser gets it wrong, its the browser's fault!)

/applause

> Of course, that doesn't help the bottom line, so browser testing is still
> neccessary

for markup? rarely IME

For design/layout & CSS - yes, browser testing is required

--
William Tasso

http://williamtasso.com/words/what-is-usenet.asp

GreyWyvern

2006-09-24, 6:59 pm

And lo, Nikita the Spider didst speak in alt.www.webmaster:

> Toby Inkster <usenet200608@tobyinkster.co.uk> wrote:
>
>
> Sounds like an Open Source project waiting to happen!

Who's up for it? ;)

Grey

--
The technical axiom that nothing is impossible sinisterly implies the
pitfall corollary that nothing is ridiculous.
- http://www.greywyvern.com/orca#search - Orca Search: Full-featured
spider and site-search engine

BBM

2006-09-24, 6:59 pm

Scott Bryce wrote:
> Matt Probert wrote:
>
>
> I didn't run it through a validator. Since HTML isn't a language, I
> don't think of it as debugging. But there were errors. Lots of them.
You mean that the HyperText Markup *Language* is not a language?
Is English a language? ~_^

Charles Sweeney

2006-09-24, 6:59 pm

Nikita the Spider wrote

> In article <Xns9843E5A5C28A9mecharlessweeneycom@130.133.1.4>,
> Charles Sweeney <me@charlessweeney.com> wrote:
>
> No offense, Charles, as I realize you were trying to be helpful, but I
> think people need to be very wary of syntax checkers and not mistake
> them for proper validation.

Howdy. You missed my point. It's about syntax checking VERSUS
"validation".

A syntax error is an error. A "validation error" can be (and usually is)
an instance of not adopting a recommendation.

--
Charles Sweeney
http://CharlesSweeney.com