Re: Idea for the validators

On Sun, 27 Jul 2003, James Nash wrote:

> I'm a web-designer and I make a point of sticking to the standards when
> I make pages.

Good, provided that "standards" are put into proper perspective. The W3C
recommendations are useful, but by no means sufficient for creating
good pages, and sometimes one even needs to deviate from them in order to
make pages work well across browsers.

> To show that I have I usually include your "Valid HTML
> (or whatever)" buttons on my pages as links to your validation service.

Why? If there were an online spelling checker, would you also link to it?
If yes, why?

> I was recently showing one of my pages to a friend and he didn't know
> what the "valid html" meant

That's very common. Actually, many Web surfers do not even know what
"html" means, and don't need to know (for several values of "need").

> so he clicked on it and got the validation results.

So the link name or symbol was misleading, was it not? But more
importantly, the link itself was worse than useless, for the purposes of
your page (unless, perhaps, the page itself is specifically about Web

> However, not being a web-designer himself and not knowing what
> html was or why it was important he was quite confused by the page.

So would he need to know anything about it? Why? If he intends to start
Web authoring some day, it is _surely_ better to start from something
completely different than validation.

> This gave me an idea: Why don't you offer an alternative, in-layman's
> terms results page that people like me can link to instead?

My idea is that authors should stop putting those icons on their pages,
and the W2C should stop encourageing authors to use them. (This includes
CSS, WAI, and other icons.) They might have a useful purpose _at most_
on pages that specifically discuss the design of a site, i.e. on those
"About this site" page. Even there, words and sentences say much more than
obscure abbreviations.

It is _very_ difficult to explain validation in layman's terms, if you set
the boundary condition that the explanation should be closer to factual
truth than to nonsense. I have once tried to do that (naturally for
"laymen" _among Web authors_):

> (i.e.: don't talk in terms of resources and xhtml transitional and code
> snippits)

That would be both impossible and pointless. Validation is _all about_
such things. It is difficult to explain general relativity in layman's
terms, but it would get pretty absurd if all references to physical
reality and physical concepts were forbidden.

> I'm thinking that just says something along the lines of:

Let's see... (I'm criticizing your proposal just to prove what I have
written above. Your proposal largely reflects what is often written about
validation, so I'm effectively criticizing that, not this particular

> This page is valid html!

Is there any reason to shout? If you said, for some reason, that the
language of your document is grammatically correct and checked with
MS Word spelling checker (a good idea BTW, provided that you don't take
its judgement as a word of &Deity;), would you write it as an exclamation?

> This means the author of the page you came from took the care to create
> a web-page that adheres to the relevant standards.

No, it does not mean that. It could be valid for various reasons. And it
does _not_ imply conformance to relevant "standards", just conformance to
the formalized syntax rules of _a_ specification among many
specifications that are relevant to Web authoring.

> These standards are
> important since they ensure that information on the web is accessible
> to all people, regardless of what kind of computer or web-browser they
> are using.

No, they absolutely surely do not guarantee anything like that.
Validity is not about such things at all. It may have a positive
contribution in that direction, but just a minor contribution.
(And even conforming to the WAI guidelines would not ensure accessibility
to all people. Besides, practically all pages that carry the WAI icon do
not even comply with WAI guidelines.)

> To find out more about internet standards and their relevance please
> visit these links...
> [some useful links, w3c, wasp etc...]

The W3C is not an Internet standards body and does not claim to be.
Although the word "standard" can be used loosely, we should use the phrase
"Internet standard" to refer only to the documents declared as Internet
standards by the relevant organization, namely IETF.

> Of course if the page turns out to be not valid it should say something
> like:
> Oops! - -

Well, _that_ would be mildly amusing. But what would be the message
conveyed to the user, really? Would it serve the purpose of the page?

> You guys can probably word it a lot better than me

No, I don't think anyone can. Validity and validation (which should be
kept as separate issues - it is quite possible to create valid documents
reliably without a validator, e.g. by using software that generates valid
markup only) are not something that could or should be explained to the
Web browsing public. To authors or wannabe authors, something could be
said, hopefully in the tutorials or textbooks they read. If they have
missed such material, there's probably quite a lot other things to learn
before validation.

> - -  I'm doubtful whether [the validator is]
> useful for promoting the use of standards to 'normal' web users.

It isn't, and it doesn't need to be. Promoting the use of standards to
normal Web _authors_ is a different thing, but this does not mean the
validity logos would serve even that purpose, or that they should be put
onto pages that have mostly other people than authors as their readers.
Even to authors, other messages are far more important than the idea of

> What I'd like to see one day is the little W3C logo being recognised as
> a kind of seal of quality by normal users!

Well, that's an issue that I won't touch here, except by mentioning that
seals of quality generally don't work on the Web. There was a time when
seals like "Best 5%!" were common, and people soon learned not to take
them at face value, or any (positive) value.

Jukka "Yucca" Korpela,

Received on Monday, 28 July 2003 09:12:24 UTC