By using this site, you agree to our updated Privacy Policy and our Terms of Use. Manage your Cookies Settings.
424,660 Members | 1,579 Online
Bytes IT Community
+ Ask a Question
Need help? Post your question and get tips & solutions from a community of 424,660 IT Pros & Developers. It's quick & easy.

3 questions on validation

P: n/a
Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html

1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore. I did search google. Do you know such
site? What's its url?

2- Generally speaking, the most damaging markup code errors are improper
nesting. Damaging in the sense that it breaks the layout. Any comment on
this? Do you agree?

3- I've written a webpage explaining what is validation (HTML, CSS and
also WCAG). That page is targeting users who know pretty much nothing
about markup code, web authoring issues, etc. or who know little and who
might be using FrontPage and lower class products of this sort. I'd like
to hear some feedback on that page, preferably constructive criticisms,
suggestions, useful feedback, etc.

http://www.gtalbot.org/Varia/Validat...lanations.html

Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html

Gérard
--
remove blah to email me
Sep 29 '05 #1
Share this Question
Share on Google+
41 Replies


P: n/a
Gérard Talbot wrote:
Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html


You seem to have missed the point that validation is a "tool" not a
"goal"

--
-=tn=-

Sep 29 '05 #2

P: n/a
Gérard Talbot wrote:
1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore. I did search google. Do you know such
site? What's its url?
The W3C Markup Validation Service <http://validator.w3.org/> is available in
several languages. Which you get depends on the Accept-Language header your
browser sends.
2- Generally speaking, the most damaging markup code errors are improper
nesting. Damaging in the sense that it breaks the layout. Any comment on
this? Do you agree?


To be honest - I don't care. Validation lets certain errors be identified
easily. If they can be identified they can be fixed. I've never been in a
situation where I had so many errors and such a small timescale that I had
to prioritise them.

--
David Dorward <http://blog.dorward.me.uk/> <http://dorward.me.uk/>
Home is where the ~/.bashrc is
Sep 29 '05 #3

P: n/a
Gérard Talbot <ne***********@gtalbot.org> wrote:
1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore.
Tha might have been a misunderstanding. It may have been a phoney
validator.
2- Generally speaking, the most damaging markup code errors are improper
nesting. Damaging in the sense that it breaks the layout. Any comment on
this? Do you agree?
Breaking the layout isn't the worst damage. Improper nesting may or may not
cause serious damage.
3- I've written a webpage explaining what is validation (HTML, CSS and
also WCAG). - -
http://www.gtalbot.org/Varia/Validat...lanations.html


The main problem is that you don't understand what validation is.
See http://www.cs.tut.fi/~jkorpela/html/validation.html

The first statement in your document is not correct; only a relatively
small minority of pages conforms to the HTML 4.01 specification. The two
strongly emphasized statements in section "HTML 4.01 validation" are wrong,
the second one being very wrong. The rest is not about validation but
confuses validation with other types of checking.

I'm afraid that not a single statement about validation on your page is
factually correct.

--
Yucca, http://www.cs.tut.fi/~jkorpela/
Pages about Web authoring: http://www.cs.tut.fi/~jkorpela/www.html

Sep 29 '05 #4

P: n/a
Gérard Talbot <ne***********@gtalbot.org> wrote in
news:3q************@uni-berlin.de:
3- I've written a webpage explaining what is validation (HTML, CSS
and also WCAG). That page is targeting users who know pretty much
nothing about markup code, web authoring issues, etc. or who know
little and who might be using FrontPage and lower class products of
this sort. I'd like to hear some feedback on that page, preferably
constructive criticisms, suggestions, useful feedback, etc.


Nice. Can I link to it and use it for my HTML/CSS classes? I teach to
use valid HTML 4.01 and CSS1 with some of CSS2.1. I teach to validate
their code, and you've written some good explanations as to why.

I found one typo. Formating, in the CSS section should have two ts.

--
Stan McCann "Uncle Pirate" http://stanmccann.us/pirate.html
Webmaster/Computer Center Manager, NMSU at Alamogordo
http://alamo.nmsu.edu/ There are 10 kinds of people.
Those that understand binary and those that don't.
Sep 29 '05 #5

P: n/a
On Thu, 29 Sep 2005, Gérard Talbot wrote:
http://www.gtalbot.org/Varia/Validat...lanations.html


Hurled me a cookie without warning. That's rude. Of course the
browser now has yet another site in its list from which it will never
accept cookies...

But to the point.

Your page obfuscates the difference between "in conformance to the
specifications", "properly coded" and "valid". It also drifts into
the use of the term "well formed", which seems to be a specifically
XML-relevant term that doesn't formally apply to HTML.

"valid" is a technical term in HTML/SGML context, and has an exact
meaning. It's what a formal validator tests.

This would become clear if the reader would consult the W3C first:
you do indeed offer relevant links at the end of this section, but it
seems to me that after being befuddled by your introductory text, they
might be so confused as to miss the distinctions.

I'm all in favour of writing valid markup, but that's only one part of
making a web page. The mere fact that the markup is "valid" doesn't
guarantee anything about its quality in other respects.

sorry.
Sep 29 '05 #6

P: n/a
Travis Newbury a écrit :
Gérard Talbot wrote:
Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html

You seem to have missed the point that validation is a "tool" not a
"goal"

--
-=tn=-


A tool for developers: correct. But here, in that document, I'm
addressing mere visitors who see a valid HTML 4.01 (or CSS) icon and a
link explaining what this is about.

Gérard
--
remove blah to email me
Sep 29 '05 #7

P: n/a
David Dorward a écrit :
Gérard Talbot wrote:


2- Generally speaking, the most damaging markup code errors are improper
nesting. Damaging in the sense that it breaks the layout. Any comment on
this? Do you agree?

To be honest - I don't care. Validation lets certain errors be identified
easily. If they can be identified they can be fixed. I've never been in a
situation where I had so many errors and such a small timescale that I had
to prioritise them.


From a web authoring perspective, I personally share the same policy.
No validation errors on my website and I value validity and conformance
very much.
But from an end-user perspective, the ordinary visitor/end-user is going
to indirectly find that misplaced or overlapping elements is the most
damaging effect because of its visual/graphical obviousness .. and
improper nesting always create differential rendering in browsers
including overlapping.

Such document certainly won't interest readers (ordinary visitors) if I
start explaining how missing alt attribute for <img> or unescaped &'s is
invalid according to formal DTD syntax, you see.

Gérard
--
remove blah to email me
Sep 29 '05 #8

P: n/a
Jukka K. Korpela a écrit :
Gérard Talbot <ne***********@gtalbot.org> wrote:

1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore.

Tha might have been a misunderstanding. It may have been a phoney
validator.


I'll do my best to find it again.
2- Generally speaking, the most damaging markup code errors are improper
nesting. Damaging in the sense that it breaks the layout. Any comment on
this? Do you agree?

Breaking the layout isn't the worst damage.


Well, for non-web-authors, it might be. That is what I assumed.

Improper nesting may or may not cause serious damage.
Crashing the browser application certainly is more serious damage ...
but it is also rather rare.

3- I've written a webpage explaining what is validation (HTML, CSS and
also WCAG). - -
http://www.gtalbot.org/Varia/Validat...lanations.html

The main problem is that you don't understand what validation is.
See http://www.cs.tut.fi/~jkorpela/html/validation.html


Wait, wait. I may have written a not-so perfect document but I must
underline here that I am trying to address mere people with that
document. I may not have properly presented the context of such document
in my original post. At the bottom of each page, there will be a valid
HTML 4.01 icon and a valid CSS icon (both the W3C ones) and by each of
these icon, there will be a link offering explanation on what is validation.
The first statement in your document is not correct; only a relatively
small minority of pages conforms to the HTML 4.01 specification.
Sorry.. I don't understand you. This is my first statement:

"HTML 4.01 is the base language of web pages. The HTML 4.01 language has
specifications which have been set and have been agreed after years of
discussion by the World Wide Web Consortium (W3C), an international
consortium that involves the top 500 IT corporations in the world."

The two strongly emphasized statements in section "HTML 4.01 validation" are wrong,
This is the first strongly emphasized statement:
"the best starting step to ensure a webpage appears and works as
expected for all modern web browsers and web rendering software is to
write HTML code in conformance to the specifications."
I really fail to see what's wrong in there.

the second one being very wrong.
This is the 2nd strongly emphasized statement:
"The clickable W3C HTML 4.01 button-image on webpages provides a way for
any vistor to verify that such webpages are properly coded."
Well, here, I have not linked an icon to a page but when a [W3C HTML
4.01] button-image can be properly linked to the validator, then, I'd
say that people can actually verify that a webpage is properly coded.
A [W3C HTML 4.01] button-image linked to the W3C validator will verify
how such code complies with formal rules of syntax,...

The rest is not about validation but confuses validation with other types of checking.

I'm afraid that not a single statement about validation on your page is
factually correct.


Ok. I heard you.

Gérard
--
remove blah to email me
Sep 29 '05 #9

P: n/a
Stan McCann a écrit :
Gérard Talbot <ne***********@gtalbot.org> wrote in
news:3q************@uni-berlin.de:

3- I've written a webpage explaining what is validation (HTML, CSS
and also WCAG). That page is targeting users who know pretty much
nothing about markup code, web authoring issues, etc. or who know
little and who might be using FrontPage and lower class products of
this sort. I'd like to hear some feedback on that page, preferably
constructive criticisms, suggestions, useful feedback, etc.

Nice. Can I link to it and use it for my HTML/CSS classes? I teach to
use valid HTML 4.01 and CSS1 with some of CSS2.1. I teach to validate
their code, and you've written some good explanations as to why.

I found one typo. Formating, in the CSS section should have two ts.


Don't link it too fast. I might change the directory and I certainly
want first to test the content of the document with regulars in this
newsgroup and see more other 2nd opinions.

Gérard
--
remove blah to email me
Sep 29 '05 #10

P: n/a
On 29/09/2005 18:56, Jukka K. Korpela wrote:
Gérard Talbot <ne***********@gtalbot.org> wrote:
[snip]
http://www.gtalbot.org/Varia/Validat...lanations.html


[snip]
The two strongly emphasized statements in section "HTML 4.01
validation" are wrong, the second one being very wrong.
That would seem a little unfair.

As I read the first, it says as much as you have at the start of "So
what do I get from validation". Granted, it could be misread by some as
guaranteeing identical rendering across all browsers, but that's not
what it says.

The second seems wrong on technicalities. The usual URL used for
validation of a document relies on the Referer header, which obviously
doesn't cover 'any' visitor, and passing automated validation doesn't
necessarily mean valid, which is what I assume is meant by 'properly coded'.

[snip]
I'm afraid that not a single statement about validation on your page
is factually correct.


Now that's certainly not true. The first sentence in the fourth paragraph:

"Passing HTML validation does not necessarily guarantee that the
webpage is an overall good webpage or a well designed one."

is quite accurate. :P

Mike

--
Michael Winter
Prefix subject with [News] before replying by e-mail.
Sep 29 '05 #11

P: n/a
Alan J. Flavell a écrit :
On Thu, 29 Sep 2005, Gérard Talbot wrote:

http://www.gtalbot.org/Varia/Validat...lanations.html

Hurled me a cookie without warning. That's rude. Of course the
browser now has yet another site in its list from which it will never
accept cookies...


That,s most likely my web host server doing that. I don't create/set/use
cookies in my code.
But to the point.

Your page obfuscates the difference between "in conformance to the
specifications", "properly coded" and "valid".
Yeah... not too long ago, I asked about this and I probably did not
modify that document.

It also drifts into the use of the term "well formed", which seems to be a specifically
XML-relevant term that doesn't formally apply to HTML.

I was told that. I will change that in my document.
Originally, in my mind, well-formed meant without any improper nesting.
But here, I have to remove this everywhere.. Thanks for pointing me on this.
"valid" is a technical term in HTML/SGML context, and has an exact
meaning. It's what a formal validator tests.

This would become clear if the reader would consult the W3C first:
you do indeed offer relevant links at the end of this section, but it
seems to me that after being befuddled by your introductory text, they
might be so confused as to miss the distinctions.

Ok.
I'm all in favour of writing valid markup, but that's only one part of
making a web page. The mere fact that the markup is "valid" doesn't
guarantee anything about its quality in other respects.


Well, isn't that what I am saying too?
"Passing HTML validation does not necessarly guarantee that the webpage
is an overall good webpage or a well designed one."
is in the HTML 4.01 validation section ...

Gérard
--
remove blah to email me
Sep 29 '05 #12

P: n/a
Gérard Talbot wrote:
Travis Newbury a écrit :
Gérard Talbot wrote:
Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html


You seem to have missed the point that validation is a "tool" not a
"goal"


A tool for developers: correct. But here, in that document, I'm
addressing mere visitors who see a valid HTML 4.01 (or CSS) icon and a
link explaining what this is about.


Why is a visitor interested at all? The visitor only cares if the page
looks OK and is usable. What would be useful would be a logo announcing
NON-conformance, to be place on pages with invalid code, so that users
having trouble with a page, when they see the logo, will say, "Oh, no
wonder, the code's invalid. No surprise that my browser isn't displaying
it correctly."
Sep 29 '05 #13

P: n/a
"Alan J. Flavell" <fl*****@ph.gla.ac.uk> wrote in
news:Pi*******************************@ppepc56.ph. gla.ac.uk:
I'm all in favour of writing valid markup, but that's only one part
of making a web page. The mere fact that the markup is "valid"
doesn't guarantee anything about its quality in other respects.

sorry.


You've nothing to be sorry for stating your comments, especially when
you give constructive criticism as Gerard asked for. As usual, Alan,
you have some good things to say. This last comment caught my eye in
that he does mention this, just not real forcefully.

Personally, I found the information put together in one place good and
hope to use it with my teaching. I'm sure it will get cleaned up and
be even better and more useful with the commentary he is getting.

--
Stan McCann "Uncle Pirate" http://stanmccann.us/pirate.html
Webmaster/Computer Center Manager, NMSU at Alamogordo
http://alamo.nmsu.edu/ There are 10 kinds of people.
Those that understand binary and those that don't.
Sep 29 '05 #14

P: n/a
Harlan Messinger a écrit :
Gérard Talbot wrote:
Travis Newbury a écrit :
Gérard Talbot wrote:

Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html
You seem to have missed the point that validation is a "tool" not a
"goal"

>

A tool for developers: correct. But here, in that document, I'm
addressing mere visitors who see a valid HTML 4.01 (or CSS) icon and a
link explaining what this is about.

Why is a visitor interested at all? The visitor only cares if the page
looks OK and is usable. What would be useful would be a logo announcing
NON-conformance, to be place on pages with invalid code, so that users
having trouble with a page, when they see the logo, will say, "Oh, no
wonder, the code's invalid. No surprise that my browser isn't displaying
it correctly."


Actually more and more browsers are doing that in various ways, are
announcing erroneous or problematic code which may affect the rendering
in pages in various ways.

Icab with its frown icon, Dillo browser apparently (I have not verified
their claim), Amaya 9.2.1 does that indicating clearly the parsing
errors, HTML Tidy (more of a linter but nevertheless), as a Firefox
extension, will report unknown elements, unknown attributes, improper
nesting, wrong code, etc.

Gérard
--
remove blah to email me
Sep 29 '05 #15

P: n/a
On Thu, 29 Sep 2005, Gérard Talbot wrote, quoting me:

[...]
I'm all in favour of writing valid markup, but that's only one
part of making a web page. The mere fact that the markup is
"valid" doesn't guarantee anything about its quality in other
respects.
Well, isn't that what I am saying too?


I hope so! I just felt that the salad of technical terms was a bit
muddled.

My counsel would be to carefully avoid the use of any of the specific
technical terms, /except/ in contexts where they are used in their
precise sense. Rather, develop the background by concentrating on
non-specific language which talks about factors, both objective and
subjective, which contribute to the quality of a final web page/site,
and then point to the extent to which these useful tools can
contribute... I feel that, if it's done right, the W3C material would
then practically fall into place when you link to it at the right
moment.
"Passing HTML validation does not necessarly guarantee that the
webpage is an overall good webpage or a well designed one." is in
the HTML 4.01 validation section ...


Your conclusion is OK. I was worrying more about how you got there,
if you see what I mean.

In the end, it's your document, of course - I'm only offering you my
reactions.

good luck
Sep 29 '05 #16

P: n/a
Jukka K. Korpela a écrit :
Gérard Talbot <ne***********@gtalbot.org> wrote:

1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore.

Tha might have been a misunderstanding. It may have been a phoney
validator.

2- Generally speaking, the most damaging markup code errors are improper
nesting. Damaging in the sense that it breaks the layout. Any comment on
this? Do you agree?

Breaking the layout isn't the worst damage. Improper nesting may or may not
cause serious damage.

3- I've written a webpage explaining what is validation (HTML, CSS and
also WCAG). - -
http://www.gtalbot.org/Varia/Validat...lanations.html

The main problem is that you don't understand what validation is.
See http://www.cs.tut.fi/~jkorpela/html/validation.html

The first statement in your document is not correct; only a relatively
small minority of pages conforms to the HTML 4.01 specification. The two
strongly emphasized statements in section "HTML 4.01 validation" are wrong,
the second one being very wrong. The rest is not about validation but
confuses validation with other types of checking.

I'm afraid that not a single statement about validation on your page is
factually correct.


Regarding the HTML 4.01 validation section:
-------------------------------------------

Ok, I modified the HTML 4.01 validation section. If you grant me that
the first paragraph (about HTML 4.01 being the base language and its
relation to W3C) is not really important, then the whole HTML 4.01
validation section has 8 sentences. I removed the expression "markup
code" and replaced it with "HTML code" instead, to simplify things a bit
for the eventual readers. Another reason is that all my webpages will be
HTML 4.01 strict; I will not be using XHTML at all.

Generally speaking, when I say "properly coded", I mean complying with
formal rules of syntax and semantics of the SGML but I definately do not
want to mention SGML anywhere, so I refer to HTML 4.01 specification ...
and nothing more.

I don't want to develop too much into details for the targeted audience.

Gérard
--
remove blah to email me
Sep 29 '05 #17

P: n/a

Content-Transfer-Encoding: 8Bit
Gérard Talbot wrote:
Don't link it too fast. I might change the directory


Look here:
http://www.w3.org/Provider/Style/URI
Sep 29 '05 #18

P: n/a
Harlan Messinger a écrit :
Gérard Talbot wrote:
Travis Newbury a écrit :
Gérard Talbot wrote:

Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html
You seem to have missed the point that validation is a "tool" not a
"goal"

>

A tool for developers: correct. But here, in that document, I'm
addressing mere visitors who see a valid HTML 4.01 (or CSS) icon and a
link explaining what this is about.

Why is a visitor interested at all? The visitor only cares if the page
looks OK and is usable. What would be useful would be a logo announcing
NON-conformance, to be place on pages with invalid code, so that users
having trouble with a page, when they see the logo, will say, "Oh, no
wonder, the code's invalid. No surprise that my browser isn't displaying
it correctly."


Another point. If a visitor has a problem - whatever it might be - with
a page, he can check the validity of that page thanks to a properly
linked image-button HTML 4.01 valid and then possibly conclude that the
problem is not caused by invalid code.

In that same spirit, browser manufacturers like mozilla.org and Opera
have groups, organizations and now related features trying to tackle the
issue of websites having problems being rendered in/with their
respective browsers.
"Open the web" community for Opera and Report a broken website to the
Bugzilla/mozilla Evangelism team via a built-in wizard or via this page:
http://reporter.mozilla.org/app/

Another point. B.1 Notes on invalid documents
http://www.w3.org/TR/html4/appendix/notes.html#h-B.1
goes on to saying:
"This [HTML 4.01] specification does not define how conforming user
agents handle general error conditions (...) We also recommend that user
agents provide support for notifying the user of such errors. Since user
agents may vary in how they handle error conditions, authors and users
must not rely on specific error recovery behavior."
I tried to convey those ideas in the HTML 4.01 validation section of my
document.

Gérard
--
remove blah to email me
Sep 29 '05 #19

P: n/a
In article <3q************@uni-berlin.de>,
Gérard Talbot <ne***********@gtalbot.org> wrote:
1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore. I did search google. Do you know such
site? What's its url?


http://www.validome.org/

--
Henri Sivonen
hs******@iki.fi
http://hsivonen.iki.fi/
Mozilla Web Author FAQ: http://mozilla.org/docs/web-developer/faq.html
Sep 29 '05 #20

P: n/a
With neither quill nor qualm, =?ISO-8859-1?Q?G=E9rard_Talbot?= quothed:
Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html

1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore. I did search google. Do you know such
site? What's its url?

2- Generally speaking, the most damaging markup code errors are improper
nesting. Damaging in the sense that it breaks the layout. Any comment on
this? Do you agree?

3- I've written a webpage explaining what is validation (HTML, CSS and
also WCAG). That page is targeting users who know pretty much nothing
about markup code, web authoring issues, etc. or who know little and who
might be using FrontPage and lower class products of this sort. I'd like
to hear some feedback on that page, preferably constructive criticisms,
suggestions, useful feedback, etc.

http://www.gtalbot.org/Varia/Validat...lanations.html

Cross-posted to: comp.infosystems.www.authoring.html and alt.html
Followup-to: comp.infosystems.www.authoring.html


I like it a lot, -think it's informative and intelligent. My only real
criticism is that your teeny w3c logo image sucks.

--
Neredbojias
Contrary to popular belief, it is believable.
Sep 29 '05 #21

P: n/a
Henri Sivonen a écrit :
In article <3q************@uni-berlin.de>,
Gérard Talbot <ne***********@gtalbot.org> wrote:

1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore. I did search google. Do you know such
site? What's its url?

http://www.validome.org/


Right on the nose! Thank you very much, Henri! This is appreciated; I
was going nuts trying to find it back.

Gérard
--
remove blah to email me
Sep 29 '05 #22

P: n/a
Jukka K. Korpela a écrit :
Gérard Talbot <ne***********@gtalbot.org> wrote:

1- One day, I stumbled across a website that offers to validate
webpages. What was really special about that site is that the validation
service was available in Russian, German, Spanish and English. Now, I
can't find this site anymore.


Tha might have been a misunderstanding. It may have been a phoney
validator.

Henri Sivonen found it!
http://www.validome.org/

Gérard
--
remove blah to email me
Sep 29 '05 #23

P: n/a
Michael Winter wrote:
The two strongly emphasized statements in section "HTML 4.01
validation" are wrong, the second one being very wrong.
That would seem a little unfair.


I wasn't trying to be fair. I see no reason to be fair with a document
that just adds to the worldwide supply of bogosity about validation.

Anyone who purports to explain to others what something is should first
understand it himself, or at least learn it in the process of writing.
There's no reason to say that something that is completely wrong is just
poorly formulated; that would just be an encouragement to go on in
explaining things without understanding them.
As I read the first, it says as much as you have at the start of "So
what do I get from validation".
It doesn't. Maybe you already knew what validation is and therefore read
it as saying something else than it actually says. Or you read the two
documents carelessly.
Granted, it could be misread by some as
guaranteeing identical rendering across all browsers, but that's not
what it says.
The emphasized text makes a strong promise in that direction,
particularly when read by a person who does not know what validation is
and has been affected by the usual propaganda about validation - that
is, by a typical assumed reader. "The best starting step to ensure a
webpage appears and works as expected for all modern web browsers - -".
The best? Ensure? All? That's _seriously_ wrong.
The second seems wrong on technicalities.
Validation is about technicalities. If you are wrong on them, you are
wrong on everything. Besides, it's no technicality to promote the icon
and to babble about users checking validity. It's horrendously wrong.
I'm afraid that not a single statement about validation on your page
is factually correct.


Now that's certainly not true.


"I'm afraid that - -" is something you cannot know about; I read the
most emphatic parts of the document, looked at a few others, noting they
are all wrong, and that made me expect that everything is wrong.

OK, "I'm afraid that" could also be read as a phrase with no
denotational meaning, just meant to soften what follows, so to say. In
that case, I'm guilty as charged. This does not change the fact that the
page is bogosity and much worse than useless. Trying to "fix" it is
pointless.
The first sentence in the fourth paragraph:

"Passing HTML validation does not necessarily guarantee that the
webpage is an overall good webpage or a well designed one."

is quite accurate. :P


And contradicts the emphatic promise that the page makes. Besides, to be
exact, "HTML validation" is an incorrect term, and the sentence is
misleading, since "does not necessarily guarantee" says, or can easily
be read as saying, that it sometimes guarantees, but not necessarily. In
any case, the word "necessarily" does not belong there. So the sentence
isn't quite accurate. It's also wrongly emphasized, the same way as the
following would be: "Passing a spelling check that MS Word performs does
not necessarily guarantee that a document is an overall good document or
a well designed one."
Sep 30 '05 #24

P: n/a
Gérard Talbot <ne***********@gtalbot.org> writes:
Gérard Talbot wrote:
2- Generally speaking, the most damaging markup code errors are improper
nesting. Damaging in the sense that it breaks the layout. Any comment on
this? Do you agree?


Such document certainly won't interest readers (ordinary visitors) if
I start explaining how missing alt attribute for <img> or unescaped
&'s is invalid according to formal DTD syntax, you see.


Hmm, I'd say that *both* of those have the potential to be as (or
more) damaging than badly nested elements.

<a href="......&lang=..."> will prevent the link from working in some
browsers (see the current Alexa/validation thread for the original) -
some browsers will error-recover it as the author expects, some won't.

<img src="..."> can potentially make the page entirely
incomprehensible in some browsing situations. No browser can do
error-recovery on this.

--
Chris
Sep 30 '05 #25

P: n/a
On Thu, 29 Sep 2005, David Dorward wrote:
The W3C Markup Validation Service <http://validator.w3.org/> is available in
several languages. Which you get depends on the Accept-Language header your
browser sends.


In which languages? I tried de, fr, sa without success.

Sep 30 '05 #26

P: n/a
On Thu, 29 Sep 2005, Gérard Talbot wrote:
"The clickable W3C HTML 4.01 button-image on webpages provides a way for
any vistor to verify that such webpages are properly coded." ^^^^^^^^^^^ Well, here, I have not linked an icon to a page but when a [W3C HTML
4.01] button-image can be properly linked to the validator, then, I'd
say that people can actually verify that a webpage is properly coded.

^^^^^^^^^^^

Not "verify that" but "check whether". And even that might not
be possbile. If you follow the validator's advice, you end up with

<noframes>
...
<img ... alt="Valid HTML 4.01 Frameset">
...
</noframes>

for a valid frameset. ;-)

Sep 30 '05 #27

P: n/a
On Fri, 30 Sep 2005, Chris Morris wrote:
<a href="......&lang=..."> will prevent the link from working in
some browsers (see the current Alexa/validation thread for the
original) -


Could I have a real example of this, please? Except, in some cases,
for the Latin-1 character entity names from earlier versions of HTML,
all of my tests for

http://ppewww.ph.gla.ac.uk/~flavell/...mgetbyurl.html

seemed to be showing browsers doing what the misguided author meant -
rather than what the specs would require. And that included the
version of Lynx which I was testing. It would be great to be able to
exhibit some real examples of browsers doing the wrong thing with
*all* of the additional HTML4 character entities :-}

thanks
Sep 30 '05 #28

P: n/a
Andreas Prilop wrote:
On Thu, 29 Sep 2005, David Dorward wrote:
The W3C Markup Validation Service <http://validator.w3.org/> is available
in several languages. Which you get depends on the Accept-Language header
your browser sends.
In which languages? I tried de, fr, sa without success.


hmm, looks like I might have been getting it mixed up with the CSS Validator
(which certainly supports French).

--
David Dorward <http://blog.dorward.me.uk/> <http://dorward.me.uk/>
Home is where the ~/.bashrc is
Sep 30 '05 #29

P: n/a
On Fri, 30 Sep 2005, David Dorward wrote:
hmm, looks like I might have been getting it mixed up with the CSS Validator
(which certainly supports French).


Right! It has de, es, fr versions - but no sa though.

Sep 30 '05 #30

P: n/a
"Alan J. Flavell" <fl*****@ph.gla.ac.uk> writes:
On Fri, 30 Sep 2005, Chris Morris wrote:
<a href="......&lang=..."> will prevent the link from working in
some browsers (see the current Alexa/validation thread for the
original) -


Could I have a real example of this, please? Except, in some cases,
for the Latin-1 character entity names from earlier versions of HTML,
all of my tests for

http://ppewww.ph.gla.ac.uk/~flavell/...mgetbyurl.html

seemed to be showing browsers doing what the misguided author meant -
rather than what the specs would require. And that included the
version of Lynx which I was testing. It would be great to be able to
exhibit some real examples of browsers doing the wrong thing with
*all* of the additional HTML4 character entities :-}


Hmm, and including the ancient Lynx 2-4-2, even. Possibly, like <br />
in text/html, it was such a common thing that browsers _had_ to be
bug-compatible with their predecessors. Ignoring the apparent entity
seems to be consistent behaviour right back to the line-mode browser,
anyway.

--
Chris
Sep 30 '05 #31

P: n/a
On 30/09/2005 06:00, Jukka K. Korpela wrote:
Michael Winter wrote:


[snip]
As I read the first, it says as much as you have at the start of "So
what do I get from validation".


It doesn't. Maybe you already knew what validation is and therefore read
it as saying something else than it actually says.


I am interpreting it in a more favourable light than you obviously are.
I added more emphasis to the phrase "starting step" than others would. I
would also have very different expectations of modern browsers than any
lay-person might. However, that doesn't make it completely wrong. It
/does/ make it inappropriate and misleading for the intended audience,
but that's a more useful message for the OP, in my opinion, if he is
intent on retaining the article.

[snip]

Mike

--
Michael Winter
Prefix subject with [News] before replying by e-mail.
Sep 30 '05 #32

P: n/a
Gérard Talbot <ne***********@gtalbot.org> wrote:
The first statement in your document is not correct; only a relatively
small minority of pages conforms to the HTML 4.01 specification.
Sorry.. I don't understand you.


So it seems.
This is my first statement:

"HTML 4.01 is the base language of web pages. The HTML 4.01 language has
specifications which have been set and have been agreed after years of
discussion by the World Wide Web Consortium (W3C), an international
consortium that involves the top 500 IT corporations in the world."


No, that's at least two statements. Read your first statement and compare
it with my explanation. The long second sentence has errors too (starting
from the fact that there is only one HTML 4.01 specification), but I was
not commenting on them.

--
Yucca, http://www.cs.tut.fi/~jkorpela/
Pages about Web authoring: http://www.cs.tut.fi/~jkorpela/www.html

Sep 30 '05 #33

P: n/a
Jukka K. Korpela a écrit :
Gérard Talbot <ne***********@gtalbot.org> wrote:

The first statement in your document is not correct; only a relatively
small minority of pages conforms to the HTML 4.01 specification.


Sorry.. I don't understand you.

So it seems.

This is my first statement:

"HTML 4.01 is the base language of web pages. The HTML 4.01 language has
specifications which have been set and have been agreed after years of
discussion by the World Wide Web Consortium (W3C), an international
consortium that involves the top 500 IT corporations in the world."

No, that's at least two statements. Read your first statement and compare
it with my explanation. The long second sentence has errors too (starting
from the fact that there is only one HTML 4.01 specification), but I was
not commenting on them.


Ok, let me get this straight now.
I said in a document this statement:
"HTML 4.01 is the base language of web pages."
Then you replied this explanation:
"The first statement in your document is not correct; only a relatively
small minority of pages conforms to the HTML 4.01 specification."

Did I get all this right?

Gérard
--
remove blah to email me
Sep 30 '05 #34

P: n/a
Jukka K. Korpela a écrit :
Michael Winter wrote:
The two strongly emphasized statements in section "HTML 4.01
validation" are wrong, the second one being very wrong.

That would seem a little unfair.

I wasn't trying to be fair. I see no reason to be fair with a document
that just adds to the worldwide supply of bogosity about validation.

Anyone who purports to explain to others what something is should first
understand it himself,


What exactly makes you so sure that I don't understand what validation
is, means, etc..? You can't read minds over the internet and then, just
like that, establish what others understand and don't understand.
or at least learn it in the process of writing.
Your posts bash pretty hard someone who wanted to get balanced feedback
and constructive comments on that page to begin with. I was open-minded,
you see.
There's no reason to say that something that is completely wrong is just
poorly formulated; that would just be an encouragement to go on in
explaining things without understanding them.

Another simple thing you missed. That page is linked from no page in my
website but to other resources pages (W3C pages) providing more
explanations.

Before linking that page in a part of my website and before translating
it, I wanted to make a final check - a public one - of the document by
inviting comments, feedback from this newsgroup.
As I read the first, it says as much as you have at the start of "So
what do I get from validation".

It doesn't. Maybe you already knew what validation is and therefore read
it as saying something else than it actually says. Or you read the two
documents carelessly.
Granted, it could be misread by some as guaranteeing identical
rendering across all browsers, but that's not what it says.

The emphasized text makes a strong promise in that direction,


Ok. I read the document again and more carefully and I see what you mean
now with the "strong promise in that direction".
particularly when read by a person who does not know what validation is
and has been affected by the usual propaganda about validation - that
is, by a typical assumed reader. "The best starting step to ensure a
webpage appears and works as expected for all modern web browsers - -".
The best? Ensure? All? That's _seriously_ wrong.

I wish I could discuss that a bit more... Such statement is more of/from
a web author perspective than meeting the visitor's perspective
regarding this issue. To claim that my statement

"the best starting step to ensure a webpage appears and works as
expected for all modern web browsers and web rendering software is to
write HTML code in conformance to the specifications"

is seriously wrong is quite a stunning claim in itself. You'll have a
hard time substantiating this. Whenever I meet a problem in a webpage, a
layout one - in particular when 2 modern browsers render different
layout - or a problem of another nature, I first check the validatity of
the page and then correct markup errors. Realistically speaking, there
is no other better first step to do. A long history of posts in this
newsgroup (and in alt.html) will back me on this.
The second seems wrong on technicalities.

Validation is about technicalities. If you are wrong on them, you are
wrong on everything. Besides, it's no technicality to promote the icon
and to babble about users checking validity. It's horrendously wrong.


In 2002, the W3C estimated that 98% of all webpages on the web would
fail validation test. Using an icon to promote validation, when done
properly, is making some sense in that regard, in that perspective.
Almost all of the sites listed here
http://www.w3csites.com/
have/use/display some kind of (custom or not) valid icon at the bottom
of their website pages: are they all pointlessly and/or horrendously wrong?
I'm afraid that not a single statement about validation on your page
is factually correct.

Now that's certainly not true.

"I'm afraid that - -" is something you cannot know about; I read the
most emphatic parts of the document, looked at a few others, noting they
are all wrong, and that made me expect that everything is wrong.

OK, "I'm afraid that" could also be read as a phrase with no
denotational meaning, just meant to soften what follows, so to say. In
that case, I'm guilty as charged.


It took you quite a lot of words to explain, give detail and nuances on
what you meant with the expression "I'm afraid that". If I want the
normal visitor to understand what valid HTML and validity means, I don't
want such explanations to be too long. So, in my mind, approximation may
happen and details, sharp explanations will be avoided in that document.
This does not change the fact that the
page is bogosity and much worse than useless.
What you say here is excessive... The draft version you read had
problems, misuse of terms. Acknowledged...

Trying to "fix" it is pointless.

....but you're saying much more than that. You're saying the whole
purpose of the document is worthless, even pointless.

The first sentence in the fourth paragraph:

"Passing HTML validation does not necessarily guarantee that the
webpage is an overall good webpage or a well designed one."

is quite accurate. :P

And contradicts the emphatic promise that the page makes. Besides, to be
exact, "HTML validation" is an incorrect term, and the sentence is
misleading, since "does not necessarily guarantee" says, or can easily
be read as saying, that it sometimes guarantees, but not necessarily. In
any case, the word "necessarily" does not belong there.


I agree with you on this: "necessarly" does not belong in that sentence.
Even the verb "guarantee" may not be the best; the verb "mean" fits
better the intended idea that I wanted to convey.

So the sentence isn't quite accurate. It's also wrongly emphasized, the same way as the
following would be: "Passing a spelling check that MS Word performs does
not necessarily guarantee that a document is an overall good document or
a well designed one."


Of course it does not mean that the document is a good document but it
certainly checks one significant aspect of editing of the document and
is an unavoidable part of the whole process of creating a good document
and/or a well designed one.
Strangely enough, my initial version was using the spell checking
analogy. I wanted to cut off the document to mere bones and lean stuff:
I may have gone too far or done so awkwardly.

Gérard
--
remove blah to email me
Oct 1 '05 #35

P: n/a
In article <3q************@uni-berlin.de>,
Gérard Talbot <ne***********@gtalbot.org> wrote:
http://www.gtalbot.org/Varia/Validat...lanations.html


| Therefore, the best starting step to ensure a webpage appears
| and works as expected for all modern web browsers and web rendering
| software is to write HTML code in conformance to the specifications.

But validity does not imply conformance. Also, it is possibly to be
valid but still incompatible with the real world.

http://hsivonen.iki.fi/test/minimization.html

| Generally speaking, a webpage using valid HTML code is smaller in
| size (so downloading is faster) and is also rendered
| considerably faster in modern browsers.

What's the basis of those (false) claims?

| The W3C provides a public HTML validator service that
| automatically checks a submitted web page against the formal
| rules (semantic, syntax) of the HTML specification and then
| report any error found.

The W3C validator makes no semantic checks whatsoever.

| The clickable W3C HTML 4.01 button-image on webpages provides a
| way for any vistor to verify that such webpages' HTML code is
| properly coded.

http://www.cs.tut.fi/~jkorpela/html/...tion.html#icon

| Passing HTML validation does not necessarly guarantee that the
| webpage is an overall good webpage or a well designed one.

Nor does it guarantee conformance. It does not even guarantee that the
parse tree in the browser is the same as in the validator.

How about just linking to
http://www.cs.tut.fi/~jkorpela/html/validation.html ?

--
Henri Sivonen
hs******@iki.fi
http://hsivonen.iki.fi/
Mozilla Web Author FAQ: http://mozilla.org/docs/web-developer/faq.html
Oct 1 '05 #36

P: n/a
In article <3q************@uni-berlin.de>,
Gérard Talbot <ne***********@gtalbot.org> wrote:
Your posts bash pretty hard someone who wanted to get balanced feedback
and constructive comments on that page to begin with.


It is quite normal for authors of accurate Web documents (to which
anyone is free to link) to become irked when others publish less
accurate or downright bogus documents about the same subject matter.

I know this from experience having published documents of above-average
accuracy about doctype sniffing and PNG gamma issues.

--
Henri Sivonen
hs******@iki.fi
http://hsivonen.iki.fi/
Mozilla Web Author FAQ: http://mozilla.org/docs/web-developer/faq.html
Oct 1 '05 #37

P: n/a
Henri Sivonen a écrit :
In article <3q************@uni-berlin.de>,
Gérard Talbot <ne***********@gtalbot.org> wrote:

http://www.gtalbot.org/Varia/Validat...lanations.html

| Therefore, the best starting step to ensure a webpage appears
| and works as expected for all modern web browsers and web rendering
| software is to write HTML code in conformance to the specifications.

But validity does not imply conformance. Also, it is possibly to be
valid but still incompatible with the real world.

http://hsivonen.iki.fi/test/minimization.html

| Generally speaking, a webpage using valid HTML code is smaller in
| size (so downloading is faster) and is also rendered
| considerably faster in modern browsers.

What's the basis of those (false) claims?


- CSS is assumed when a document complies with web standards, W3C
recommendations. I read somewhere that valid HTML documents often assume
CSS implementation where code reusability is best and optimal.
- Quirks mode implies more checking, more efforts in the code to render
malformed code or invalid code; David Hyatt in his blog gave examples of
this. Generally speaking, valid code is parsed and rendered faster on
web standards compliant browsers because newer browser versions are
geared, optimized, tuned to comply furthermore with web standards.
| The W3C provides a public HTML validator service that
| automatically checks a submitted web page against the formal
| rules (semantic, syntax) of the HTML specification and then
| report any error found.

The W3C validator makes no semantic checks whatsoever.

It will report unknown elements and unknown attributes and mismatch of
type value. OTOH, it won't report e.g. that this document may be
misusing tables for layout instead of rendering tabular data. So there
is some semantic checking.
| The clickable W3C HTML 4.01 button-image on webpages provides a
| way for any vistor to verify that such webpages' HTML code is
| properly coded.

http://www.cs.tut.fi/~jkorpela/html/...tion.html#icon

| Passing HTML validation does not necessarly guarantee that the
| webpage is an overall good webpage or a well designed one.

Nor does it guarantee conformance.
Correct. It does not guarantee conformance but it's a start. A first
step toward conformance. A good first step considering that 98% or so of
billions of webpages on the web can't pass validation test.

It does not even guarantee that the parse tree in the browser is the same as in the validator.

Ok. That, I really don't know.
How about just linking to
http://www.cs.tut.fi/~jkorpela/html/validation.html ?


Well, I would rather prefer to use a document that better fits my
perspective. Jukka does not agree with valid icons of any sort.

Gérard
--
remove blah to email me
Oct 1 '05 #38

P: n/a

Content-Transfer-Encoding: 8Bit

Gérard Talbot wrote:
Ok, let me get this straight now.
I said in a document this statement:
"HTML 4.01 is the base language of web pages."


But it isn't!

The *base* language of web pages (the language upon which all webpages
are based) is the version of HTML which was invented by Tim Berners-Lee
in 1989. He in turn based his invention on SGML (Standard Generalized
Mark-up Language) and on Bill Atkinson's Hypercard for the Macintosh,
which was itself based on Vannevar Bush's memex and Ted Nelson's Xanadu.
HTML 4.01 is a refinement (not the forst and not the last), not the base.

Here is an example of a document writen in "the base language of web pages":
http://www.w3.org/History/19921103-h...Technical.html

Note the lack of any sort of images or graphics. A look at the source
code is quite instructive, as is a close look at the original HyperText.m
at http://www.w3.org/History/1991-WWW-N...on/HyperText.m
and at the Proposal at http://www.w3.org/History/1989/proposal.html
--
Guy Macon <http://www.guymacon.com/>

Oct 1 '05 #39

P: n/a
Guy Macon a écrit :
Content-Transfer-Encoding: 8Bit

Gérard Talbot wrote:

Ok, let me get this straight now.
I said in a document this statement:
"HTML 4.01 is the base language of web pages."

But it isn't!

The *base* language of web pages (the language upon which all webpages
are based) is the version of HTML which was invented by Tim Berners-Lee
in 1989. He in turn based his invention on SGML (Standard Generalized
Mark-up Language) and on Bill Atkinson's Hypercard for the Macintosh,
which was itself based on Vannevar Bush's memex and Ted Nelson's Xanadu.
HTML 4.01 is a refinement (not the forst and not the last), not the base.

Here is an example of a document writen in "the base language of web pages":
http://www.w3.org/History/19921103-h...Technical.html

Note the lack of any sort of images or graphics. A look at the source
code is quite instructive, as is a close look at the original HyperText.m
at http://www.w3.org/History/1991-WWW-N...on/HyperText.m
and at the Proposal at http://www.w3.org/History/1989/proposal.html


I just uploaded a revised version of the document. This point has been
modified. Now, it goes like this:

"HTML is the base language of webpages. The HTML language has
specifications which have been set after years of discussion by the
World Wide Web Consortium (W3C), an international consortium that
involves the top 500 IT corporations in the world. The latest and final
specification of the HTML language is HTML 4.01."

And I really don't want to get into details like XHTML 1.1 is the latest
spec of the HTML language... otherwise, I won't meet the purpose of the
document.

Gérard
--
remove blah to email me
Oct 1 '05 #40

P: n/a
Guy Macon <http://www.guymacon.com/> writes:
Here is an example of a document writen in "the base language of web pages":
http://www.w3.org/History/19921103-h...Technical.html

Note the lack of any sort of images or graphics.


Also note the lack of any sort of formatting or indenting in the code... Do
Swiss keyboards lack tab keys? :-)

sherm--

--
Cocoa programming in Perl: http://camelbones.sourceforge.net
Hire me! My resume: http://www.dot-app.org
Oct 1 '05 #41

P: n/a
Henri Sivonen a écrit :
In article <3q************@uni-berlin.de>,
Gérard Talbot <ne***********@gtalbot.org> wrote:

http://www.gtalbot.org/Varia/Validat...lanations.html

| Therefore, the best starting step to ensure a webpage appears
| and works as expected for all modern web browsers and web rendering
| software is to write HTML code in conformance to the specifications.

But validity does not imply conformance.


True. But validity is a necessary step toward achieving conformance. The
way I understood this is that validity is a subset of conformance to the
specification. The words "starting step" are important in the above
sentence.

I want to try to stay away from technical terms, technical discussion in
the document as much as possible. I provide links which can elaborate
more on this validity-versus-conformance issue.

What I want the ordinary visitor to understand is that those icons,
particularly the HTML 4.01 icon, serve the purpose of checking whether
the document meets *_basic coding requirements_* for a webpage: that's
what I think the icon button should serve as a purpose.
Also, it is possibly to be
valid but still incompatible with the real world.

http://hsivonen.iki.fi/test/minimization.html

| Generally speaking, a webpage using valid HTML code is smaller in
| size (so downloading is faster) and is also rendered
| considerably faster in modern browsers.

What's the basis of those (false) claims?

In the revised version, I changed that sentence to

"Generally speaking, a webpage using valid HTML code along with valid
CSS code is smaller in size (so downloading is faster) and is also
rendered considerably faster in modern browsers."

| The W3C provides a public HTML validator service that
| automatically checks a submitted web page against the formal
| rules (semantic, syntax) of the HTML specification and then
| report any error found.

The W3C validator makes no semantic checks whatsoever.
I added that validation checks most of the formal rules, not all. The
checking of most of the formal rules can be done automatically; others
must be verified by examining manually the markup code.
I added document structure in the formal rules.

| The clickable W3C HTML 4.01 button-image on webpages provides a
| way for any vistor to verify that such webpages' HTML code is
| properly coded.


I changed and replaced "properly coded" with "complies with most of such
formal rules"

Gérard
--
remove blah to email me
Oct 1 '05 #42

This discussion thread is closed

Replies have been disabled for this discussion.