W3C attack on HTML

Someone in W3C just cannot accept the idea of so much time (your own
and everyone else) wasted for nothing (for XHTML to be exact). So
instead of concentrating on something useful instead they decided to
squeeze HTML out - as if it can push anyone for XHTML.

To things to be aware of in this "W3C against of HTTP war" (a crazy
title - but close to the truth):

1. As of recently even fully correct Strict page with DTD linked *and*
served with server Content-Type header "text/html":

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd">
<html>
<head>
<title>Empty</title>
<meta http-equiv="Content-Type"
content="text/html; charset=iso-8859-1">
</head>

<body>
<p>No content</p>
</body>
</html>

leads to a huge warning:

<quote>
Unknown Parse Mode!
The MIME Media Type (text/html) for this document is used to serve both
SGML and XML based documents, and it is not possible to disambiguate it
based on the DOCTYPE Declaration in your document. Parsing will
continue in SGML mode.
</quote>

Please feel free to ignore this warning: it has no technical sense,
factually wrong, misleading and contradicting to W3C own specs.
2. Starting this week this warning is used as a formal excuse to not
propose any more "Valid HTML 4.01" banner even if the validation itself
is passed.

If you passed the validation (green banner) and still want to place the
banner on your site, the relevant image url is
<http://www.w3.org/Icons/valid-html401> and the relevant HTML code
could be like:

<p><a
href="http://validator.w3.org/check?uri=referer"><img
src="http://www.w3.org/Icons/valid-html401"
width="88" height="31"
alt="Valid HTML 4.01"
style="border: 1px none"></a></p>
If in the future W3C server stops serving the banner, it can be
re-linked from a number of locations, say from
<http://www.geocities.com/schools_ring/html401/valid-html401.bmp>.
But as of today such extreme is not necessary (yet?).

Apr 24 '06 #1

Subscribe Post Reply

2061

frederick

VK wrote:

Someone in W3C just cannot accept the idea of so much time (your own
and everyone else) wasted for nothing (for XHTML to be exact). So
instead of concentrating on something useful instead they decided to
squeeze HTML out - as if it can push anyone for XHTML.
Next time you launch a tirade against something, I suggest that you be
a bit more careful to verify your "facts"!
1. As of recently even fully correct Strict page with DTD linked *and*
served with server Content-Type header "text/html":

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd">

[snip]

It's hardly surprising that the validator didn't like your doctype:
it's invalid!
--
AGw.

Apr 24 '06 #2

Harlan Messinger

VK wrote:

Someone in W3C just cannot accept the idea of so much time (your own
and everyone else) wasted for nothing (for XHTML to be exact). So
instead of concentrating on something useful instead they decided to
squeeze HTML out - as if it can push anyone for XHTML.

To things to be aware of in this "W3C against of HTTP war" (a crazy
title - but close to the truth):

1. As of recently even fully correct Strict page with DTD linked *and*
served with server Content-Type header "text/html":

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd">

Incorrect. The word "Strict" doesn't belong after the "4.01".

Apr 24 '06 #3

Eric B. Bednarz

"frederick" <fr*******@southernskies.co.uk> writes:

VK wrote:
[more of the same]
Next time you launch a tirade against something, I suggest that you be
a bit more careful to verify your "facts"!
Well, the validator front-end buzzer *is* silly, not to say FUBAR. The
OP is just too lazy to get a clue and construct a test case. :)

<http://validator.w3.org/check?uri=http%3A%2F%2Fbednarz.nl%2Ftmp%2Fcase.htm l&ss=1>
<http://validator.w3.org/check?uri=http%3A%2F%2Fbednarz.nl%2Ftmp%2Fpi.html& ss=1>

So far for bullschildt-heuristics.
It is also fun to compare the erroneous error message for

<http://validator.w3.org/check?uri=http%3A%2F%2Fbednarz.nl%2Ftmp%2Fsys.html &ss=1>

with the fallback action for

<http://validator.w3.org/check?uri=http%3A%2F%2Fbednarz.nl%2Ftmp%2Ffully-tagged.html&ss=1>

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd">

It's hardly surprising that the validator didn't like your doctype:
it's invalid!

Is it? Please explain. The cited document type declaration itself is
hardly 'invalid' (if you set 'override' to 'yes' in your catalog and add
an entry for the public identifier to your local copy of the HTML 4.01
strict DTD -- or anything else, the public and the system identifier of
HTML 4.0 point to different DTDs since years and noboby cares -- you can
even validate the document instance set locally).
--
||| hexadecimal EBB
o-o decimal 3771
--oOo--( )--oOo-- octal 7273
205 goodbye binary 111010111011

Apr 24 '06 #4

Harlan Messinger wrote:

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd">

Incorrect. The word "Strict" doesn't belong after the "4.01".

Uhm... Since when? No later than since last week I humbly presume as I
used this template (if HTML strict was really really requested) for a
year at least - using Validator of course.

And even if Strict became not allowed, what does it have to do with
that spooky crap from Validator? There is "text/html" content type
clearly stated by server, there is DTD linked to check against, so what
about that "I cannot tell is it HTML or XML, so presuming SGML this is
a valid HTML" (??)

Apr 24 '06 #5

Harlan Messinger

VK wrote:

Harlan Messinger wrote:
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd"> Incorrect. The word "Strict" doesn't belong after the "4.01".

Uhm... Since when?

Since as far back as I've been using it (several years), and if you take
five seconds to actually look at the URL in your declaration you'll see
it for yourself.
No later than since last week I humbly presume as I
used this template (if HTML strict was really really requested) for a
year at least - using Validator of course.

And even if Strict became not allowed, what does it have to do with
that spooky crap from Validator? There is "text/html" content type
clearly stated by server, there is DTD linked to check against, so what
about that "I cannot tell is it HTML or XML, so presuming SGML this is
a valid HTML" (??)

Seems clear to me. Your doctype declaration was invalid, so the
validator ignored it. Then it tried its best to figure out on its own
what you were trying to have validated. It looked at your MIME type for
a clue, and found it to be text/html, but since both SGML documents and
XML documents (read: SGML-based HTML documents and XML-based XHTML
documents) can be served as text/html, it still couldn't reach a
definitive conclusion. However, again in an attempt to be helpful, it
decided to *try* SGML-based HTML

In other words, instead of taking the old approach, "This is a bad
doctype, so I'm not going to tell you anything till you fix it or use
the drop-down list to tell me what you think it is I'm looking at," it
tried other strategies to see if it couldn't help you without further
effort on your part. And then, in a magnificent show of ingratitude, you
posted here to complain about it.

Apr 24 '06 #6

Jim Moe

Eric B. Bednarz wrote:

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd">

It's hardly surprising that the validator didn't like your doctype:
it's invalid!

Is it? Please explain.

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">

--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)

Apr 24 '06 #7

Benjamin Niemann

Jim Moe wrote:

Eric B. Bednarz wrote:
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd">

It's hardly surprising that the validator didn't like your doctype:
it's invalid!

Is it? Please explain.

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">

Both doctype declarations are perfectly valid. The first one has an unknown
public identifier (which does not make it invalid), which means that the
validator does not recognize it as a HTML doctype and it fetches the DTD
from the (correct) system identifier. It does not try to recognize HTML vs.
XHTML from the system identifier or from the DTD file itself - I cannot ad
hoc think of a reliable way to implement this.

--
Benjamin Niemann
Email: pink at odahoda dot de
WWW: http://pink.odahoda.de/

Apr 24 '06 #8

Neredbojias

To further the education of mankind, Harlan Messinger
<hm*******************@comcast.net> vouchsafed:

In other words, instead of taking the old approach, "This is a bad
doctype, so I'm not going to tell you anything till you fix it or use
the drop-down list to tell me what you think it is I'm looking at," it
tried other strategies to see if it couldn't help you without further
effort on your part. And then, in a magnificent show of ingratitude, you
posted here to complain about it.

Haha, yes, that's right, but when Microsoft does the same thing,
_everybody_ complains.

The "old approach" is the correct approach.

--
Neredbojias
Infinity has its limits.

Apr 24 '06 #9

Eric B. Bednarz

Jim Moe <jm***************@sohnen-moe.com> writes:

Eric B. Bednarz wrote:
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
"http://www.w3.org/TR/html401/strict.dtd">

It's hardly surprising that the validator didn't like your doctype:
it's invalid!

Is it? Please explain.

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">

That is not an explanation. Since you dropped the supporting prose of
my question, let's investigate this on a chicks-with-pics level then.

<http://bednarz.nl/tmp/valid.png>

<http://bednarz.nl/tmp/invalid.png>
The problem with the W3C validator is that it accepts arbitrary input in
an uncontrolled environment and uses undocumented misfeatures to deal
with that. The latter can work in a local environment, not on the web.

Before an SGML parser can even process the prolog, it needs to know the
SGML declaration. Since there is no formal way of advertising a
document type, one must resolve to voodoo, and the default SGML
declaration and general prolog sniffing heuristics of the W3C validation
service have changed in the past and conveniently are never transparent
for the user. *That* is basically why the OP is scratching her head.
It has nothing to do whatsoever with an 'invalid doctype'. Again, the
above document type declaration *is not invalid*, in the worst case the
system identifier cannot be generated (because there's no catalog entry
for the public identifier and the system identifier cannot be resolved
for whatever reason).

It's not the 'W3C *Markup Validation* Service', it's the '*W3C Markup*
Validation Service'. Which would be fine with me if it was documented.
As it stands, I'd recommend a Wierscher all the same.
--
||| hexadecimal EBB
o-o decimal 3771
--oOo--( )--oOo-- octal 7273
205 goodbye binary 111010111011

Apr 24 '06 #10

Neredbojias

To further the education of mankind, Eric B. Bednarz <bednarz@fahr-zur-
hoelle.org> vouchsafed:

Jim Moe <jm***************@sohnen-moe.com> writes:
Eric B. Bednarz wrote:
It's hardly surprising that the validator didn't like your doctype:
it's invalid!

Is it? Please explain.
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">

That is not an explanation. Since you dropped the supporting prose of
my question, let's investigate this on a chicks-with-pics level then.

<http://bednarz.nl/tmp/valid.png>

<http://bednarz.nl/tmp/invalid.png>

All right, I didn't see one single chick in either of those pics!

..... It's not the 'W3C *Markup Validation* Service', it's the '*W3C Markup*
Validation Service'. Which would be fine with me if it was documented.
As it stands, I'd recommend a Wierscher all the same.

Are you speaking of a hypothetical award or suggesting that the CSE
validator is superior to the w3c's?

--
Neredbojias
Infinity has its limits.

Apr 25 '06 #11

Stephen Poley

On Mon, 24 Apr 2006 17:15:22 -0400, Neredbojias
<http://www.neredbojias.com/fliam.php?cat=alt.html> wrote:

To further the education of mankind, Harlan Messinger
<hm*******************@comcast.net> vouchsafed:
In other words, instead of taking the old approach, "This is a bad
doctype, so I'm not going to tell you anything till you fix it or use
the drop-down list to tell me what you think it is I'm looking at," it
tried other strategies to see if it couldn't help you without further
effort on your part. And then, in a magnificent show of ingratitude, you
posted here to complain about it.

Haha, yes, that's right, but when Microsoft does the same thing,
_everybody_ complains.

Er no - the Microsoft approach would be to take a wild guess and not
tell you anything. Or maybe to supply a validator built in to Windows
which silently changes the document if it finds anything it doesn't
like.

--
Stephen Poley

http://www.xs4all.nl/~sbpoley/webmatters/

Apr 25 '06 #12

Jim Moe

Eric B. Bednarz wrote:

> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
> "http://www.w3.org/TR/html401/strict.dtd">

It's hardly surprising that the validator didn't like your doctype:
it's invalid!

Is it? Please explain.

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">

That is not an explanation. Since you dropped the supporting prose of
my question, let's investigate this on a chicks-with-pics level then.

I ran the test code using the original DTD through the W3C validator. It
validated.
I ran the test code using the modified DTD through the W3C validator. It
validated.
Apparently both forms are acceptable.
I can only presume VK's original attempt contained a typo that did not
make it to his posting (the hazard of not providing an URL), or a
temporary dementia at the W3C validator site, or some other error beyond
our ken.

--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)

Apr 25 '06 #13

Jim Moe

Stephen Poley wrote:

Haha, yes, that's right, but when Microsoft does the same thing,
_everybody_ complains.

Er no - the Microsoft approach would be to take a wild guess and not
tell you anything. Or maybe to supply a validator built in to Windows
which silently changes the document if it finds anything it doesn't
like.

Or claim it a feature, embed it in the operating system, and call it
innovation.

--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)

Apr 25 '06 #14

Neredbojias

To further the education of mankind, Stephen Poley
<sb******************@xs4all.nl> vouchsafed:

On Mon, 24 Apr 2006 17:15:22 -0400, Neredbojias
<http://www.neredbojias.com/fliam.php?cat=alt.html> wrote:
To further the education of mankind, Harlan Messinger
<hm*******************@comcast.net> vouchsafed:
In other words, instead of taking the old approach, "This is a bad
doctype, so I'm not going to tell you anything till you fix it or
use the drop-down list to tell me what you think it is I'm looking
at," it tried other strategies to see if it couldn't help you
without further effort on your part. And then, in a magnificent show
of ingratitude, you posted here to complain about it.

Haha, yes, that's right, but when Microsoft does the same thing,
_everybody_ complains.

Er no - the Microsoft approach would be to take a wild guess and not
tell you anything. Or maybe to supply a validator built in to Windows
which silently changes the document if it finds anything it doesn't
like.

I don't disagree that MS has exhibited calumny in this respect, but my
point is that I'd much rather see (-hear...) that something is broken if
it is broken and not have some dubious piece of software (of any brand)
"patch it up" for me. Of course the term "within reason" applies here,
and if the mistake is noted and accepted in a logical manner, that would
probably suffice.

--
Neredbojias
Infinity has its limits.

Apr 25 '06 #15

David Dorward

Harlan Messinger wrote:

Seems clear to me. Your doctype declaration was invalid, so the
validator ignored it.
It wasn't invalid, it had an unrecognised public identifier.
Then it tried its best to figure out on its own
what you were trying to have validated.
So the validator worked from the DTD at the URL specified at the URL in the
system identifier.
It looked at your MIME type for
a clue, and found it to be text/html, but since both SGML documents and
XML documents (read: SGML-based HTML documents and XML-based XHTML
documents) can be served as text/html,
Hate Appendix C, hate hate hate hate hate. :)
it still couldn't reach a definitive conclusion. However, again in an
attempt to be helpful, it decided to *try* SGML-based HTML

Which is a fairly sensible default as there shouldn't be any non-XHTML 1.0
DTDs floating around needing XML parsing.

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

Apr 25 '06 #16

Jim Moe wrote:

>> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"
>> "http://www.w3.org/TR/html401/strict.dtd"> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">

I ran the test code using the original DTD through the W3C validator. It
validated.
I ran the test code using the modified DTD through the W3C validator. It
validated.
Apparently both forms are acceptable.
I can only presume VK's original attempt contained a typo that did not
make it to his posting (the hazard of not providing an URL), or a
temporary dementia at the W3C validator site, or some other error beyond
our ken.

"4.01 Strict//EN" validates (green banner) but gives you a warning atop
I quoted in the OP. It also doesn't propose you to use "Valid HTML"
banner.

"4.01//EN" validates (green banner), no warning, proposes to use "Valid
HTML" banner.

Presuming I'm a nuts and it's "always" (since 1999) was this way and
all cases of "4.01 Strict//EN" are just games of my "beautiful mind"
:-( :-)

W3C's own documentation is shinny clear about that: "Oceania always
faught with Eastasia, it never faught with Eurasia" :-) I just found
some wague mentions of a patch dated 2002
<http://lists.w3.org/Archives/Public/www-validator/2002Mar/0094.html>

MSDN still remembers something about "4.01 Strict//EN":
<http://msdn.microsoft.com/workshop/author/dhtml/reference/objects/DOCTYPE.asp>
<http://msdn.microsoft.com/library/en-us/dnie60/html/cssenhancements.asp>

But of course "what Micro$oft may possibly know about standards!" :-)

Any way, it is fine by me to use "4.01//EN" without "Strict". No big
deal. Still kind of stupid to validate against of a linked DTD and
clame at the end that you have no clue that kind of document was
that... IMHO.

Apr 25 '06 #17

Nick Kew

David Dorward wrote:

Harlan Messinger wrote:

Seems clear to me. Your doctype declaration was invalid, so the
validator ignored it.

It wasn't invalid, it had an unrecognised public identifier.

Yep. No problem with the substance of that. But such a big,
prominent warning is misleading: it looks like complaining of
an error (which it isn't - though in this particular instance
it *is* something the user should correct).

Then it tried its best to figure out on its own
what you were trying to have validated.

So the validator worked from the DTD at the URL specified at the URL in the
system identifier.

Indeedie.

It looked at your MIME type for
a clue, and found it to be text/html, but since both SGML documents and
XML documents (read: SGML-based HTML documents and XML-based XHTML
documents) can be served as text/html,

Hate Appendix C, hate hate hate hate hate. :)

Zigackly.

it still couldn't reach a definitive conclusion. However, again in an
attempt to be helpful, it decided to *try* SGML-based HTML

Which is a fairly sensible default as there shouldn't be any non-XHTML 1.0
DTDs floating around needing XML parsing.

Yes, it should parse as SGML.

The circumstances in which a warning message is appropriate is when the
document is in fact XML (e.g. someone declared XHTML1.1 and served
text/html). In that case it will *correctly* parse as SGML, and almost
certainly generate a bunch of errors, even if the document is valid
XHTML. That's very confusing to the user, and a warning about what's
going on is the lesser evil.

The W3C validator's problem is that the warning is itself so prominent
as to cause confusion. A simple system message saying "Unrecognised
public identifier foo" would be entirely appropriate, and *if* the
document fails validation, a brief note/link about the reason for and
effect of parsing XML as SGML.

--
Nick Kew

Apr 25 '06 #18

Jack

Jim Moe wrote:

I can only presume VK's original attempt contained a typo that did
not make it to his posting (the hazard of not providing an URL), or a
temporary dementia at the W3C validator site, or some other error
beyond our ken.

The OP's URLs lacked a 'c', as in 'w3c.org'. Someone else drew attention
to this fact, but not very loudly.

--
Jack.

Apr 25 '06 #19

Jack wrote:

The OP's URLs lacked a 'c', as in 'w3c.org'. Someone else drew attention
to this fact, but not very loudly.

?

The official W3C site is <http://www.w3.org>, not w3c.org

w3c.org is just a forwarder for typos.

Apr 25 '06 #20

Jack

VK wrote:

Jack wrote:
The OP's URLs lacked a 'c', as in 'w3c.org'. Someone else drew
attention to this fact, but not very loudly.

?

The official W3C site is <http://www.w3.org>, not w3c.org

w3c.org is just a forwarder for typos.

Oh! You live and learn.

--
Jack.

Apr 25 '06 #21

Harlan Messinger

Neredbojias wrote:

To further the education of mankind, Harlan Messinger
<hm*******************@comcast.net> vouchsafed:
In other words, instead of taking the old approach, "This is a bad
doctype, so I'm not going to tell you anything till you fix it or use
the drop-down list to tell me what you think it is I'm looking at," it
tried other strategies to see if it couldn't help you without further
effort on your part. And then, in a magnificent show of ingratitude, you
posted here to complain about it.
Haha, yes, that's right, but when Microsoft does the same thing,
_everybody_ complains.

The problem with Microsoft is that it *changes* what you've done. The
W3C validator isn't changing anything. All it's doing is attempting to
save you a step in getting information.

The "old approach" is the correct approach.

Apr 25 '06 #22

Harlan Messinger

Neredbojias wrote:

To further the education of mankind, Stephen Poley
<sb******************@xs4all.nl> vouchsafed:
On Mon, 24 Apr 2006 17:15:22 -0400, Neredbojias
<http://www.neredbojias.com/fliam.php?cat=alt.html> wrote:
To further the education of mankind, Harlan Messinger
<hm*******************@comcast.net> vouchsafed:

In other words, instead of taking the old approach, "This is a bad
doctype, so I'm not going to tell you anything till you fix it or
use the drop-down list to tell me what you think it is I'm looking
at," it tried other strategies to see if it couldn't help you
without further effort on your part. And then, in a magnificent show
of ingratitude, you posted here to complain about it.
Haha, yes, that's right, but when Microsoft does the same thing,
_everybody_ complains. Er no - the Microsoft approach would be to take a wild guess and not
tell you anything. Or maybe to supply a validator built in to Windows
which silently changes the document if it finds anything it doesn't
like.

I don't disagree that MS has exhibited calumny in this respect, but my
point is that I'd much rather see (-hear...) that something is broken if
it is broken and not have some dubious piece of software (of any brand)
"patch it up" for me.

The W3C validator *is* telling you that your doctype is broken. At
least, it's implying as much by telling you that the one you've declared
can't be used to identify the document's type.
Of course the term "within reason" applies here,
and if the mistake is noted and accepted in a logical manner, that would
probably suffice.

Apr 25 '06 #23

Harlan Messinger

David Dorward wrote:

Harlan Messinger wrote:
Seems clear to me. Your doctype declaration was invalid, so the
validator ignored it.

It wasn't invalid, it had an unrecognised public identifier.

Touché.

Apr 25 '06 #24

David Dorward

VK wrote:

"4.01 Strict//EN" validates (green banner) but gives you a warning atop
I quoted in the OP. It also doesn't propose you to use "Valid HTML"
banner.
No, its some Doctype with an unrecognised public identifier (but working
system identifier). Since it keys off the public identifier it doesn't
recognise it as being HTML 4.01.
"4.01//EN" validates (green banner), no warning, proposes to use "Valid
HTML" banner.

Presuming I'm a nuts and it's "always" (since 1999) was this way and
all cases of "4.01 Strict//EN" are just games of my "beautiful mind"
Correct.

Note however that the XHTML 1.0 Strict DTD _does_ have the word Strict in
the public identifier.
Any way, it is fine by me to use "4.01//EN" without "Strict". No big
deal. Still kind of stupid to validate against of a linked DTD and
clame at the end that you have no clue that kind of document was
that... IMHO.

It doesn't though. It makes the claim at the beginning. To paraphrase: I
don't recognise your public identifier and since the HTML Working Group
came up with Appendix C a few years ago, I can't be sure that the document
is SGML, but I'll assume it is for now, so if there are errors, then they
might be due to the DTD being an XML DTD.

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

Apr 25 '06 #25

Neredbojias

To further the education of mankind, Harlan Messinger
<hm*******************@comcast.net> vouchsafed:

Neredbojias wrote:
To further the education of mankind, Harlan Messinger
<hm*******************@comcast.net> vouchsafed:
In other words, instead of taking the old approach, "This is a bad
doctype, so I'm not going to tell you anything till you fix it or
use the drop-down list to tell me what you think it is I'm looking
at," it tried other strategies to see if it couldn't help you
without further effort on your part. And then, in a magnificent show
of ingratitude, you posted here to complain about it.

Haha, yes, that's right, but when Microsoft does the same thing,
_everybody_ complains.

The problem with Microsoft is that it *changes* what you've done. The
W3C validator isn't changing anything. All it's doing is attempting to
save you a step in getting information.

Perhaps, but it seems to me that an inexperienced user could easily think
(misunderstand) that there is not a real problem.

--
Neredbojias
Infinity has its limits.

Apr 25 '06 #26

Neredbojias

To further the education of mankind, Harlan Messinger
<hm*******************@comcast.net> vouchsafed:

I don't disagree that MS has exhibited calumny in this respect, but my
point is that I'd much rather see (-hear...) that something is broken if
it is broken and not have some dubious piece of software (of any brand)
"patch it up" for me.

The W3C validator *is* telling you that your doctype is broken. At
least, it's implying as much by telling you that the one you've declared
can't be used to identify the document's type.

And that is my argument, the difference between "implying" and "telling".
Implications are hardly robust and seldom succinct as they should be in
such a venue.

--
Neredbojias
Infinity has its limits.

Apr 25 '06 #27

Robert Baer

-------- SNIPped as following case shows a possible problem.
This HTML code exactly as shown *passes*; isn't there something
incorrect about the useage of "style" after the H3?

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<title>Oregon well data</title>
<style type="text/css" >
</style>
</head>

<body>


<div style="position: absolute; height: 432px; width: 576px; top: 30px;
left: 100px; " >
<img src="ORwells.gif" alt="" usemap="#OR" style="border-style:none" >
</div>

<H3>
< style="position: absolute; top: 1px; left: 300px; " ><i>© 2006
Oil 4 Less LLC</i>
</H3>

</body>
</html>

Apr 27 '06 #28

Eric B. Bednarz

Robert Baer <ro********@earthlink.net> writes:

This HTML code exactly as shown *passes*; isn't there something
incorrect about the useage of "style" after the H3?
[...]
<H3>
< style="position: absolute; top: 1px; left: 300px; " ><i>© 2006
Oil 4 Less LLC</i>
</H3>

No. It's just character data up to the start tag of the I element.
--
||| hexadecimal EBB
o-o decimal 3771
--oOo--( )--oOo-- octal 7273
205 goodbye binary 111010111011

Apr 27 '06 #29

David Dorward

Robert Baer wrote:

This HTML code exactly as shown *passes*; isn't there something
incorrect about the useage of "style" after the H3? < style="position: absolute; top: 1px; left: 300px; " >

Since the less than character is followed by a space, the construct isn't an
opening tag.

Your code is equivalent to:

< style="position: absolute; top: 1px; left: 300px; " >

Firefox renders this correctly. The W3C Markup Validator parses this
correctly (as can be seen by turning the Parse Tree mode on). I haven't
tested in any other user agents.
--
David Dorward <http://blog.dorward.me.uk/> <http://dorward.me.uk/>
Home is where the ~/.bashrc is

Apr 27 '06 #30

Similar topics