Local server HTML validator
Torbjørn Pettersen
HTML code. The way I do it is simply more or less redoing to
complete site, testing it on a web server I have set up on my
local network.
I have downloaded, and installed CSE HTML Validator Pro,
but I don't get the same results with that as I do with the online
validator on W3.org. And I can't upload files to W3.org either,
due to all the ASP code I use.
So, the question is, where can I find a validator that validates
like the one on W3.org? (Windows 2000 server).
--
Torbjørn Pettersen
Editor/Webmaster
FantaFiction
Spartanicus
>I have downloaded, and installed CSE HTML Validator Pro,
>but I don't get the same results with that as I do with the online
>validator on W3.org.
CSE is not a validator, despite it's name, it's a linter.
>So, the question is, where can I find a validator that validates
>like the one on W3.org? (Windows 2000 server).
--
Spartanicus
Brian
> where can I find a validator that validates like the one on W3.org?
At w3.org?
http://validator.w3.org/source/
--
Brian (remove "invalid" from my address to email me)
http://www.tsmchughs.com/
Torbjørn Pettersen
> >I have downloaded, and installed CSE HTML Validator Pro,
> >but I don't get the same results with that as I do with the online
> >validator on W3.org.
>
> CSE is not a validator, despite it's name, it's a linter.
A linter? What's that?
> >So, the question is, where can I find a validator that validates
> >like the one on W3.org? (Windows 2000 server).
>
> http://arealvalidator.com/
Thanks. Testing it now, looks good so far. Except that I'm getting
more work. ;-)
Torbjørn Pettersen
> > where can I find a validator that validates like the one on W3.org?
>
> At w3.org?
>
> http://validator.w3.org/source/
I already looked at that page, and half way down it my head started
spinning. This Friday I think I'll skip the beer, and just read this a
couple of times. ;-)
If it doesn't have a "download the ready to install here" button, I'm stuck. :-/
Darin McGrew
There's an explanation in the FAQ:
http://www.htmlhelp.com/faq/html/basics.html#error-check
--
Darin McGrew, mcg...@stanfordalumni.org, http://www.rahul.net/mcgrew/
Web Design Group, da...@htmlhelp.com, http://www.HTMLHelp.com/
"If you aren't part of the solution, then you are part of the precipitate."
Saqib Ali
> spinning.
I know what you mean. The first time I installed w3c locally on my
server, I was pulling out hair. I didn't wanna go through the hassle
of installing on other intranet servers, so I just wrote my own PHP
based validator. Plus, in addition to XHTML/HTML I also wanted to
validate DocBook XML, so writing my own validator made more sense. And
it works fine, and much easier to install on new servers. :).
In Peace,
Saqib Ali
http://validate.sf.net
Brian
> I just wrote my own PHP based validator. Plus, in addition to
> XHTML/HTML
And is this released to the public? :-)
Nick Kew
sa...@stonebeat.org (Saqib Ali) writes:
> of installing on other intranet servers, so I just wrote my own PHP
> based validator.
Isn't that just slightly revisionist? Nothing against your validator,
but that's not quite how you were describing your motivation at the time.
> Plus, in addition to XHTML/HTML I also wanted to
> validate DocBook XML, so writing my own validator made more sense. And
> it works fine, and much easier to install on new servers. :).
Your validator, in common with other HTML validators, uses OpenSP,
which has known limitations with XML. mod_validator, from
<URL:http://apache.webthing.com/>, has the edge in terms of supporting
different markup types, though it's not packaged for simple-install.
But for a single user, a desktop validator is likely to make more sense
than a webserver-based one. That leaves arealvalidator for windows,
or validator-lite for other platforms, as someone already said.
--
Nick Kew
Nick's manifesto: http://www.htmlhelp.com/~nick/
Saqib Ali
> Saqib Ali wrote:
>
> > I just wrote my own PHP based validator. Plus, in addition to
> > XHTML/HTML
>
> And is this released to the public? :-)
Here is the SourceForge webpage for the validator:
http://sourceforge.net/projects/validate/
The first release is coming soon.
You test-drive the validator @ http://validate.sf.net
Saqib Ali
> than a webserver-based one. That leaves arealvalidator for windows,
> or validator-lite for other platforms, as someone already said.
hmm, for desktop validator, I think OpenSP or XSLTproc is better
option. They both are free, easy-to-use and included with all Linux
distros. I think it is included with MacOS X as well. You can d/l it
for free for windows.
Saqib Ali
> but that's not quite how you were describing your motivation at the time.
hmm. i m not sure what you mean by this. Originally I didn't intend to
make a validate DocBook XMl content, but the motivation was to create
a validator that is simple to install, in 2 steps.
1) Install OpenSP. (which is no longer required on Linux, since most
Linux distro come with OpenSP).
2) GZip the PHP files.
And start validating your (X)HTML.
Later on I decided to include DocBook XML, and also added the ability
to convert DocBook XML to HTML/PDF.
Brian
> for a single user, a desktop validator is likely to make more sense
> than a webserver-based one.
I don't know, I'd like to have a way to install a validator on my
localhost-only Apache (which I use to test my sites). That would allow
me to more quickly check my work. I use Windows. I could try to download
the w3c code, and get all the components, but I haven't made time for
that yet. But if something simpler were to come along... :-)
Albert Wiersch
CSE HTML Validator doesn't use the same engine as the W3 validator so the
results won't be the same. You might find this interesting, though:
http://www.htmlvalidator.com/htmlval/whycseisbetter.html
--
Albert Wiersch
http://www.htmlvalidator.com/
"Torbjørn Pettersen" <tpe AT broadpark DOT no> wrote in message
news:407c...@news.broadpark.no...
Brian
http://allmyfaqs.com/faq.pl?How_to_post
Albert Wiersch wrote:
> CSE HTML Validator doesn't use the same engine as the W3 validator so
> the results won't be the same. You might find this interesting,
> though: http://www.htmlvalidator.com/htmlval/whycseisbetter.html
You might find this more interesting:
http://validator.w3.org/check?uri=http://www.htmlvalidator.com/htmlval/whycseisbetter.html
Physician, heal thyself!
Jukka K. Korpela
wrote:
> CSE HTML Validator doesn't use the same engine as the W3 validator so
> the results won't be the same.
As we have repeatedly pointed out, it is not a validator at all, and you
have read that and you have decided to keep advertizing it as a
validator. You have now reached a new level of bogosity with the babble
about "engine".
Validators do much less good than many people think and say. But surely a
commercial product that _has_ to be sold under a false name of a
validator is even less interesting.
--
Yucca, http://www.cs.tut.fi/~jkorpela/
Pages about Web authoring: http://www.cs.tut.fi/~jkorpela/www.html
Nick Kew
sa...@stonebeat.org (Saqib Ali) writes:
>> But for a single user, a desktop validator is likely to make more sense
>> than a webserver-based one. That leaves arealvalidator for windows,
>> or validator-lite for other platforms, as someone already said.
>
> hmm, for desktop validator, I think OpenSP or XSLTproc is better
> option.
Those are commandline tools. validator-lite is just a GTK-based GUI
frontend to OpenSP, that takes care of the complexities that make it
intimidating to many users. And it's GPL (free) and easy-to-use.
> They both are free, easy-to-use and included with all Linux
> distros.
I think a lot of people find OpenSP challenging. Managing the SGMLDECL,
catalogue files and lib is not a simple task even for the computer-
literate unless they are working from a package such as one of the
validator distributions.
Nick Kew
Brian <use...@julietremblay.com.invalid> writes:
> I don't know, I'd like to have a way to install a validator on my
> localhost-only Apache (which I use to test my sites).
Fair enough - your choice.
> I use Windows.
Compiling OpenSP for windows is not for the faint-hearted, though it's now
been done. Hmmm, now that I think about it, there's a project now at
sourcefirge to add a validation bar to MSIE using a locally-installed
OpenSP. Check Bjoern Hoehrmann's posts to www-validator over recent months.
> I could try to download
> the w3c code, and get all the components, but I haven't made time for
> that yet. But if something simpler were to come along... :-)
Take your pick: w3c validator (Perl/CGI), WDG validator (Perl/CGI),
mod_validator (Apache module), or Saqib's PHP. One of these days
I'll grab Bjoern's OpenSP binary for windows and try building
mod_validator against it. That'll then be both the most powerful
and most efficient server-based validator, as it already is for
Unix/Linux.
Albert Wiersch
"Brian" <use...@julietremblay.com.invalid> wrote in message
> You might find this more interesting:
>
>
http://validator.w3.org/check?uri=http://www.htmlvalidator.com/htmlval/whycseisbetter.html
>
There's nothing wrong with that as I've chosen to use some non-standard
markup. Some developers like to do that, while some don't. Unfortunately, if
you do like to do that, then the W3C validator may not be as useful.
I do try to use standard markup as much as possible, but sometimes there is
a reason to use something other than standard markup.
Albert Wiersch
"Jukka K. Korpela" <jkor...@cs.tut.fi> wrote in message
news:Xns94CC1A41BBCF...@193.229.0.31...
> "Albert Wiersch" <mrinternetREMVETHIS...@wiersch.com>
> wrote:
>
> > CSE HTML Validator doesn't use the same engine as the W3 validator so
> > the results won't be the same.
>
> As we have repeatedly pointed out, it is not a validator at all, and you
> have read that and you have decided to keep advertizing it as a
> validator. You have now reached a new level of bogosity with the babble
> about "engine".
Yes, it has been repeatedly pointed out that CSE HTML Validator is not a
validator in the strict, technical sense of the word. I never said that it
was. It is a validator in the common meaning of the word (what most people
think of when they think HTML Validator). I'm not going to continue to argue
about this, but it's important that people understand this. The program is
not marketed under a false name. Never has been, never will be. Look up
"validate" in a dictionary and you'll see that there's more than one
definition.
> Validators do much less good than many people think and say.
It's interesting that you say this. That's why I created CSE HTML
Validator -- to be more useful and do more good than "real" ones like the
W3C validator, since it isn't limited to validating only to formal standards
that can be expressed in DTDs. So it's great if that's what you want and you
want to limit yourself to that. But if you want more, the W3C validator
can't give it to you.
Anyway, use what you want, but I don't think it's appropriate to trash a
great product just because you don't like the name.
C A Upsdell
in message news:ts6dndiicpO...@august.net...
> Yes, it has been repeatedly pointed out that CSE HTML Validator is not a
> validator in the strict, technical sense of the word. I never said that it
> was. It is a validator in the common meaning of the word (what most people
> think of when they think HTML Validator).
You could avoided this contentious issue if you had called your product the
CSE HTML Code Checker, or CSE HTML Lint program, or CSE HTML QC Toolkit, or
something more consistent with other, similar tools. Or if you had built a
real validator into your product, which users could use either separately or
in combination with your other code checks. Indeed, adding a true validator
to your product would both enhance its functionality and forever end this
kind of dispute.
Saqib Ali
I understand the argument about empty TAGs (e.g. </p>), however empty
tags is well-formed according to XML standards.
A Validator, by definition, should only check for well-formedness and
validity. Anything beyond that, is not the function of the validator.
I really like CSE HTML Validator, but in real-life lot of sites use
empty tags, so CSE HTML validator generates tons of warning messages.
Just my .2 cents.
In Peace,
Saqib Ali
http://validate.sf.net
"Albert Wiersch" <mrinternetREMVETHIS...@wiersch.com> wrote in message news:<Lb2dnVLExu8...@august.net>...
David Dorward
> There's nothing wrong with that as I've chosen to use some non-standard
> markup.
Why does the document claim that you conform the HTML 4.01 Transitional DTD
then? Why not use a DTD that includes your non-standard markup?
> Some developers like to do that, while some don't. Unfortunately,
> if you do like to do that, then the W3C validator may not be as useful.
Gosh! The validator might not be useful if you don't care if your code is
valid or not! What an insight!
> I do try to use standard markup as much as possible, but sometimes there
> is a reason to use something other than standard markup.
What's the reason for this then? The valign attribute is well supported when
applied to table data cells and table heading cells, and the vertical-align
property will allow you to:
.myTable * { vertical-align: middle; }
Not to mention that valign='middle' is the default anyway!
--
David Dorward <http://blog.dorward.me.uk/> <http://dorward.me.uk/>
Albert Wiersch
"Saqib Ali" <sa...@stonebeat.org> wrote in message
news:d22b4f0d.0404...@posting.google.com...
> Albert,
>
> I really like CSE HTML Validator, but in real-life lot of sites use
> empty tags, so CSE HTML validator generates tons of warning messages.
What are the warning messages you are referring to? CSE should be happy with
empty tags when used properly.
Thanks,
Albert
Albert Wiersch
"David Dorward" <dor...@yahoo.com> wrote in message
news:c5lap6$gpu$4$8302...@news.demon.co.uk...
> Albert Wiersch wrote:
>
> > There's nothing wrong with that as I've chosen to use some non-standard
> > markup.
>
> Why does the document claim that you conform the HTML 4.01 Transitional
DTD
> then? Why not use a DTD that includes your non-standard markup?
Because I start with the closest standard DTD and add what I think is useful
from there. Do you really think that browsers care if I add an extra
attribute that they don't understand for the benefit of those using a
certain browser?
> What's the reason for this then?
I do not remember the reason now. I put it in there years ago and there was
most likely a good reason. Are you willing to say that that attribute
doesn't have any effect in all the browsers and versions ever written? I
doubt it. Now I may very well decide to remove it in the future, but it was
put in for a reason. And neither can one say having it hurt anything, other
than an extra millisecond or so of page load time.
Albert
Saqib Ali
> validator in the strict, technical sense of the word. I never said that it
> was. It is a validator in the common meaning of the word (what most people
> think of when they think HTML Validator). I'm not going to continue to argue
> about this, but it's important that people understand this. The program is
> not marketed under a false name. Never has been, never will be. Look up
> "validate" in a dictionary and you'll see that there's more than one
> definition.
definitions are usually taken in context - The dictionay also says
friendship is a synonym for intimacy. That doesn't mean I am intimate
with Linus. :)
Getting back to XML:
A valid document conforms to the XML syntax rules and follows the
guidelines of a Document Type Definition (DTD), or in other word: A
valid XML document is well-formed and conforms to a XML DTD.
The process of comparing the XML document to the DTD is called
validation, which is performed using a validation parser (aka
validator).
Alan J. Flavell
> Yes, it has been repeatedly pointed out that CSE HTML Validator is not a
> validator in the strict, technical sense of the word. I never said that it
> was.
Eh? Which parts of http://www.w3.org/TR/html401/conform.html
HTML document
An HTML document is an SGML document that meets the constraints of
this specification.
are causing comprehension difficulties here? The term "validator" has
a definite meaning in relation to SGML, it's not open to
re-negotiation.
> It is a validator in the common meaning of the word
Fine. In other words, it is not an "HTML Validator".
> (what most people think of when they think HTML Validator).
In what way is the opinion of "most people" relevant, when the
question is about a clearly-defined technical term?
> I'm not going to continue to argue about this,
I don't see that there's anything to argue about. The term "HTML
Validator" has a precise meaning. The product is not an "HTML
Validator". You've confirmed this yourself. The only remaining
problem is that - as you've rightly identified - the false claim fools
"most people".
Jukka K. Korpela
wrote:
>> Why does the document claim that you conform the HTML 4.01
>> Transitional DTD then? Why not use a DTD that includes your
>> non-standard markup?
>
> Because I start with the closest standard DTD and add what I think is
> useful from there.
So you apparently do not know how to work with DTDs, in order to create a
customized DTD.
It is not surprising that when advertizing your phoney "validator", you
indicate lack of knowledge of DTDs (i.e., the very thing that real
validators work with).
Apparently you don't understand that every time to post your
advertisement (which, along with poor excuses for it, seems to be the
only thing you post here), you make the "CSE HTML Validator" look more
and more ridiculous. So your advertisements have a considerable impact,
just a negative one.
Albert Wiersch
>
> So you apparently do not know how to work with DTDs, in order to create a
> customized DTD.
I do not normally work with DTDs but I do know that having a customized DTD
isn't going to do me any good. What browser is going to work better with a
customized DTD? What person viewing my site will have a better experience
because I used a customized DTD? It's better to use something standard in my
opinion (with, of course, the exceptions where something proprietary makes
sense).
David Dorward
> I do not remember the reason now. I put it in there years ago and there
> was most likely a good reason. Are you willing to say that that attribute
> doesn't have any effect in all the browsers and versions ever written?
I'm willing to bet that it doesn't have any effect that can't be achieved
using one of the three methods I described that don't deviate from the
standard.
Saqib Ali
> empty tags when used properly.
Try this URL
http://www.htmlhelp.com/cgi-bin/validate.cgi?url=http%3A%2F%2Fwww.xml-dev.com%2Fblog%2Ftest.php&warnings=yes
Jukka K. Korpela
wrote:
> I do not normally work with DTDs but I do know that having a
> customized DTD isn't going to do me any good.
You already told us you don't understand what validation is, or you
deliberately misrepresent your commercial product as a validator. (This
is an "inclusive or".)
> It's better to use something standard in my opinion
So do you think that claiming your document to comply with HTML 4.01
syntax when it actually doesn't is "using something standard"?
There might be good reasons to lie about the DTD to browsers, but you
apparently don't know them.
Steve Pugh
wrote:
>"Jukka K. Korpela" <jkor...@cs.tut.fi> wrote in message
>news:Xns94CCC15E8FEE...@193.229.0.31...
>>
>> So you apparently do not know how to work with DTDs, in order to create a
>> customized DTD.
>
>I do not normally work with DTDs but I do know that having a customized DTD
>isn't going to do me any good.
Really?
1. Your code will no longer to telling lies by claiming to be
something that it is not.
2. You'll be able to take advantage of the benefits available via
validation.
3. Anyone looking at your code will be able to look up your DTD and
see how you have defined the extensions you are using. If the DTD is
commented or contains the URL of full documentation they can also read
what you think the extensions should do in supporting browsers.
>What browser is going to work better with a customized DTD?
Depends. Customs DTDs tend to trigger Standards mode in modern
browsers so the literal answer to the above question is "quite a lot
of them".
Of course as you're using non-standard code the chances are that you
want Quirks mode anyway (some of the non-standard attributes are
ignored in Standards mode), so in actual fact you want browsers to
work worse not better. ;-)
>What person viewing my site will have a better experience
>because I used a customized DTD?
What person viewing your site will have a worse experience if you use
no doctype at all?
>It's better to use something standard in my opinion
But you're not...
>(with, of course, the exceptions where something proprietary makes
>sense).
At which point it's no longer standard so you should stop claiming
that it is. If you don't write standard HTML then don't use a standard
doctype declaration - either use one that references a custom DTD or
don't use one at all. These two options give you Standards and Quirks
mode respectively.
Steve
--
"My theories appal you, my heresies outrage you,
I never answer letters and you don't like my tie." - The Doctor
Steve Pugh <st...@pugh.net> <http://steve.pugh.net/>
Albert Wiersch
>
> So do you think that claiming your document to comply with HTML 4.01
> syntax when it actually doesn't is "using something standard"?
Yes. Just because there's one or two elements that aren't standard, doesn't
mean that no standards were used. I used plenty of standards in my pages.
Albert Wiersch
"Steve Pugh" <st...@pugh.net> wrote in message
news:jpjt70tp02lg3l38p...@4ax.com...
>
> 1. Your code will no longer to telling lies by claiming to be
> something that it is not.
I'm not a purist. I do not care if I "tell a lie" by using a transitional
DTD and then using one or two proprietary pieces of markup. What matters is
that people are able to benfit from the information on my web site. That's
what I believe should be the concern of a webmaster, not whether they are
telling a lie because their document isn't 100% valid according to a DTD.
> 2. You'll be able to take advantage of the benefits available via
> validation.
I can take plenty of advantage of validators using the benefits of CSE HTML
Validator. And I can still use the W3C validator if I want to, ignoring the
error that it generates for proprietary markup.
> 3. Anyone looking at your code will be able to look up your DTD and
> see how you have defined the extensions you are using. If the DTD is
> commented or contains the URL of full documentation they can also read
> what you think the extensions should do in supporting browsers.
And how many people do you suppose are going to look at someone's pages like
that? If I wanted to do such a thing for the .001% of people who do, then I
would. I have better things to do, however.
>
> >What browser is going to work better with a customized DTD?
>
> Depends. Customs DTDs tend to trigger Standards mode in modern
> browsers so the literal answer to the above question is "quite a lot
> of them".
The real issue here is whether standards mode is on or not, not whether the
DTD is custom or not, since whether the browser is in standard or quirk mode
affects how the page is rendered. A custom DTD is not needed to control
quirks mode. I can use a standard DTD.
> At which point it's no longer standard so you should stop claiming
> that it is. If you don't write standard HTML then don't use a standard
> doctype declaration - either use one that references a custom DTD or
> don't use one at all. These two options give you Standards and Quirks
> mode respectively.
As an author, I decide what I think is best and write my pages accordingly.
You can choose how you want to write your pages. Very few people care about
writing 100% technically compliant pages. What is important is that the
viewer be able to benefit from the information on the web page, not that
it's 100% compliant with technical standards. My belief is that it is more
important to please people than to please a validator. Of course validators
are useful, but I target my pages primarily toward people and the browsers
that show the pages to people, not validators.
Albert Wiersch
> I'm willing to bet that it doesn't have any effect that can't be achieved
> using one of the three methods I described that don't deviate from the
> standard.
That may be true, but it might not be. The "valign" attribute for the
"table" element probably has an effect on an older browser, which is why I
put it in there a few years ago. The methods you mention may or may not have
the same effect in that browser. There are so many browsers and versions of
browsers available, and everyone knows that don't all behave the same way.
But it could probably safely be removed now since most people use updated
browsers now where that attribute doesn't have any effect... so I may do
that.
Albert Wiersch
>
> Try this URL
>
http://www.htmlhelp.com/cgi-bin/validate.cgi?url=http%3A%2F%2Fwww.xml-dev.com%2Fblog%2Ftest.php&warnings=yes
That's not CSE HTML Validator generating those warnings. CSE has no problem
with <p/>.
Jukka K. Korpela
wrote:
> I do not care if I "tell a lie"
We regulars know that, but thanks for making it clear to all of us.
Albert Wiersch
> "Albert Wiersch" <mrinternetREMVETHIS...@wiersch.com>
> wrote:
>
> > I do not care if I "tell a lie"
>
> We regulars know that, but thanks for making it clear to all of us.
You know, a lie is taking something out of context for the purpose of
deliberately deceiving and/or giving a wrong impression. You just did this
when you selectively quoted a few words from my message--completely out of
context.
Alan J. Flavell
> You know, a lie is taking something out of context for the purpose
> of deliberately deceiving and/or giving a wrong impression. You just
> did this when you selectively quoted a few words from my
> message--completely out of context.
Quite a number of us are following the context closely, (not all of
whom are actually posting to this thread, since some of them have
already noticed that it's pointless), and are now in no doubt about
your position. When you find yourself down a hole, it's said to be
wise to stop digging. I sure don't recommend you calling for a
mechanical excavator to help dig faster.
James Pickering
>Quite a number of us are following the context closely, (not all of
>whom are actually posting to this thread, since some of them have
>already noticed that it's pointless), and are now in no doubt about
>your position ..........
That certainly goes for me.
James Pickering
http://www.jp29.org/cadr.htm
Jan Roland Eriksson
>"Albert Wiersch" <mrinternetREMVETHIS...@wiersch.com>
>wrote:
>>"Jukka K. Korpela" <jkor...@cs.tut.fi> wrote in message
>>news:Xns94CCC15E8FEE...@193.229.0.31...
>>>So you apparently do not know how to work with DTDs,
>>>in order to create a customized DTD.
>>I do not normally work with DTDs but I do know that having
>>a customized DTD isn't going to do me any good.
>Really?
[...]
>3. Anyone looking at your code will be able to look up your
>DTD and see how you have defined the extensions you are using.
>If the DTD is commented or contains the URL of full documentation...
From a formal SGML standpoint, we need to correct this statement :-)
Comments in a "Document Type Definition" entity is not allowed to be
read as "documentation" of semantics since the DTDefinition itself is
only about markup syntax in the first place.
Also the inclusion of a link, to the governing "Formal Prose" part of an
application of SGML, is questionable. The SGML Declaration (if any) and
the DTDefinition comes as a result of the "Formal Prose" definition for
an application of SGML so in effect we should look at this as "traveling
down a one way street".
Formal Prose definition comes first, if that one calls for
changes/additions to the SGML Reference Concrete Syntax -- write an SGML
declaration to the fact. Finally, and still based on the prose, write a
DTDefinition that expresses the syntactical rules for the markup that
conforms to the prose. The rest of it, i.e. the semantic definitions for
that markup, will still only be available in the "Formal Prose" part.
Only humans can think and understand, computers don't think, they just
do what they are told to do, be that right or wrong, they will still do
it :-)
>>What browser is going to work better with a customized DTD?
>Depends. Customs DTDs tend to trigger Standards mode in modern
>browsers so the literal answer to the above question is "quite a lot
>of them".
You changed subject; now you are talking about a "Document Type
Declaration". That's only a reference to some entity who's absolute
location may, or may not, be possible to find.
DOCTYPE-Switching in browsers, based on DTDeclarations, has nothing to
do with the possible existence of a DTDefinition that matches any
DTDeclaration in use.
--
Rex
Stan Brown
>"Brian" <use...@julietremblay.com.invalid> wrote in message
>news:107rigq...@corp.supernews.com...
>
>There's nothing wrong with that as I've chosen to use some non-standard
>markup. Some developers like to do that, while some don't. Unfortunately, if
>you do like to do that, then the W3C validator may not be as useful.
But if you're knowingly writing invalid HTML, doesn't the whole idea
of "validating" it go out the window?
It's like asking someone to check your arithmetic but insisting they
they should consider 2x3 to be 5.
--
Stan Brown, Oak Road Systems, Cortland County, New York, USA
http://OakRoadSystems.com/
HTML 4.01 spec: http://www.w3.org/TR/html401/
validator: http://validator.w3.org/
CSS 2 spec: http://www.w3.org/TR/REC-CSS2/
2.1 changes: http://www.w3.org/TR/CSS21/changes.html
validator: http://jigsaw.w3.org/css-validator/
Stan Brown
Then why should anyone be interested in anything you say?
Stan Brown
>"Albert Wiersch" <mrinternetREMVETHIS...@wiersch.com>
>wrote:
>> CSE HTML Validator doesn't use the same engine as the W3 validator so
>> the results won't be the same.
>
>have read that and you have decided to keep advertizing it as a
>validator. You have now reached a new level of bogosity with the babble
>about "engine".
Question: I volunteer at the local library, and they have engaged an
intern to produce a Web site for them. (This library comes very late
to the Internet. The long-time director basically believed in
nothing but print and maybe audiocassettes. The new director is
trying to make up for 40 years of neglect.)
The intern has installed this CSE "Validator" on the library
computer and is presumably using it to test, or develop, or
something. Since she obviously doesn't know any better, how do I
explain to her in words of one syllable what is wrong?
Albert Wiersch
"Stan Brown" <the_sta...@fastmail.fm> wrote in message
news:MPG.1ae91f46d...@news.odyssey.net...
>
> The intern has installed this CSE "Validator" on the library
> computer and is presumably using it to test, or develop, or
> something. Since she obviously doesn't know any better, how do I
> explain to her in words of one syllable what is wrong?
*Most* people, including her, are better off with CSE HTML Validator than
using a "real" validator. I wouldn't be too quick to discount a useful
product because a couple of people here have an issue with the name.
Be sure to read these pages:
http://www.htmlvalidator.com/htmlval/whycseisbetter.html
and
http://www.htmlvalidator.com/htmlval/testimonials.html
Albert Wiersch
>
> But if you're knowingly writing invalid HTML, doesn't the whole idea
> of "validating" it go out the window?
No, of course not. The exception is only if one is a purist and is limited
to only using standards and 100% technical correctness. Most of the world
isn't so technical and perfectionistic (and that can be a good thing!).
I get plenty of value and use out of validating HTML that contains some
"invalid" markup that I choose to use.
Albert Wiersch
> >I'm not a purist. I do not care if I "tell a lie"
>
> Then why should anyone be interested in anything you say?
Normally it is a good thing to try to understand what someone is saying by
taking it in context. Why not try requoting me properly?
If you don't care what I have to say, that's your choice. I don't mind.
Take care,
Albert Wiersch
http://www.htmlvalidator.com/
Albert Wiersch
"Jan Roland Eriksson" <jre...@newsguy.com> wrote in message
news:5pau70pm3dtkj032s...@4ax.com...
> On Thu, 15 Apr 2004 19:28:13 +0100, Steve Pugh <st...@pugh.net> wrote:
>
>
> From a formal SGML standpoint, we need to correct this statement :-)
>
> Comments in a "Document Type Definition" entity is not allowed to be
> read as "documentation" of semantics since the DTDefinition itself is
> only about markup syntax in the first place.
>
> Also the inclusion of a link, to the governing "Formal Prose" part of an
> application of SGML, is questionable. The SGML Declaration (if any) and
> the DTDefinition comes as a result of the "Formal Prose" definition for
> an application of SGML so in effect we should look at this as "traveling
> down a one way street".
>
> Formal Prose definition comes first, if that one calls for
> changes/additions to the SGML Reference Concrete Syntax -- write an SGML
> declaration to the fact. Finally, and still based on the prose, write a
> DTDefinition that expresses the syntactical rules for the markup that
> conforms to the prose. The rest of it, i.e. the semantic definitions for
> that markup, will still only be available in the "Formal Prose" part.
Good information!
So, if I interpreted your post correctly, it seems that it makes sense to
check HTML according to the "formal prose" as well as the DTD? Since a
"real" DTD validator only checks according to the DTD, wouldn't it be a good
idea to also check according to the "formal prose", perhaps by using a
program that goes beyond DTD checking?
David Dorward
> I get plenty of value and use out of validating HTML that contains some
> "invalid" markup that I choose to use.
That is a contradiction. HTML can not be validating if it contains invalid
markup.
And the amount of value you get out if it is obvious if you can't remember
why you used non-standard markup to achieve something that can be achieved
easily with standard markup except *maybe* in some browser you can't
identify (thus removing the need for you to provide evidence that you
really did have a good reason).
(reference: <news:se-dnSHjmri...@august.net>)
David Dorward
> That's not CSE HTML Validator generating those warnings. CSE has no
> problem with <p/>.
Shame. Browsers do.
Jukka K. Korpela
wrote:
> Good information!
We are not surprised at seeing that factual information about SGML was
new to the advertizer of the so-called "CSE HTML Validator".
> So, if I interpreted your post correctly, it seems that it makes
> sense to check HTML according to the "formal prose" as well as the
> DTD?
Yes. Have to tried doing that to your markup?
> Since a "real" DTD validator only checks according to the DTD,
> wouldn't it be a good idea to also check according to the "formal
> prose", perhaps by using a program that goes beyond DTD checking?
Maybe for some SGML applications, if someone wrote such a program. Surely
you cannot do that without knowing SGML, without knowing what "formal
prose" means here, and without actually reading the prose of the
specification of the SGML application in question.
In the case of HTML, the possibilities are fairly limited. You could, for
example, report <p></p> as valid but incorrect (at least at some level -
it violates a specific "should not" statement), but even <p> </p> (a
common product of wysiwyg software) is questionable, since the prose is
obscure in this issue. (Is a paragraph consisting of a no-break space
empty?) And although <p class="heading2">Introduction</p> rather
undeniably violates the prose of the HTML specifications, while
unquestionably valid, it is hardly possible to capture such cases, except
perhaps desultorily on a heuristic basis. And the more cases you try to
cover with heuristics, the more errors in error messages you generate.
The "CSE HTML Validator", which is not a validator (and not even an
imaginary validator) but is being sold and advertized as a validator,
despite several mentions of this fact in public through years,
does not even come close to _trying_ to be a checker that consists of a
validator and additional checks against the prose of the HTML
specifications. It just reflects its author's ideas, opinions, and
tastes, which may occasionally coincide with the formalized requirements
in HTML DTDs, or the prose of HTML specifications, or other W3C
recommendations - but its author hasn't bothered distinguishing such
issues from rules he just made up.
Jukka K. Korpela
> The intern has installed this CSE "Validator" on the library
> computer and is presumably using it to test, or develop, or
> something. Since she obviously doesn't know any better, how do I
> explain to her in words of one syllable what is wrong?
That's quite a challenge, but it is useful to try and learn to express
oneself with monosyllabic words at times.
It [the "CSE HTML Validator" - sorry I can't _name_ it using words of one
syllable :-) ] just tells what its author likes. Don't use it if you
don't know the real rules well enough to tell what's right and what's
wrong in what it spits out. It's known to make false claims 'bout things
being wrong when they are in fact right. It works by rules that someone
just made up.
(Sorry about the occasional bisyllabics.)
Torbjørn Pettersen
> I have downloaded, and installed CSE HTML Validator Pro,
> but I don't get the same results with that as I do with the online
> validator on W3.org. And I can't upload files to W3.org either,
> due to all the ASP code I use.
>
> So, the question is, where can I find a validator that validates
> like the one on W3.org? (Windows 2000 server).
OK, this thread has answered a lot of questions for me.
I'm totally new to validated HTML. I knew about W3, and the
online HTML validator. I did have a very basic knowledge about
validated HTML, and the idea behind it.
I have a web site with a ::lot:: of pages, with a ::lot:: of very bad coding.
To make the pages I had been using FrontPage, which didn't really
do things better. ;-)
To clean things up, and avoid getting all the dirty FrontPage code,
I switched over to DreamWeaver, and set up a web server on my
local network. Then I started redoing my web site, using my local
web server as a test server. Thus having my "dirty" web site online,
and not slapping my visitors in the face with a lot of errors when
visiting my site.
As I use a lot of ASP code, uploading the files to the W3 validator
didn't work, and I went out looking for something I could install
on my server, to check the pages locally. I knew I needed a HTML
validator I could download, and I that is what I searched for using
Google. CSE HTML Validator came up as a clear winner on the
search, as I saw it listed very many places. So, I downloaded it,
installed it, and was happy to see that it worked.
Then I uploaded one of my pages, and tested it with the online
validator at W3.org..... That's when I found out that something
smelt rotten fish. :-/ And after posting my original message here,
and reading all the answers, I found that it didn't just ::smell::
rotten, it ::was:: rotten, very rotten.
Conclusion:
If CSE HTML Validator had it's base (and server) in Norway
I could have reported the case to the Consumer Ombudsman
who would have found it/him to violate Norwegian laws, and he
would have to change the name of his product. Most likely he
would also have been fined.
As it is I have wasted a ::lot:: of work using this, for me, totally
useless piece of software, and would recommend everyone in
my situation (newbie to HTML validation) to avoid it at all cost.
--
Torbjørn Pettersen
Editor/Webmaster
FantaFiction
Steve Pugh
>On Thu, 15 Apr 2004 19:28:13 +0100, Steve Pugh <st...@pugh.net> wrote:
>
>>3. Anyone looking at your code will be able to look up your
>>DTD and see how you have defined the extensions you are using.
>>If the DTD is commented or contains the URL of full documentation...
>
>From a formal SGML standpoint, we need to correct this statement :-)
Of course HTML and the 'formal SGML standpoint' are barely acquainted,
but let's play along. ;-)
>Comments in a "Document Type Definition" entity is not allowed to be
>read as "documentation" of semantics since the DTDefinition itself is
>only about markup syntax in the first place.
Fair enough.
However, when someone creates a custom HTML DTD adding comments to say
"LOOK! I CHANGED THIS BIT" or "Netscape proprietary extension" are
essential not only for other interested parties but also for the
author when he returns in six months time and tries to find the
changes he made (I know, I've been there).
>Also the inclusion of a link, to the governing "Formal Prose" part of an
>application of SGML, is questionable. The SGML Declaration (if any) and
>the DTDefinition comes as a result of the "Formal Prose" definition for
>an application of SGML so in effect we should look at this as "traveling
>down a one way street".
I see what you're saying but given the situation I described in my
post, how else is an author to get at the formal prose?
Is there anything actually wrong with a comment containing the URL of
the formal prose?
>>Depends. Customs DTDs tend to trigger Standards mode in modern
>>browsers so the literal answer to the above question is "quite a lot
>>of them".
>
>You changed subject; now you are talking about a "Document Type
>Declaration".
Whoops. Yes. I thought I'd remembered to get this right, but this one
slipped through.
Albert Wiersch
> "Albert Wiersch" <mrinternetREMVETHIS...@wiersch.com>
> wrote:
>
> > So, if I interpreted your post correctly, it seems that it makes
> > sense to check HTML according to the "formal prose" as well as the
> > DTD?
>
> Yes. Have to tried doing that to your markup?
All the time!
> In the case of HTML, the possibilities are fairly limited. You could, for
> example, report <p></p> as valid but incorrect (at least at some level -
> it violates a specific "should not" statement), but even <p> </p> (a
> common product of wysiwyg software) is questionable, since the prose is
> obscure in this issue. (Is a paragraph consisting of a no-break space
> empty?) And although <p class="heading2">Introduction</p> rather
> undeniably violates the prose of the HTML specifications, while
> unquestionably valid, it is hardly possible to capture such cases, except
> perhaps desultorily on a heuristic basis. And the more cases you try to
> cover with heuristics, the more errors in error messages you generate.
>
> The "CSE HTML Validator", which is not a validator (and not even an
> imaginary validator) but is being sold and advertized as a validator,
> despite several mentions of this fact in public through years,
> does not even come close to _trying_ to be a checker that consists of a
> validator and additional checks against the prose of the HTML
> specifications.
Actually, it is heavily based on the prose part of HTML. Sure, when you get
that high level, there is some opinion involved, but you seem to think that
the whole program is opinion, when it's not. And the "opinion" that it is
based on is a high-quality opinion. Sure, there may be some questions and
problems, but nothing is perfect, especially when you do as much checking as
CSE does.
As you say, there is also some question as to what the HTML prose means, but
in many (most) cases it is clear. And there is A LOT more checking you can
do than simply DTD checking. For instance, accessibility checking is much
more important now.
Of course, I'm not going to get into the "validator" name dispute as that
wouldn't do any good. But I would recommend that you trash the W3C too, as
you trash CSE HTML Validator, because they use the term "CSS Validator" when
there's no such thing according to your definition. And the W3C markets it
as a "CSS Validator"!
Albert Wiersch
"Jukka K. Korpela" <jkor...@cs.tut.fi> wrote in message
news:Xns94CD66517E45...@193.229.0.31...
> Stan Brown <the_sta...@fastmail.fm> wrote:
>
>
> It [the "CSE HTML Validator" - sorry I can't _name_ it using words of one
> syllable :-) ] just tells what its author likes. Don't use it if you
> don't know the real rules well enough to tell what's right and what's
> wrong in what it spits out. It's known to make false claims 'bout things
> being wrong when they are in fact right. It works by rules that someone
> just made up.
Boy, some people just can't stand CSE HTML Validator! You'd be doing a
disservice if you convince someone to dump a useful tool because it doesn't
work the way *YOU* want it to. Especially if you fabricate information about
it like "It works by rules that someone just made up.". I'd like to see the
look on their face when they see the results from the W3C validator after
using CSE.
What would you have them do instead? Become HTML/SGML experts before using
any HTML checking tools? Limit their use of HTML checking to only DTD level?
Something tells me that an expert HTML designer would still want to do
higher level checking and not be so limited.
Albert Wiersch
> Albert Wiersch wrote:
>
> > That's not CSE HTML Validator generating those warnings. CSE has no
> > problem with <p/>.
>
> Shame. Browsers do.
Let me be a little more specific.
CSE has no problem with that by default. But if you think browsers (such as
an older browser) will have a problem with it and you want to be warned of
such constructs, then it is easily changed in CSE by checking an option.
Albert Wiersch
> As it is I have wasted a ::lot:: of work using this, for me, totally
> useless piece of software, and would recommend everyone in
> my situation (newbie to HTML validation) to avoid it at all cost.
In your case, you wanted only (and nothing else but) EXACT DTD checking, and
did not want to check for problems beyond DTD checking. In that case, the
product is not designed for you or for anyone else who wants to do the same.
I wouldn't recommend it in that case either (though I would recommend that
you reconsider limiting yourself to DTD checking). But people with those
requirements is a very small percentage of people. To most people, without
your exact requirements, it would be very useful. That's important to keep
in mind.
Thanks,
Albert
Albert Wiersch
"Albert Wiersch" <mrinternetREMVETHIS...@wiersch.com> wrote
in message news:8YydnaDuRu_...@august.net...
>
> In your case, you wanted only (and nothing else but) EXACT DTD checking,
and
> did not want to check for problems beyond DTD checking. In that case, the
> product is not designed for you or for anyone else who wants to do the
same.
> I wouldn't recommend it in that case either (though I would recommend that
> you reconsider limiting yourself to DTD checking). But people with those
> requirements is a very small percentage of people. To most people, without
> your exact requirements, it would be very useful. That's important to keep
> in mind.
>
Also, I didn't ask what edition you were using. If the results were so
different from the W3C, then you may have been using the lite edition, which
is clearly stated on our site that it only provides basic checking. Much
more in depth checking (beyond DTD checking) is available in the other
editions. If you were using the lite edition and demanding at least W3C
validator compliant pages, then you would have most likely found the
standard or professional editions far more useful.
But if your requirement remains to only make your pages pass the W3C
validator and want to do no further checking, then CSE in any edition may
not be want you want. But I have had users report that they use CSE to
develop their pages, and then, for a final check, use the W3C validator.
That way, they have great benefit from CSE-- as it helps them easily write
good HTML (going beyond DTD checking) and they don't have the hassle of
accessing the W3C validator. When they pass CSE, they can then see if there
is anything additional they want to change to be completely OK by the W3C
validator.
CSE also has a "Standards compliant check" which makes the validation more
like the W3C (though still not exact-- you still get the additional checking
that CSE does, but without things such as proprietary markup).
Stan Brown
>> The intern has installed this CSE "Validator" on the library
>> computer and is presumably using it to test, or develop, or
>> something. Since she obviously doesn't know any better, how do I
>> explain to her in words of one syllable what is wrong?
>
>That's quite a challenge, but it is useful to try and learn to express
>oneself with monosyllabic words at times.
>
>It [the "CSE HTML Validator" - sorry I can't _name_ it using words of one
>syllable :-) ] just tells what its author likes. Don't use it if you
>don't know the real rules well enough to tell what's right and what's
>wrong in what it spits out. It's known to make false claims 'bout things
>being wrong when they are in fact right. It works by rules that someone
>just made up.
Thanks. I guess I'd have to look at all the source code and scope
out what was right and what was wrong, then match it to what the
"CSE HTML Validator" says and point out the faux pas. Sounds like a
lot of work, and at the end of the day I might be blamed not
thanked.
(All one-syllable words, and quite a struggle!)
>(Sorry about the occasional bisyllabics.)
:-)
Stan Brown
comp.infosystems.www.authoring.html:
>"Stan Brown" <the_sta...@fastmail.fm> wrote in message
>news:MPG.1ae91e247...@news.odyssey.net...
>>
>> But if you're knowingly writing invalid HTML, doesn't the whole idea
>> of "validating" it go out the window?
>
>No, of course not. The exception is only if one is a purist and is limited
>to only using standards and 100% technical correctness. Most of the world
>isn't so technical and perfectionistic (and that can be a good thing!).
So someone who thinks "valid" should mean valid and a "validator"
should validate is a purist, but you (by your own words) don't mind
telling a lie.
And that would make you a ...
Nick Kew
Stan Brown <the_sta...@fastmail.fm> writes:
> something. Since she obviously doesn't know any better, how do I
> explain to her in words of one syllable what is wrong?
It is important to note that validation has a very precise meaning.
Unfortunately the issue is confused by the fact that some products
falsely claim to "validate", whilst in fact applying an arbitrary
selection of tests that are not derived from any standard. Such tools
may be genuinely useful, but should be used alongside true validation,
not in place of it.
</blockquote>
Or in shorter words, "Would you prefer to work to open, published
standards, or trust to the individual choices of one man who is known
to be a liar and doesn't understand how the standards work"?
(BTW, "some products" in the above of course means the CSE snake-oil.
I've also seen Tidy and Bobby mis-described as validators, but not by
their authors or vendors, so we can attribute that to honest ignorance).
--
Nick Kew
Nick's manifesto: http://www.htmlhelp.com/~nick/
Nick Kew
"Torbjørn Pettersen" <tpe AT broadpark DOT no> writes:
> I wrote...
>> I have downloaded, and installed CSE HTML Validator Pro,
>> but I don't get the same results with that as I do with the online
>> validator on W3.org.
Indeed, you started this whole thread, as a clear example of a user
who installed a product in good faith, only to find you'd been conned.
> I'm totally new to validated HTML. I knew about W3, and the
> online HTML validator. I did have a very basic knowledge about
> validated HTML, and the idea behind it.
A daresay there are many in your position.
> Conclusion:
> If CSE HTML Validator had it's base (and server) in Norway
> I could have reported the case to the Consumer Ombudsman
> who would have found it/him to violate Norwegian laws, and he
> would have to change the name of his product. Most likely he
> would also have been fined.
Here (UK) we have legislation covering truth in advertising, and
accuracy of product descriptions. IANAL, but I expect CSE may be
in violation of both of those if it does business here. I believe
most (developed) countries have comparable legislation.
> As it is I have wasted a ::lot:: of work using this, for me, totally
> useless piece of software, and would recommend everyone in
> my situation (newbie to HTML validation) to avoid it at all cost.
OK, you've got some answers: for a Windows user, arealvalidator is
presumably going to be far and away the easiest to install and use.
That's not to say other tools are not useful. As a developer myself,
I offer both formal validation and an accessibility analysis tool at
<URL:http://valet.webthing.com/>. If I described the latter tool
(AccessValet) as a validator, I'd be telling the same lie as CSE.
But I'd consider AccessValet at least as useful in "real life" as
the true validator, Page Valet.
Torbjørn Pettersen
> In your case, you wanted only (and nothing else but) EXACT DTD checking, and
> did not want to check for problems beyond DTD checking. In that case, the
> product is not designed for you or for anyone else who wants to do the same.
> I wouldn't recommend it in that case either (though I would recommend that
> you reconsider limiting yourself to DTD checking). But people with those
> requirements is a very small percentage of people. To most people, without
> your exact requirements, it would be very useful. That's important to keep
> in mind.
I think most peple would think like me, and a lot of people would think that
when CSE reports no errors, their code ::is:: valid HTML as understood by
most people, i.e. as set by W3C.
You even claim that CSE is ::better:: than W3C's validatior. But how can it
be when it doesn't follow the W3C's validation rules? Close to fraud if you
ask me.
What I'm after is to put that little gif on my page, which says that that my
HTML is "Valid HTML 4.01!". This can't happen with your software.
Thanks, but no thanks,
Torbjørn
Torbjørn Pettersen
> Also, I didn't ask what edition you were using. If the results were so
> different from the W3C, then you may have been using the lite edition, which
> is clearly stated on our site that it only provides basic checking. Much
> more in depth checking (beyond DTD checking) is available in the other
> editions. If you were using the lite edition and demanding at least W3C
> validator compliant pages, then you would have most likely found the
> standard or professional editions far more useful.
I tried both, and neither of them complies to the W3C standard.
The question is; Why would I use I software I can't trust to give me correct information?
Albert Wiersch
"Nick Kew" <ni...@hugin.webthing.com> wrote in message
news:kqv4l1...@webthing.com...
>
> Or in shorter words, "Would you prefer to work to open, published
> standards, or trust to the individual choices of one man who is known
> to be a liar and doesn't understand how the standards work"?
Or you could try telling the truth and stop slandering people.
Do you call everyone who disagrees with you a liar? Do you slander them? I
hope not.
Albert Wiersch
"Torbjørn Pettersen" <tpe AT broadpark DOT no> wrote in message
news:407f...@news.broadpark.no...
> Albert Wiersch wrote...
>
> I think most peple would think like me, and a lot of people would think
that
> when CSE reports no errors, their code ::is:: valid HTML as understood by
> most people, i.e. as set by W3C.
>
> You even claim that CSE is ::better:: than W3C's validatior. But how can
it
> be when it doesn't follow the W3C's validation rules? Close to fraud if
you
> ask me.
Read the page I pointed out and it explains why it's better. It can find
problems that the W3C validator doesn't. Granted, there may be issues that
the W3C finds that CSE doesn't, but in most cases CSE is better to use if
your HTML is intended to be viewed on real-world browsers by real-world
people.
The arguments that CSE is "fraud" and "worthless" are completely untrue, as
thousands of people have found it to be quite useful.
>
> What I'm after is to put that little gif on my page, which says that that
my
> HTML is "Valid HTML 4.01!". This can't happen with your software.
Sure it can. Just use the W3C validator after using CSE to get the best of
both. Just use CSE's "standards compliant mode" in the standard or
professional edition and there won't be much more for you to do (if
anything). Plus, you'd benefit from the ease of use of CSE, not having to
access an online validator, and from the additional checking CSE does.
Brian
> Stan Brown <the_sta...@fastmail.fm> wrote:
>
>> The intern has installed this CSE "Validator" on the library
>> computer
>> how do I explain to her in words of one syllable what is wrong?
>
> That's quite a challenge, but it is useful to try and learn to
> express oneself with monosyllabic words at times.
:-)
> "CSE HTML Validator"
> Don't use it if you don't know the real rules well enough to tell
> what's right and what's wrong in what it spits out. It's known to
> make false claims 'bout things being wrong when they are in fact
> right. It works by rules that someone just made up.
A "linter" or heuristics checker is not unlike a word proecessor's
grammar checker. I've told students not to accept a proposed change from
a grammar checker unless they understand the rule being applied and
agree with the proposed change. Everytime I use MS Word's checker, it
flaggs constructs that are not errors. Even in French, which is my
second language, I find I often know better then the program. This is
rather natural when one is checking for things for which there's no hard
and fast rules.
--
Brian (remove "invalid" from my address to email me)
http://www.tsmchughs.com/
Alan J. Flavell
> As it is I have wasted a ::lot:: of work using this, for me, totally
> useless piece of software,
As Nick Kew already said on his web page:
<blockquote>
Unfortunately the issue is confused by the fact that some products
falsely claim to "validate", whilst in fact applying an arbitrary
selection of tests that are not derived from any standard. Such tools
may be genuinely useful, but should be used alongside true validation,
not in place of it.
</blockquote>
Many of us already knew which product Nick was referring to in that
somewhat coded message. However, you will note that Nick says "such
tools may be genuinely useful", and I would say the same: but only if
they're presented honestly as what they are - a heuristic checker or
linter, not to be confused with "validation" in the specialised sense
which is appropriate when the term is encountered in SGML contexts.
BOTH kinds of tool can, indeed, be genuinely useful, when used
appropriately; in fact, validation (in the technical sense) *can* be
potentially misleading, to someone who is unaware of the issues, since
there are lots of things that are technically valid (in the SGML
sense) in HTML which are not, in fact, supported by the generally-used
client agents, and certainly *will* cause indigestion if they are -
deliberately or inadvertently - used in a WWW context, irrespect of
the verdict of a formal validator on the subject.
IMHO and YMMV, anyhow.
Brian
> What I'm after is to put that little gif on my page, which says that
> that my HTML is "Valid HTML 4.01!".
Why? "Valid" icons are a mere distraction for visitors. Such icons or
declarations should be limited to a page describing the site mechanics,
which might include what tools were used to build the site, what
standards adhered to, etc.
Producing good web documents is the job of every author. Part of that
job is making sure the documents are valid. The public need not know
that you've done a good job. They should simply be able to access and
process your document easily and enjoyably.
David Dorward
> Of course, I'm not going to get into the "validator" name dispute as that
> wouldn't do any good. But I would recommend that you trash the W3C too, as
> you trash CSE HTML Validator, because they use the term "CSS Validator"
http://groups.google.com/groups?selm=Xns93304C0DDA3Djkorpelacstutfi%40193.229.0.31&oe=UTF-8&output=gplain
http://groups.google.com/groups?selm=Xns944D7E5F6A1F3jkorpelacstutfi%40193.229.0.31&oe=UTF-8&output=gplain
Jukka K. Korpela
wrote:
> Boy, some people just can't stand CSE HTML Validator!
You misspelled "continued dishonesty". HTH.
C A Upsdell
news:Pine.LNX.4.53.04...@ppepc56.ph.gla.ac.uk...
> BOTH kinds of tool can, indeed, be genuinely useful, when used
> appropriately; in fact, validation (in the technical sense) *can* be
> potentially misleading, to someone who is unaware of the issues, since
> there are lots of things that are technically valid (in the SGML
> sense) in HTML which are not, in fact, supported by the generally-used
> client agents, and certainly *will* cause indigestion if they are -
> deliberately or inadvertently - used in a WWW context, irrespect of
> the verdict of a formal validator on the subject.
Y'know, years ago when I used C and C++, I used Gimpel's Lint program to
check my code for problems, and I remember that Lint would emit both error
and warning messages: error messages for syntactic errors; and several
levels of warning messages for questionable code; and Lint allowed me to
control which conditions would trigger the warnings.
Albert Wiersch could enhance his product and avoid this kind of controversy
by making his product behave much the same as my old Lint program: by
adding true validation to his product, emitting error messages for
validation errors, and warning messages for questionable code.
So Albert, why don't you? This would make your product more useful, which I
would think would enhance your sales (and profits!), and might also convert
some of your detractors into disciples: you could do well and do good, both
at the same time.
Why not, Albert?
Albert Wiersch
"C A Upsdell" <cupsdell0311XXX@-@-...@XXXrogers.com> wrote in message
news:U5Vfc.10652$QG5....@news04.bloor.is.net.cable.rogers.com...
> "Alan J. Flavell" <fla...@ph.gla.ac.uk> wrote in message
> news:Pine.LNX.4.53.04...@ppepc56.ph.gla.ac.uk...
>
> check my code for problems, and I remember that Lint would emit both error
> and warning messages: error messages for syntactic errors; and several
> levels of warning messages for questionable code; and Lint allowed me to
> control which conditions would trigger the warnings.
>
> Albert Wiersch could enhance his product and avoid this kind of
controversy
> by making his product behave much the same as my old Lint program: by
> adding true validation to his product, emitting error messages for
> validation errors, and warning messages for questionable code.
That's how CSE already works, but without the "real" validator. It generates
errors for more severe issues and then warnings, messages, and comments for
other issues. And most message types can easily be changed. It's extremely
configurable. You can even add or change the messages to suit your style.
> So Albert, why don't you? This would make your product more useful, which
I
> would think would enhance your sales (and profits!), and might also
convert
> some of your detractors into disciples: you could do well and do good,
both
> at the same time.
As for adding a real validator, I'd certainly look into it if there was
demand for it, but there's hardly anyone asking me to add it, and I'm not
going to add it just for the few people (two or three maybe?) who don't like
the name (who wouldn't buy it anyway). I'd be more than happy to add it if I
thought my users would find real value in it (provided the effort justifies
the benefit). Who knows though, maybe it will get added in a future version.
I have considered it before but couldn't justify it at the time.
Thanks,
Albert
Albert Wiersch
> Albert Wiersch wrote:
>
> > Of course, I'm not going to get into the "validator" name dispute as
that
> > wouldn't do any good. But I would recommend that you trash the W3C too,
as
> > you trash CSE HTML Validator, because they use the term "CSS Validator"
>
>
http://groups.google.com/groups?selm=Xns93304C0DDA3Djkorpelacstutfi%40193.229.0.31&oe=UTF-8&output=gplain
>
http://groups.google.com/groups?selm=Xns944D7E5F6A1F3jkorpelacstutfi%40193.229.0.31&oe=UTF-8&output=gplain
Thanks, I see that Jukka is unhappy with the name "CSS Validator" (he says
it is misleading), but I don't see him bashing the W3C like he bashes CSE.
Seems quite unfair in my opinion. If you bash one, you ought to bash the
other in the same manner and magnitude.
C A Upsdell
in message news:08mdncMub9m...@august.net...
> As for adding a real validator, I'd certainly look into it if there was
What there is, is a demand for are tools for finding problems (or potential
problems) in web pages. By not having a true validator, you are letting
some problems slip by ... thereby not meeting the demand as well as you
could ... and surely missing out on some sales.
I myself would be interested in a more comprehensive toolkit than what you
offer.
Brian
> C A Upsdell wrote...
>
>> Albert Wiersch could enhance his product and avoid this kind of
>> controversy by making his product behave much the same as my old
>> Lint program: by adding true validation to his product, emitting
>> error messages for validation errors, and warning messages for
>> questionable code.
>
> That's how CSE already works, but without the "real" validator.
In other words, that's not how CSE works, since C A Upsdell suggested
that you add a real validator.
BTW, why did you put "real" in quotes in your message? Does it mean you
don't really believe in real validators?
Brian
> David Dorward wrote ...
>
>> http://groups.google.com/groups?selm=Xns93304C0DDA3Djkorpelacstutfi%40193.229.0.31
>>
>> http://groups.google.com/groups?selm=Xns944D7E5F6A1F3jkorpelacstutfi%40193.229.0.31
>
> Thanks, I see that Jukka is unhappy with the name "CSS Validator" (he
> says it is misleading), but I don't see him bashing the W3C like he
> bashes CSE.
W3C representatives aren't arguing in ciwah that the name "CSS
Validator" is appropriate. If they were, I've little doubt that they
would come under fire.
Albert Wiersch
"C A Upsdell" <cupsdell0311XXX@-@-...@XXXrogers.com> wrote in message
news:7dWfc.55510$2oI1....@twister01.bloor.is.net.cable.rogers.com...
> "Albert Wiersch" <mrinternetREMVETHIS...@wiersch.com>
wrote
> in message news:08mdncMub9m...@august.net...
>
> What there is, is a demand for are tools for finding problems (or
potential
> problems) in web pages. By not having a true validator, you are letting
> some problems slip by ... thereby not meeting the demand as well as you
> could ... and surely missing out on some sales.
>
> I myself would be interested in a more comprehensive toolkit than what you
> offer.
Thanks. I appreciate your comments.
Albert Wiersch
"Brian" <use...@julietremblay.com.invalid> wrote in message
news:1080doo...@corp.supernews.com...
> Albert Wiersch wrote:
>
> BTW, why did you put "real" in quotes in your message? Does it mean you
> don't really believe in real validators?
I should have said DTD validator to avoid the confusion.
Brian
> I should have said DTD validator to avoid the confusion.
Except that, in terms of SGML, there is no other kind.
Albert Wiersch
"Brian" <use...@julietremblay.com.invalid> wrote in message
> Albert Wiersch wrote:
>
> W3C representatives aren't arguing in ciwah that the name "CSS
> Validator" is appropriate. If they were, I've little doubt that they
> would come under fire.
By using the name, aren't they saying that it's appropriate?
Anyway, I'm not arguing about it anymore. There's no point to it.
Thanks,
Albert Wiersch
http://www.htmlvalidator.com/
Brian
>
>> W3C representatives aren't arguing in ciwah that the name "CSS
>> Validator" is appropriate. If they were, I've little doubt that
>> they would come under fire.
[I actually wrote the quoted part above, though it does not appear that
way in your message. Please edit more carefully. Your postings make it
look like you said things that I said or the other way around.]
Albert Wiersch wrote:
> By using the name, aren't they saying that it's appropriate?
They aren't saying it's appropriate *in ciwah*. If they were...
> Anyway, I'm not arguing about it anymore.
We'll see.
Alan J. Flavell
> "Brian" <use...@julietremblay.com.invalid> wrote in message
> news:1080f2a...@corp.supernews.com...
> > Albert Wiersch wrote:
> >
> > W3C representatives aren't arguing in ciwah that the name "CSS
> > Validator" is appropriate. If they were, I've little doubt that they
> > would come under fire.
>
> By using the name, aren't they saying that it's appropriate?
CSS is not an SGML application, and makes no claim to being one. In
relation to CSS, the term "validation" is not pre-defined. What the
W3C validator does is to investigate the CSS syntax, and report errors
(which is what a validator does, in general terms), and optionally
offer warnings (which is no business of a validator as such, but is
nevertheless useful). It seems to me that the CSS thingy is primarily
a validator, and that its claim to be one does not clash with any
pre-defined specification of what a CSS validator should be.
So: it's a matter of personal taste (rather than of objective
definition) whether the term is appropriate.
In relation to HTML (starting with HTML2.0/RFC1866 - remember that?
- if you don't, at least some of us do) the term "validation" already
-had- a defined meaning, and was not open to marketing puffery.
> Anyway, I'm not arguing about it anymore.
Didn't we already get to hear that before? The hole is getting
deeper.
> There's no point to it.
Perhaps someone will bother to refer it under the Trades Description
Act.
Even Nick was willing to concede that it could be a useful tool, if
offered honestly for what it is: I say the same myself. As long as
you misrepresent it - despite your public admission that it's a lie,
with the implication that calling it a "validator" is a purely
marketing strategy to impress the gullible - we have no alternative
than to warn folks off it.
Rijk van Geijtenbeek
<use...@julietremblay.com.invalid> wrote:
> Albert Wiersch wrote:
>
>> I should have said DTD validator to avoid the confusion.
>
> Except that, in terms of SGML, there is no other kind.
What passes as HTML on the web has, on average, very little to do with
SGML... Albert has a point here :)
--
Rijk van Geijtenbeek
The Web is a procrastination apparatus:
It can absorb as much time as is required to ensure that you
won't get any real work done. - J.Nielsen
C A Upsdell
news:Pine.LNX.4.53.04...@ppepc56.ph.gla.ac.uk...
Surely it would be possible to devise a strict grammar for CSS, in which
case software that identified syntax errors with respect to this grammar
could reasonably be called a validator.
Nick Kew
"C A Upsdell" <cupsdell0311XXX@-@-...@XXXrogers.com> writes:
>> CSS is not an SGML application, and makes no claim to being one. In
>> relation to CSS, the term "validation" is not pre-defined ...
>
> Surely it would be possible to devise a strict grammar for CSS, in which
> case software that identified syntax errors with respect to this grammar
> could reasonably be called a validator.
Indeed, that's exactly what tools like the W3C CSS validator do (bugs aside).
But your words "could reasonably be called", when contrasted with something
more definite like "is by definition", make Alan's point. And - crucially -
once we have "is by definition", we need a formal and well-understood
definition not merely of the syntax to be validated, but of a validation
process. No such thing exists for CSS, which is why it's considered
to be legitimately open to armwaving like "could reasonably be called".
Nick Kew
"Rijk van Geijtenbeek" <ri...@opera.com> writes:
> What passes as HTML on the web has, on average, very little to do with
> SGML... Albert has a point here :)
and probably not the most important.
It justifies having checking tools other than validators.
It doesn't justify selling a lie to innocent end-users.
Brian
> On Fri, 16 Apr 2004, Brian wrote:
>
>> Albert Wiersch wrote:
>>
>>> I should have said DTD validator to avoid the confusion.
>>
>> Except that, in terms of SGML, there is no other kind.
>
> What passes as HTML on the web has, on average, very little to do
> with SGML...
You appear to know more about SGML than I do (that wouldn't be hard;
there are likely scores of folks here know more that I do). But I'll be
brave and ask questions.
If the HTML one finds on the www has little to do with SGML -- no
argument from me on that one -- does that change the meaning of HTML
validation? Does that divorce HTML validation entirely from SGML?
> Albert has a point here :)
I'm not so sure. Whatever one may say about HTML on the www, I don't
think that's an excuse to mischaracterize a linter as a validator and
confuse the matter for the ignorant.
Rijk van Geijtenbeek
<use...@julietremblay.com.invalid> wrote:
> Rijk van Geijtenbeek wrote:
..
>> Albert has a point here :)
>
> I'm not so sure. Whatever one may say about HTML on the www, I don't
> think that's an excuse to mischaracterize a linter as a validator and
> confuse the matter for the ignorant.
By creating a CSS checker and calling it a 'CSS validator', the W3C itself
indicates that there is more to the word 'validator' than parsing an SGML
document according to a DTD...
I think it shouldn't be too hard to build in 'real' DTD validation in
Albert's product, it would make it a better product. This criticism of CSE
is years old, and obviously either Albert is very stubborn, or he doesn't
understand why it can be useful to know a document is valid in the sense
of condforming to a specified DTD. Either way, it doesn't inspire much
confidence in the quality of the product.
Jan Roland Eriksson
<fla...@ph.gla.ac.uk> wrote:
[crossed, and fup's set, to ciwas]
[...]
>CSS is not an SGML application, and makes no claim to being one.
[...]
A thought that strikes me rather late in the day is if it would be
possible to create an SGML declaration that is fitting for the available
CSS syntax, i.e. could a retrofit work be done that brings the CSS
syntax into an SGML conforming state?
At time of writing this post I don't know if the idea of an SGML
declaration for available CSS syntax, and a DTDefinition that makes use
of that SGML declaration, is possible to turn into reality.
But it sure has a nice ring to it to be able to have a unified method of
true validation (based on nsgmls most probably) for both (X)HTML markup
and CSS stylesheets; No?
--
Rex .. <r...@css.nu> .. <http://css.nu/>
Jan Roland Eriksson
<use...@julietremblay.com.invalid> wrote:
>Rijk van Geijtenbeek wrote:
[...]
>> What passes as HTML on the web has, on average, very little to do
>> with SGML...
[...]
>If the HTML one finds on the www has little to do with SGML --
>...
>-- does that hange the meaning of HTML validation?
No.
>Does that divorce HTML validation entirely from SGML?
No. The HTML spec claims to describe an application of SGML; we all know
it has its flaws, and right out errors, in that respect.
Nevertheless, validation as per SGML terms is defined and shall be
available to all defined SGML applications.
Now if folks in general could only be brought to understand the inherent
meaning of a "markup syntax validation" and that it's not a God given
thing that automatically leads to a good www appearance :-)
>> Albert has a point here :)
>I'm not so sure. Whatever one may say about HTML on the www, I don't
>think that's an excuse to mischaracterize a linter as a validator and
>confuse the matter for the ignorant.
I don't like Albert's naming approach either, but mind you all, his CSE
has a long track record as it goes back for some 6 years as I know of it
today. Back in the "early days of CSE" few people in the www area at
large, had a grip on the fact that the word "valid" might have been
claimed for special meaning in some obscure thing called SGML.
Albert can easily solve his long time criticized use of the words "HTML
validator" by writing an email to Liam Quinn and ask to buy his code for
a fully working front end of nsgmls. Liam "hacked up" a MS-Windows front
for nsgmls and asked $25 as a one time fee from people who wanted it.
Liam has not, to my knowledge, presented any new updates for some years
gone by but I still find that my chipped in $25 is worth all cents in
it.
Albert! (I know you are listening :-) Your next step as an update to
your CSE shall be the incorporation of 'onsgmls' and whatever code it
takes to make that program work seamless inside your CSE product.
Others have said the same already, just go ahead and do it will you :-)
Daniel R. Tobias
> As Nick Kew already said on his web page:
>
> <blockquote>
> Unfortunately the issue is confused by the fact that some products
> falsely claim to "validate", whilst in fact applying an arbitrary
> selection of tests that are not derived from any standard. Such tools
> may be genuinely useful, but should be used alongside true validation,
> not in place of it.
> </blockquote>
I have some similar comments on my page about validators:
http://webtips.dan.info/validators.html
--
== Dan ==
Dan's Mail Format Site: http://mailformat.dan.info/
Dan's Web Tips: http://webtips.dan.info/
Dan's Domain Site: http://domains.dan.info/
Neal
DOT no> wrote:
> What I'm after is to put that little gif on my page, which says that
> that my
> HTML is "Valid HTML 4.01!". This can't happen with your software.
Like Brian said, a web author who advertises "Valid HTML!" is a little
like a policeman wearing a pin that says "I'm not corrupt" or a babysitter
with a T-shirt reading "I'm not a pedophile". You'd kinda hope they
wouldn't need to tout something we generally accept as being required for
competency and trustworthiness at their job.
Torbjørn Pettersen
> > What I'm after is to put that little gif on my page, which says that
> > that my HTML is "Valid HTML 4.01!".
>
> Why? "Valid" icons are a mere distraction for visitors. Such icons or
> declarations should be limited to a page describing the site mechanics,
> which might include what tools were used to build the site, what
> standards adhered to, etc.
If the web sites contents is of such a character that your visitors get
distracted by an icon, it might be time to consider if your content is
really worth a web site.
When the majority of web sites out there ::don't:: validate, having a
"Valid HTML 4.01!" icon serves two purposes; it tells your visitors
you have done an effort to make your site for as broad an audience
as you can, and it draws attention to valid HTML, and can help make
others do the same effort as you.
Torbjørn
Torbjørn Pettersen
> > What I'm after is to put that little gif on my page, which says that
> > that my
> > HTML is "Valid HTML 4.01!". This can't happen with your software.
>
> Like Brian said, a web author who advertises "Valid HTML!" is a little
> like a policeman wearing a pin that says "I'm not corrupt" or a babysitter
> with a T-shirt reading "I'm not a pedophile". You'd kinda hope they
> wouldn't need to tout something we generally accept as being required for
> competency and trustworthiness at their job.
So, you mean that the majority of web sites out there would pass the test?
And do you also think that the majority of people browsing the web know
what valid HTML is? Or that the majority of web builders know? If so,
then your post is deep and meaningful, if not..... well, then it really needs
to be re-thinked.
As I said in my reply to Brian;
Nick Kew
Neal <nea...@spamrcn.com> writes:
> Like Brian said, a web author who advertises "Valid HTML!" is a little
> like a policeman wearing a pin that says "I'm not corrupt" or a babysitter
> with a T-shirt reading "I'm not a pedophile".
A better analogy would be a senior politician or judge with an
"I got where I am by honest means" label. It would be such an
unusual thing as to stand out. But the claim made by the HTML
badges has the advantage of being verifiable.
Neal
DOT no> wrote:
> So, you mean that the majority of web sites out there would pass the
> test?
They should be expected to.
> And do you also think that the majority of people browsing the web know
> what valid HTML is?
They shouldn't have to know.
> Or that the majority of web builders know?
Most do. Some don't recognize its value.
> If so,
> then your post is deep and meaningful, if not..... well, then it really
> needs
> to be re-thinked.
Nope.
> As I said in my reply to Brian;
> When the majority of web sites out there ::don't:: validate, having a
> "Valid HTML 4.01!" icon serves two purposes; it tells your visitors
> you have done an effort to make your site for as broad an audience
> as you can, and it draws attention to valid HTML, and can help make
> others do the same effort as you.
As Nick pointed out, a far better analogy is something like a box of food
at the grocer's which bears the label "Fit for human consumption!"
Sure, there's meaning behind the label. It communicates something - but
what it communicates is something most people don't care about, don't even
know about, and shouldn't even have to be touted.
Besides, you don't show the audience how the magic tricks work because
they don't really want to know. Keep the mechanics out of it. Unless your
page deals with coding valid pages, such a button isn't terribly relevant
to your target audience.
Brian
>
>>> What I'm after is to put that little gif on my page, which says
>>> that that my HTML is "Valid HTML 4.01!".
Brian wrote...
>> Why? "Valid" icons are a mere distraction for visitors. Such icons
>> or declarations should be limited to a page describing the site
>> mechanics, which might include what tools were used to build the
>> site, what standards adhered to, etc.
Torbjørn Pettersen wrote:
> If the web sites contents is of such a character that your visitors
> get distracted by an icon, it might be time to consider if your
> content is really worth a web site.
If your content is worth a web site, then you should stay focussed and
keep the page about that topic. If the page is not about html syntax,
then "valid html" is inappropriate. IMHO.
> When the majority of web sites out there ::don't:: validate, having a
> "Valid HTML 4.01!" icon
A "valid html" message has no meaning to most surfers.
> serves two purposes; it tells your visitors you have done an effort
> to make your site for as broad an audience as you can,
Why? just make it accessible. I don't see why you need to explicitly
tell them you made it accessible.
If you want to tell your audience what efforts you made to make the site
accessible, it might be appropriate -- but not on every page of the
site. Put it on one separate page, "about the site" or something. And on
that page, you might mention valid html. Although I wouldn't even do
that. An "about the site" page should be general interest. On my most
recent project, I do mention "valid html" and other such technical
matters. They are on a "mechanics" page, which begins with a warning
that the discussion that follows is technical. And it is only accessible
via the "about the site" page, which itself is only accessible
indirectly from the main site content.
> and it draws attention to valid HTML, and can help make others do the
> same effort as you.
If they're web authors, you mean. Is your page about web authoring?