Installing validator on RedHat

Hi!

 I've finally managed to get my "home version" of the validator going, from
the tarball you've provided on http://validator.w3.org/source/

I'd have a few notes about some of the issues ecountered:

(I have a RedHat7.3 + Apache 2 + [all other stuff that's needed :) ])

Note to lazy people like myself :) :

 I had OpenSP installed without beeing avare of it - my home machine is
completely offline, so it's used as a "playground" - It comes with the OpenJade
package of RedHat, and puts onsgmls in /usr/bin. Unfortunateley, it's release
1.3.4 or something like that.
  As I mentioned before, without beeing avare that I've already had OpenSP, I've
downloaded it and compiled it with default settings (had to fall back on gcc2.96
provided by RH, my install of gcc3.2.2 gave an error at ./configure saying that
"can't get the size of size_t" (???)). - the downloaded version was of course 1.5.
  Then with more or less pain I've got the interface running

(there are some _interestingly_ set up config files to say the least... If you'd
expect me to put the whole validator in /usr/local [as is in the default
validator.conf], then validator/httpd/conf/httpd.conf is all messed up. if check
should be in the same directory as all the other html stuff (as implied by the
"action" argment of the form in index.html, and the appropriate directives in
the aformentioned httpd.conf), then why put it in a different dir? etc...)

but all the results said "page not valid", but giving NO errors...

Re: Installing validator on RedHat

On Wed, 2003-07-02 at 16:49, Kelemen Zoltan wrote:

> I'd suggest changing the default setting in validator.conf to
> /usr/local/bin/onsgmls if the default compilation of OpenSP puts it there...

Well, even if the default, prefixless compilation puts it there, there
are no guarantees about the version.  And OpenSP 1.5 is being
distributed with a number of OS's that put it into /usr/bin.

> Would it take more
> than 5 minutes to get the same kind of explanations on a HTML page (and linked
> from http://validator.w3.org/source/)?

The current beta version has some improvements:
http://validator.w3.org:8001/source/

--

-- 
\/

validating JavaScript

Re: validating JavaScript

* Jake Brumby wrote:
>I recently put in a lot of effort to develop a website only to have
>to include some JavaScript supplied by Akamai for tracking users that
>the Validator finds 17 different errors in. The code must go in every
>page for the SiteWise tacking program to operate. I told my client about
>this and they followed it up with Akamai. The response I have receive is:
> 
>"The HTML specification does not cover JavaScript.  However, the
>Validator does not recognize JavaScript code from HTML and attempts
>to "validate" it as HTML.  The SiteWise code has been tested with all
>common browser versions and can be used without concern for its function
>or performance."
> 
>Is this correct?

What's the offending code?

RE: validating JavaScript

The offending code is below.

BTW - am I correct in replying to www-validator <at> w3.org and should I include previous threads below in this
email? Thanks!

<SCRIPT Language = "JavaScript">
	var wtl_loc = document.URL.indexOf('https:')==0?'https://a248.e.akamai.net/v/248/2120/1d/download.akamai.com/crs/lgsitewise.js':'http://crs.akamai.com/crs/lgsitewise.js';
	document.write("<SCR" + "IPT Language = 'JavaScript' SRC = '" + wtl_loc + "'></SCR" + "IPT>");
</SCRIPT>

<SCRIPT LANGUAGE="JavaScript">

	var wtl_TagVer = 6;
	var wtl_FWD = 0;
	var wtl_url = document.URL;
	var wtl_title = document.title;
	var wtl_TagID = 148699;
	var wtl_SID = "148699";
	var wtl_Offset = "0";
	WTL_TAG = new Image;
	WTL_TAG.ID = "WTL_TAG";

	var ORDER= "";
	var SERVER= "";
	var INVOICE= "";
	var CARTVIEW= "";
	var CARTADD= "";
	var CARTREMOVE= "";
	var CHECKOUT= "";
	var CARTBUY= "";
	var ADCAMPAIGN= "";

	//wtl_Tag6(TAG,SID,wtl_url,wtl_title,"MainContentGroup,SubContentGroup");
	wtl_Tag6(wtl_TagID,wtl_SID,wtl_url,wtl_title);

</SCRIPT>

<NOSCRIPT>
<IMG BORDER="0" WIDTH="1" HEIGHT="1" SRC="http://statse.webtrendslive.com/S148699/button6.asp?tagver=6&si=148699&offset=0&fw=0&js=No&">
</NOSCRIPT>

* Jake Brumby wrote:
>I recently put in a lot of effort to develop a website only to have
>to include some JavaScript supplied by Akamai for tracking users that
>the Validator finds 17 different errors in. The code must go in every
>page for the SiteWise tacking program to operate. I told my client about
>this and they followed it up with Akamai. The response I have receive is:
> 
>"The HTML specification does not cover JavaScript.  However, the
>Validator does not recognize JavaScript code from HTML and attempts
>to "validate" it as HTML.  The SiteWise code has been tested with all
>common browser versions and can be used without concern for its function
>or performance."
> 
>Is this correct?

What's the offending code?

RE: validating JavaScript

At 2003-07-03T21:07+0100, Jake Brumby wrote:-

[Rearranged]
> >"The HTML specification does not cover JavaScript.

Correct.

> >                                                    However, the
> >Validator does not recognize JavaScript code from HTML

That is also correct - the validator does not include a JavaScript
interpreter.

> >                                                       and attempts
> >to "validate" it as HTML.

If some JavaScript occurs within an HTML document, then it is part of an
HTML document and should be valid as such. In HTML (but not XHTML), the
SCRIPT element has CDATA content, which means that most markup is not
recognised, but the start of an end tag ("</") is, and must not appear in
the content (because it ends it). (See below)

> >Is this correct?

[Bjoern Hoehrmann]
> What's the offending code?

> The offending code is below.
>
> BTW - am I correct in replying to www-validator <at> w3.org and should I

Yes.

> include previous threads below in this email? Thanks!

Including the relevant parts of the preceding discussion is useful to
provide context, but it seems more logical to keep the conversation in
chronological order.

> <SCRIPT Language = "JavaScript">

The language attribute is deprecated (though still allowed); HTML 4
requires there to be a type attribute. Unfortunately there is (still) no
correct MIME type for JavaScript, so you can't win whatever you do.

> 	document.write("<SCR" + "IPT Language = 'JavaScript' SRC = '" + wtl_loc + "'></SCR" + "IPT>");

!!!

The validator will have complained about the "</". That can easily be
fixed by saying "<\/" or splitting the string differently, and this at
least can probably only improve compatibility.

Tim Bagot

Re: validating JavaScript

On Thu, 3 Jul 2003, Jake Brumby wrote:

> "The HTML specification does not cover JavaScript.

No, but it does cover how Javascript can be included in HTML, which
is what your correspondent seems ignorant of.

> However, the Validator does not recognize JavaScript code from HTML and
  attempts to "validate" it as HTML.

That's correct, but doesn't mean what (s)he wants it to mean.  As a rule
it's best to separate javascript from HTML.  Failing that, you need
to read TFM.  Browse the validator help pages, and point your
correspondent at something relevant.

> The SiteWise code has been tested with all common browser versions and
> can be used without concern for its function or performance."

That's a classic warning label telling you that its author is clueless.

--

-- 
Nick Kew

In urgent need of paying work - see http://www.webthing.com/~nick/cv.html

CSS validator code error

Hello,

When anybody validate the css with your validator, you give the option for
add the icon CSS Validated.
And you give this source:

<p>
 <a href="http://jigsaw.w3.org/css-validator/">
  <img style="border:0;width:88px;height:31px"
       src="http://jigsaw.w3.org/css-validator/images/vcss"
       alt="Valid CSS!">
 </a>
</p>

But the image tag is NOT CLOSED.
So it will NOT VALIDATE with the xhtml validator.

Bye
Att. Joan Miquel

Re: CSS validator code error

* Joan Miquel wrote:
>When anybody validate the css with your validator, you give the option for
>add the icon CSS Validated.
>And you give this source:
>
><p>
> <a href="http://jigsaw.w3.org/css-validator/">
>  <img style="border:0;width:88px;height:31px"
>       src="http://jigsaw.w3.org/css-validator/images/vcss"
>       alt="Valid CSS!">
> </a>
></p>
>
>But the image tag is NOT CLOSED.

Your document is either a HTML document or your preferred language is
not english. If the input document is an XHTML document and your
preffered language is english (i.e., you get an english results page)
there img element is properly closed. However, as the CSS validator
correctly states, the code fragment is an *HTML* code fragment, not an
XHTML fragment.

Will the link checker download the file of the link when I am using GET ?

Hi all ,

    I am now using 3.6.2.3 and is using the method GET. I think that it 
would download the whole file when it check each link (as opposed to HEAD). 
But I use a testing file with an <a href> pointing to a very large file, but 
it can be completed very quickly. So I am not quite sure about if I am 
right. Please advise.

    Thanks.

Perseus

_________________________________________________________________
The new MSN 8: advanced junk mail protection and 2 months FREE* 
http://join.msn.com/?page=features/junkmail