Discussion of The W3C Validation Service.

David Dorward | 2 Dec 2005 00:22

RE: Error Message Feedback


On Thu, 2005-12-01 at 15:15 -0800, Brian MacLeod wrote:

(Not trimmed as much as I would normally for the benefit of the mailing
list, which I'm CCing this thread back to, my response is below)

> My SQL database has some material generated from a PR feed.  This includes
> some characters such as curly inverted commas and hyphen characters that
> fall into some black hole (I'm still not up to speed on the ASCII/ISO/UTF
> thing yet but do understand that there is a block of problematic characters
> in the greater-than-3FH range).  Turns out that one of these characters was
> not purged on entry into the database and was showing up on the problem web
> page in the brief overview portion of the PR material.  Once I corrected the
> character, the Validator gave me good results. 
> 
> Validator problem
> 
> The presence of this hyphen character cases the Validator to baulk and give
> the non-UTF-8 error message I attached to my earlier e-mail message to you.
> This is not accompanied by anything other than the line reference to where
> the offending character was placed.

I _think_, that the problem with showing the source is that the software
would have to display the character, but since the character doesn't
exist in the specified character encoding, it isn't really a character -
its just some random bytes. Random bytes don't map to anything that
_could_ be displayed.

--

-- 
David Dorward                           <http://dorward.me.uk/>

(Continue reading)

headers

David Dorward | 2 Dec 2005 09:18

RE: Error Message Feedback


On Thu, 2005-12-01 at 15:52 -0800, Brian MacLeod wrote:
> Assuming that non-characters could be small (as in my case where it was an
> out-of-range character encoding) or very large (e.g random bytes introduced
> in quantity), just show a small sample from the beginning of the problem
> area.  Something like 6-8 bytes.  The appropriate Validator output would
> then be:
> 
> <valid UTF-8 characters>
> 
> <initial 6-8 bytes or invalid material>

But then the validator output would not be valid.

> With an output like this, the ability to debug would be preserved without
> placing much burden on the Validator tool.  In my case where the output line
> number bears no relationship to the source, I would have been able to
> pinpoint the problem in a heartbeat.

Well, the output line number does bear a relationship to the source. You
can always generate the markup from your script and examine that.

> PS I'm new to W3 discussions.  Where do I find the "thread" and other
> discussion?

http://lists.w3.org/Archives/Public/www-validator/

--

-- 
David Dorward                           <http://dorward.me.uk/>
"Anybody remotely interesting is mad, in some way or another."

(Continue reading)

headers

David Dorward | 2 Dec 2005 21:09

RE: Error Message Feedback


On Fri, 2005-12-02 at 11:28 -0800, Brian MacLeod wrote:

> I'm only suggesting that the Validator output this as part of an error,
> similar to what it does when it discovers an HTML error.  It outputs the
> incorrect HTML and flags the file as invalid.

Actually, it outputs an HTML representation of the source code to the
file.

Doing something similar for bytes which can't be interpreted as
characters isn't so simple.

--

-- 
David Dorward                           <http://dorward.me.uk/>
"Anybody remotely interesting is mad, in some way or another."
                             -- The Greatest Show in the Galaxy

headers

Djeepy46234 | 3 Dec 2005 02:22

Error Message Feedback


Hi

I know a lot of people all around the WWW use PHP on their website. The 
only problem is when there's a PHP error on a page, the page is 
automaticaly not valid because the Warnings go out of the !DOCTYPE 
declaration.

It would be great to allow PHP errors to go out of the !DOCTYPE to 
prevent those kinds of error.

Thanks
Jean-Philippe Monette

headers

Eric Sabo | 2 Dec 2005 16:19

[VE][79] Error Message Feedback - Head tag not opened error


Validating Error [79]: "end tag for element X which is not open"

[ I do believe there may be a bug in the validator. It is not finding
my <head> tag to be opened and so spits out error 79. Please see my
code - keep in mind I am prevalidating a template so the things in
brackets are the tags. I thought that they may have been confusing the
validator so I tried it on a fully rendered page and it gave the same
error. It also gives an error that the <body> tag doesn't belong where
it is, but I think that may have something to do with the <head> tag
issue. Here is my code

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
"http://www.w3.org/TR/html4/loose.dtd">
<html>
<head>
<title>[title]</title>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<link rel="stylesheet" href="templates/[template]/layout.css" type="text/css">
[scripts]
</head>
<body>

thank you in advance, and please, tell me if its my problem and not
yours! I can't seem to figure out why, for the life of me, Im getting
that error.
]

--
Eric Sabo

(Continue reading)

headers

Jukka K. Korpela | 4 Dec 2005 08:39

Re: [VE][79] Error Message Feedback - Head tag not opened error

On Fri, 2 Dec 2005, Eric Sabo wrote:

> [ I do believe there may be a bug in the validator.

There are bugs in the validator, but this isn't one of them.

> Here is my code

Is it your real code? It looks rather odd. Please post a URL
in future.

> <link rel="stylesheet" href="templates/[template]/layout.css" type="text/css">
> [scripts]

The string [scripts] is character data outside tags, so by HTML rules,
it will be taken as content implicitly wrapped inside a <body> element.
This terminates a <head> element, so any subsequent </head> tag is
an error.

My crystal ball whispers that you didn't post your real code and that the
real code has some <script> elements there and a real <body> element.
Besides, there's probably a "/" before a ">" somewhere, meant to make a 
tag self-closing in XHTML manner but actually making the document invalid
by HTML rules.

--

-- 
Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/

(Continue reading)

headers

Eric Sabo | 5 Dec 2005 05:02

Re: [VE][79] Error Message Feedback - Head tag not opened error

There are bugs in the validator, but this isn't one of them.

Is it your real code? It looks rather odd. Please post a URL
in future.

Yes, it is the real code I was submitting to the validator. I was prechecking a template I wrote for a CMS am in the process of programming. You are correct in that some <script> elements are stuck there when its processed. I ran it on my site (beta.skyphix.com/skyphix.com) which has a million other errors because its still very early beta and the dynamically generated code isn't very clean yet, and still got the error I was getting. I found some eronious poorly generated code that I stripped out and fixed the issue. The [script] was, indeed, confusing the validator and causing it to error out.

Thank you for your follow up. I guess I was suffering from some frustration and should've have prematurely claimed a "bug". It seems as though walking away from my PC and coming back to it would've solved my problem.

--
Eric Sabo
518-791-0068

headers

Nicolas ARGYROU | 5 Dec 2005 12:24

Error Message Feedback


Hello,

This is actually a suggestion for a minor addition to
the Validator: to add the Title of the page to the
Outline.

I often use the validator and I found it very useful.
When I want to revalidate pages to check if everything
is still ok, I usually don't need verbose output, but
I often need to be sure that the correct page was
checked. Adding the Title tag to the 'ouline' option
would save bandwith in those cases and still allowing
to quickly check that the correct page was validated
(even if no H1-H6 tags are present in the page).

This is only a small suggestion, although it seems
important to me since I have a personal web server
with low bandwith and dynamic IP. I assume that this
small option will nonetheless be useful to many people
that are using regularly the Validator.

Best regards,
  Nicolas Argyrou

		
__________________________________________ 
Yahoo! DSL  Something to write home about. 
Just $16.99/mo. or less. 
dsl.yahoo.com

headers

Gerard Saunders | 6 Dec 2005 06:47

validator fails to check htaccess protect pages...


Hi, We have installed the W3C Markup Validator locally however
we are having issues checking any webpage protected by a .htaccess
file. There are no issues with non-htaccess protected pages.

If we check the same pages via the validator.W3c.org checker, the
authentication is successful and the page checked.

This give the impression that the auth details are being passed
to the server which holds the page being checked...

All of our htaccess files use the "Basic Authentication" scheme, 
which from what I have read is supported. 

Does anybody have some advice why one method would work (validator.W3c.org)
and and not the other (local)? Are there any obvious http settings that would
cause this to fail?

I have tried checking pages using the http://username:password <at> host/page.html
and this works with our local version, however is not really an acceptable 
alternative.

Gerard.

headers

Robert Hughes | 6 Dec 2005 03:38

[VE][105] Error Message Feedback


Validating http://theorniphile.info/
Error [105]: "an attribute specification must start with a name or name 
token"

I get this error message when I use javascript to scramble e-mail 
addresses. I've tried rewriting the code to make it work with the 
validator but no go. Here's an example of the code I use:

<p>Send comments to me, <script language="javascript" 
type="text/javascript">
var start = "<a href='mailto:";
var name = "rhughes.enteract";
var at = " <at> ";
var url = "rcn.com?subject=re:Your birding Web site'";
var title = "title = 'E-mail me'>";
var link = "<span class='bold'>Robert Hughes</span>";
var end = "</a>";
document.write(start + name + at + url + title + link + end);
</script>.
</p>

Any suggestions?

Robert Hughes
Chicago, Illinois

22	May 2013
79	April 2013
74	March 2013
105	February 2013
18	January 2013
29	December 2012
40	November 2012
50	October 2012
41	September 2012
59	August 2012
66	July 2012
58	June 2012
33	May 2012
27	April 2012
41	March 2012
12	February 2012
28	January 2012
35	December 2011
46	November 2011
80	October 2011
56	September 2011
69	August 2011
62	July 2011
72	June 2011
52	May 2011
50	April 2011
58	March 2011
45	February 2011
62	January 2011
54	December 2010
68	November 2010
39	October 2010
38	September 2010
70	August 2010
47	July 2010
61	June 2010
50	May 2010
81	April 2010
71	March 2010
99	February 2010
63	January 2010
16	December 2009
46	November 2009
40	October 2009
58	September 2009
65	August 2009
47	July 2009
90	June 2009
70	May 2009
55	April 2009
63	March 2009
78	February 2009
83	January 2009
111	December 2008
50	November 2008
72	October 2008
112	September 2008
71	August 2008
72	July 2008
92	June 2008
124	May 2008
185	April 2008
118	March 2008
89	February 2008
77	January 2008
58	December 2007
106	November 2007
157	October 2007
180	September 2007
191	August 2007
213	July 2007
143	June 2007
149	May 2007
181	April 2007
104	March 2007
109	February 2007
111	January 2007
90	December 2006
91	November 2006
124	October 2006
105	September 2006
137	August 2006
111	July 2006
58	June 2006
82	May 2006
95	April 2006
66	March 2006
122	February 2006
94	January 2006
62	December 2005
136	November 2005
135	October 2005
113	September 2005
145	August 2005
140	July 2005
187	June 2005
158	May 2005
167	April 2005
195	March 2005
157	February 2005
170	January 2005
121	December 2004
208	November 2004
186	October 2004
177	September 2004
243	August 2004
223	July 2004
203	June 2004
371	May 2004
202	April 2004
113	March 2004
147	February 2004
138	January 2004
63	December 2003
99	November 2003
107	October 2003
163	September 2003
184	August 2003
198	July 2003
169	June 2003
124	May 2003
175	April 2003
136	March 2003
127	February 2003
150	January 2003
222	December 2002
289	November 2002
289	October 2002
109	September 2002
183	August 2002
132	July 2002
80	June 2002
156	May 2002
11	April 2002
1	January 2002

Mon	Tue	Wed	Thu	Fri	Sat	Sun

			1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30	31