W3C home > Mailing lists > Public > www-validator@w3.org > March 2004

Re: W3C validator - BUG

From: David Dorward <david@dorward.me.uk>
Date: Fri, 12 Mar 2004 16:40:45 +0000
Message-Id: <022C6178-7444-11D8-B7DA-000A957E4F00@dorward.me.uk>
Cc: www-validator@w3.org
To: Herbert Rosenau <hr@dv-rosenau.de>

On 12 Mar 2004, at 16:32, Herbert Rosenau wrote:
> David Dorward wrote:
>> On 12 Mar 2004, at 11:49, Herbert Rosenau wrote:
>>> As sample of an error the validator does itself is:
>>> Line 188, column 75:  reference to entity "Aktion" for which no 
>>> system identifier could be generated
>>>   ...t/cgi-bin/webring.cgi?Ring=428&Id=8&Aktion=Back"
>>> There is NO error in the file - but the validator misinterprets it.
>> Wrong. There is an error in the file, and the validator is quite 
>> right.
>> http://www.htmlhelp.com/tools/validator/problems.html#amp

> Then you would be able to show what error in the file is!

> Line 188, column 75:  reference to entity "Aktion" for which no system 
> identifier could be generated
>   ...t/cgi-bin/webring.cgi?Ring=428&Id=8&Aktion=Back"

And the error is that there is no HTML entity called "Aktion"

> The validator shows in the middle of the parameters given to the CGI 
> script. How can the validator know what kind, names and type the 
> script looks for?

It doesn't need to, that isn't what it is looking for. Try reading the 
reference I gave you: 

> Give a sample to write an URL with parameters that gets interpreted by 
> the validator without crying.

There is one at 

>  The validator will never known what the letters addressed to the 
> script are meaning but whines about. It is NOT the job of the 
> validator to check things a CGI script is the only that knows the 
> meaning.

Correct, but that isn't the issue here. I could say:

"Pull the red lever dwon"

You don't need to know what the lever does to know that 'dwon' isn't a 
word in English and that it should be 'down'.

> Come on with a script that shows an URL that points to a CGI script 
> that needs parameters that can be misinterpreded as an HTML keyword.

http://www.example.com/?print=1&copy=2 - oh no! A copyright symbol!

> There is nothing in W3C,

Nothing except the HTML specification. 

>  perl, PHP, Java, Javascript


>  that forbids a letter combination that looks like a valid HTML 
> keyword.

> Nowhere in the description of the keywords a link the description of 
> an url forbids the usage of any of the letters a-zA-Z in any 
> combination to build an parameter for an CGI script

The problem is with the &, not with a-zA-Z, did you read the resource I 
gave you earlier? 

David Dorward
Received on Friday, 12 March 2004 11:40:49 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:17:40 UTC