W3C home > Mailing lists > Public > www-validator@w3.org > March 2007

Re: there is no attribute "onMouseover"

From: Drake Wilson <drake@begriffli.ch>
Date: Mon, 19 Mar 2007 07:42:20 -0500
To: Melody Chamlee <melody.chamlee@mac.com>
Cc: denis4o@gmail.com, www-validator@w3.org
Message-ID: <20070319124220.GA24893@drache.begriffli.ch>
(Quoted paragraphs reordered.)

Quoth Melody Chamlee <melody.chamlee@mac.com>, on 2007-03-19 08:04:49 -0400:
> I'd suggest making the language more obvious from the perspective of  
> users who are just transitioning off of non-standard html. The  
> validator is a great system, but it is sometimes written in a way  
> that is self-evident to those who know the rules of XHTML already.   
> As this case illustrates it's not always as clear at conveying that  
> rule system to transitioning users, especially those working to  
> understand strict validation.
> There is a lot of information the user could have been overwhelmed by  
> before making it down to the solution that was specific to reducing  
> the error.

I see your point to the extent that someone who reads "there is no
attribute X" and who is used to attributes being case-insensitive.
It's entirely true that the error message could be reordered to
communicate the possibility of incorrect casing better.  Better yet,
fuzzy-match nonexistent tag and attribute names against the DTD, and
display possible corrections, similar to how spelling checkers
fuzzy-match against dictionaries.  If the names match in everything
except case, and the dialect in use is case-sensitive, display a note
to that effect for that suggestion.

> In nonstandard HTML for instance, a browser will display ANY case  
> representation of a tag, and many editors still auto-cap tags to  
> further muddy the best practice that it's better to leave markup  
> uncapped unless there is legitimate reason to do so.
> Javascript is a further confusion in that it does  prefer its  
> attributes to match both in the statement portion of the script and  
> the attribute calls in the HTML.
> Unfortunately there is no perfect solution, only best attempts.  It's  
> hard to reference both Javascript and XHTML best practices  
> simultaneously and still be clear in presenting a solution to the end  
> user.

Getting contradictory information from other sources isn't really
something you can fix in the validator, of course, and trying usually
leads to an arms race of who claims to have more authoritative
information than whom.  As you say, there's no good solution.

Now, I will still note that despite all of this, there is no
indication (that I can perceive) in the original post that the poster
read the error message at _all_---not even an "I read the message but
couldn't understand what it meant".  Perhaps I was overly harsh in my
original reply, for which I apologize if so, but I do think people
should be pointed in the direction of asking informed questions rather
than blind ones.

   ---> Drake Wilson

Received on Monday, 19 March 2007 12:42:29 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:17:51 UTC