Re: Beta: Fatal Error: No DOCTYPE specified!

Bjoern Hoehrmann wrote:


> * Terje Bless wrote:
> 
>>Bjoern Hoehrmann <derhoermi@gmx.net> wrote:
>>
>>>1) I don't see any good reason to refuse validation completly. It's very
>>>  simple to choose document types to default to
>>
>>This is a conscious choice;
> 
> Conscious or not, it's a bad choice.

I have to disagree on this point, as I think it's a sensible choice. 
The validator can't be expected to guess what the author had in mind 
when they authored the document-- assuming they had much of anything in 
mind regarding the markup, that is.  The manual override option is 
available for authors who want to use it.  For those who don't, they can 
add a DOCTYPE.

Trying to guess what the author had in mind is a sure path to confusion. 
  It's what gives us things like Web browsers that honor 'width: 400;' 
and 'color: ffffff'-- or, to be slightly more on-topic for the moment, 
'<b><i>hello</b></i>'.  I've yet to meet an author who went to use the 
old validator and couldn't figure out, within a try or three, that they 
either needed to modify their document to say what they meant, or else 
pick the manual-override options for document type and character encoding.

I do think that the options to override should be up front, though, and 
not buried on the "entended interface" page.  The validator might do 
well to have a different error message for missing DOCTYPEs, one that 
fills in more details or is re-worded to help non-experts.  I think the 
link to the document that simply contain a list of DOCTYPEs is very 
handy, and could stand to be more prominent.  But I definitely don't 
think the tool should be guessing at what I meant when I don't say it 
clearly.

-- 
Eric A. Meyer ............. emeyer@netscape.com
Standards Evangelist ...... http://developer.netscape.com/evangelism/
Netscape Communications ... http://www.netscape.com/
Author of "CSS: The Definitive Guide" (O'Reilly & Associates)

Received on Friday, 25 October 2002 10:07:06 UTC