RE: Possible Validation Error

Hi Oliver:

Thank you very much for your gracious reply.

I have attached a simple text file that will only tentatively validate,
again because it interprets our http-equiv="Content-Type" as
http-equiv="Content-Typ!".  Please note that I used a simple text editor to
strip this entire line of code, before retyping it, to eliminate the chance
of their being anything else in there, besides "e".  

I cannot duplicate the variable response mentioned below, so I may have been
mistaken about sometimes getting a "tentative" reply and other times getting
a completely OK message.  However, I am so sure that I was getting variable
responses that I will keep trying to get documentation of this anomaly for
you.

One thing is for sure: for months I have been using the strict x-html
validator on dozens of files, and they all had been just fine.  Perhaps it
is only coincidental, but it seems that since the validation site itself was
redesigned, my files are now getting "tentative" validation.

Please try to validate the attached file and let me know if it is working OK
on your end.

Many thanks,

Bob
303-844-3313 



-----Original Message-----
From: olivier Thereaux [mailto:ot@w3.org] 
Sent: Wednesday, June 23, 2004 6:53 PM
To: Sullivan, Bob RO Denver
Cc: 'www-validator@w3.org'
Subject: Re: Possible Validation Error


Hi Bob,

On Jun 24, 2004, at 7:12, Sullivan, Bob RO Denver wrote:
>  It is interesting to note that the source listing (in line 7) does 
> not show <meta http-equiv="Content-Type", but rather <meta 
> http-equiv="Content-Typ!" [emphasis added].

Could that be an encoding issue? That your editor displays "e" but 
something else is actually in there, that the validator dislikes. It 
may be a validator bug, but without a sample document, we can't 
conclude anything.

> This is the result that I am getting; please note that it identifies 
> the DOCTYPE and the Encoding that I used.  Nevertheless, it says that 
> no encoding was found!

The fact that it says no encoding was found is a consequence of the 
above, since it could not parse the content-type part. Hence the 
validator does not find the encoding used, it just happens to fall back 
to utf-8.


>  Stranger still, sometimes this same file validates just fine, i.e., 
> without a "tentative" warning.  Please note that I cannot detect a 
> pattern or any other reason why the same file is sometimes OK and why 
> it sometimes give me the "tentative" warning. 

Now that sounds strange indeed. Would you be able to send the address 
for a public document that shows these symptoms? Alternatively, if 
you've been using upload, send me the file, possible stripping out the 
bulk of the content if there is any sensitive data?

Thank you.

-- 
olivier