Very strange behavior of XHTML validator

B.H.

Hello, i'm a developer of www.kikarhashabat.co.il (Israeli news site in
Hebrew) and i'm trying to make the code validate properly  as XHTML 1.0
transitional.

Now, i get very strange output for pages that seem to be valid. Example:

URL: http://www.kikarhashabat.co.il/%D7%97%D7%A8%D7%93%D7%99%D7%9D

I get 30 errors all of them on Line 4, Column 80, which is in the middle of
simple innocent meta tag in the <head>
(<met*a*http-equiv="Content-Language" content="he" />). Error messages
look
completely unrelated (attributes construct error, Couldn't find end of Start
Tag link line 4, Couldn't find end of Start Tag img line 4...)

What's even more strange, for page that really have validation errors
(MARQUEE tag, <style> in the <body>) - http://www.kikarhashabat.co.il/ - i
get only 8 errors and they are real errors with right locations shown, but
if i hide all invalid code then all 30 weird errors come back :). Those
pages share the same wrapper template so only inner contents are different.

One common thing that i have noticed is that those errors always occur on
column 80 of the first line after <head> or <body> no matter what's there,
and this makes me think that it's maybe a validator issue. I've looked in
the FAQ, bugzilla and maillist archives but didn't find anything similar.

Thanks for your help!

Mordechay

-- 
Moshiach is coming very soon, prepare yourself!
יחי אדוננו מורינו ורבינו מלך המשיח לעולם ועד!

Received on Wednesday, 10 November 2010 22:48:09 UTC