Do you have a question? Post it now! No Registration Necessary. Now with pictures!
- Posted on
- XHTML authoring and browsers
- Eric Lindsay
November 26, 2005, 6:48 pm
rate this thread
Is this still the right place to discuss pages written in XHTML+XML and
whether browsers treat them as XML or as HTML?
I realise that Internet Explorer doesn't like XHTML pages served as
text/html, but I noticed a trick in the W3C FAQs
to get Internet Explorer to use XML.
Does that trick actually work? (I am on a Macintosh)
I also found an early test page at
which is served as Content-Type: text/xml according to the HTTP Response
Headers. Firefox page information says it is Type application/xml and
the Render mode is Standards compliance mode. Firefox, Opera and Safari
give it the same appearance, but does anyone happen to know whether they
are really parsing it as XML (all three say they can), or still doing
tag-soup parsing as HTML? Is there any way to easily figure out
which they are using? 
If I go to a broken XHTML test page
pgoiffon.free.fr/info/inet/html_ou_xhtml.ex2.xhtml which is served as
Content-Type: application/xhtml+xml; charset=utf-8 then Firefox says
Type application/xhtml+xml and the Render mode is Standards compliance
mode. So there doesn't seem anything to clearly tell whether
application/xml is treated different to application/xhtml+xml
Firefox, Opera and Safari all say the page is broken, as you would
expect from an XML parser. I take that to mean they all really used
their XML parser.
 Thanks for the tag-soup description.
 If you decided to test browsers in XHTML, served as
application/xhtml+xml or in XML, served as application/xml, would the
only way to find whether a browser used XML to be to make a broken page,
and see if each browser reports it as broken?
I found a bunch of text pages at
and it looked like Safari threw XML errors at the top of the page for
application/xhtml+xml, application/xml and text/xml but not for
text/html It say it renders the page up to the first error but actually
renders the entire page.
I thought Opera didn't show an error on any of these (contrary to its
behaviour on the other test page). However what it is doing is
rendering the page (without scroll bars to indicate more content - not
sure when it lost them) and putting the error text at the bottom of the
page. It does not show the error on the pages served text/html.
Firefox shows the error only, except on text/html which it renders.
Re: XHTML authoring and browsers
I would say no because authoring html is not like
authoring anything else. So any DTD that is
not HTML is outside of the topic here.
I think arguing the usability of xhtml or xml is best
left in some more general www-group.
Because I'm too lazy to unclutter your antispam
measures in email there is short feedback for your web site:
It does not render very nicely in small Opera 8.50 window.
Small meaning around 560 x 400 (W x H). Especially the
http://www.ericlindsay.com/airlie/index.htm was awfull.
I hate spamming and I'm not spammer but a
person who does like to communicate.
I'm willing to risk my mailbox flooded with
spam against not getting any sensible feedback.
Re: XHTML authoring and browsers
OK, that seems reasonable. I'm going to be using HTML 4.01 Strict, and
not anything newer, so the DTD questions were just a matter of curiosity.
Ouch! Mess in Opera for Macintosh also. Thanks for advising me about
the giant box problem. Hadn't seen that before, and I was using Opera 4
or 5 on Windows four or five years ago when I wrote those pages.
Border-width: 100%; is the culprit. I guess Safari and Firefox are
ignoring % since it isn't applicable, and Opera is trying to help. I
really have to clean up those pages. I can see heaps of stuff relating
to free web site locations I haven't used in five years. Plus I need to
clean out that giant logo that falls off the side of the page.
- » One GIF: different areas link to different webpages
- — Previous thread in » HTML Authoring Forum