<$BlogRSDURL$>
Proceeds from the ads below will be donated to the Bob Wuesthoff scholarship fund.

Saturday, January 08, 2005

Separate And Unequal: No Love For The Lizard At ESPN

Warning: Tech rant alert!

Some time ago, Richard noticed that the number of users to some of the political blogs (Instapundit, in this case) using Firefox is pushing 30%. Indeed, some sites are already reporting near-parity between MSIE and Firefox/Mozilla. While this probably reflects the cutting edge of Firefox usership, the overall trend is nonetheless steadily up. With new IE security bugs seemingly coming out every week, Microsoft is hard pressed to keep up.

So you'd think Mozilla browsers would get a little love at major sites. You know, like ESPN.com.

But, no.

Thanks to the ultra-stupid rules IE lives by, it resolves what to do with a given URL -- render it, save it, treat it as HTML, etc. -- by looking at the suffix of the page. Yup. Everybody else looks at the MIME type. If the server kicks out a MIME type of something besides "text/html" -- for instance, "application/octet" -- but the page has a suffix indicating HTML -- i.e., .htm or .html -- IE renders the page as HTML, but Mozilla, Firefox, etc. freak out and ask the user whether he wants to save the page.

Which I do not.

Thanks to misconfigured servers and the broken client their editorial staff almost certainly is using to browse the site with, ESPN has been about 40% invisible to me over the last three days.

It's not like they're trying to shut me out. And it's not being black in 1948 Alabama and being forced to sit in the back of the bus. Sure, it could be worse: I could be one of the unfortunate naifs who think ESPN covers pro rodeo (the page renders horribly in Firefox).

But. You understand how Rosa Parks got her start.


Comments:
If they were cutting edge people, they would be reading something better than that hack Glenn Reynolds at Instapundit.
 
Ah, but 6-4-2 looks awesome in Moz, and Firefox.
 
"...but Mozilla, Firefox, etc. freak out and ask the user whether he wants to save the page."

Ah-ha! I was wondering what was up with that.
 
Actually, what IE does is much more complicated and nastier.

What it actually does is it downloads the file, then it hands it off to WinInet.dll. WinInet.dll analyses the file, and if it finds characteristics of a particular file type (like HTML pages starting <html>) then it reassigns the MIME-type on the page.

If it doesn't, then it looks at the extension to reassign a MIME-type, and if that fails then it uses the MIME-type the server assigned.

Also, if the server-assigned MIME-type is one of a list, then it won't reassign the MIME-type (generally, this is types like text/html, image/jpeg), but if it's application/octet-stream, or text/plain, then IE's quite happy to ignore it.
 
Thanks for the note. It makes sense that IE would try to puzzle out what a server is doing, but unfortunately, that would just go along with the general brokenness Microsoft engenders.
 
Steve -- thanks for the note. I hadn't tried that, but since I use Mozilla, I'm not sure I could get away with it. I might just end up installing Firefox on this box pretty soon, though, as I ended up using it and liking it very much on my father-in-law's computer.
 

Post a Comment

Note: Only a member of this blog may post a comment.



Newer›  ‹Older
This page is powered by Blogger. Isn't yours?
Google

WWW 6-4-2