Re: Bero goes bye bye

Jef Spaleta <[email protected]> 27 Sep 2002 10:15:42 -0400
Newsgroups gmane.linux.redhat.release.limbo
Message-ID <1033136172.9031.117.camel@spatula>
On Fri, 2002-09-27 at 00:52, John wrote:
> On Thursday 26 September 2002 22:47, Thomas Dodd wrote:
> > And why won't mozilla work?
> 
> Broken website

So if its a broken website....and you know its a broken website...is it
a good idea for a browser(any browser) to render it?  Where is the
incentive for the webmaster at that site to ever create working
websites?  I mean its great for you that konq can get past the errors in
the website...but i have to wonder how the brokenness of the website is
being fixed by konq since anything that was broken has to be
reinterpreted and by konq and konq has to make a guess as to what what
meant..and once you start guessing about how to deal with errors...that
can lead to worse errors.


> 
> > Are you faking the UA in Konq?
> I haven't tried yet, but I did find the feature.

It would be very interesting to know what UA string Konq is reporting to
the admittedly broken website.  But i think there is another string
websites can check...something about browser version which seems to be
hardcoded as far as I can tell.  I found a website once that could tell
me a whole buttload of info about my browser settings...but I can't seem
to find that site now. Anyways a quick and easy way to see what UA you
have is to use a one line php script on yer php enabled webserver and
just point your browser too it..

<?php echo $HTTP_USER_AGENT; ?> 

I highly doubt konq has a feature the website is using that mozilla does
not also provide and can not get via a plugin...so its more likely than
not the website is deliberately only allowing in certain browsers by
user agent strings or some other setting condition...and konq is
probably sneaking in because its misreporting its UA as IE. And frankly
this really sucks...because when browsers misreport their UA's like this
it strengthens IE's position in any type of website traffic analysis. 
So you get access to you bank...at the price of showing up in statistics
as running an MS browser.  Can you see what konq is reporting as its
UA?  


> > Why not do the same with Mozilla (http://uabar.mozdev.org)
> 
> Because, while I've been through its settings many times, I don't recall 
> seeing such a feature.
Moz doesnt have a UA feature right in its qui settings wizard
normally...you either have to go edit a preference file by hand...or
grab the addon project called uabar from the mozdev.org website which
will give you gui access to the UA changing features.

> 
> Besides that, there's a government website important to me, and K is the only 
> browser of the three able to handle it correctly.
By why is that?  
a)Is it because the rendering engine really does handle it differently
b)is it becuase the website is broken...
c)is it becuase the website is deliberately shutting out specific
browsers via a UA string check?

if the reason is A...then kong gets a point for being better.

if the reason is B...konq gets half a point for being "smart" about
dealing with broken behavior even though its not clear if konq is
rendering the site as intented...if there are errors trying to
reinterpret them to fix than is a fuzzy game.

if the reason is C...whether or not konq gets in is going to be becuase
konq is blatantly lying about what application it is....no points
awarded for liars. And you can set the other applications to lie as
well. But browsers shouldn't have to lie to get website access....so its
not clear if this is the best behavior especially if konq is doing it by
default and you are showing up as IE in web site statistics...you aren't
helping making sure that browser independent standards are used on
websites...which would keep this kind of stuff from happening.

-jef
signature.asc (application/pgp-signature, 232 B)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQA9lGgOrWLDmRitRZURAl2ZAJ98UfMiqOf/1EVkFVShNp8EFOjHIACfai7t
dJgbulabm1tMv9k5Ac4LKJI=
=LYNq
-----END PGP SIGNATURE-----