Changing the defaults for HTTP ACCEPT, ACCEPT-LANGUAGE and USERAGENT headers

Dawit A adawit at kde.org
Sun May 8 21:28:23 BST 2011


On Sun, May 8, 2011 at 3:56 PM, Maksim Orlovich <mo85 at cornell.edu> wrote:
> 1) Does this actually fix anything? Did you do any testing?

Well the idea behind it is to have a default string that is as close
to what the other browsers are sending as possible in order to avoid
getting a different results. You can see the changes being made in the
other browsers here:

http://blog.chromium.org/2011/03/ua-string-changes-coming-in-chrome-11.html
http://hacks.mozilla.org/2010/09/final-user-agent-string-for-firefox-4/

This was all started when I was scanning through Konqueror's wishlist
bug reports and run into
https://bugs.kde.org/show_bug.cgi?id=271720

For me it does not matter what we is really done, but making the
strings sent by the two rendering engines in KDE itself as close as
possible to what the other browsers are doing makes a whole lot of
sense to me. Granted we are a little more pediantic and by default do
not send as much information as some of the other browsers might. But
still using similar formatting seems a no brainer.

> 2) 'like Gecko' was added by Apple guys for a very good reason (UA sniffing)

Ok ; so it should be left alone which is what I exactly did.

> 3) It would be incorrect to add xhtml accepts to khtml's accept-type.
> XHTML parsing is buggy, and isn't really worth prioritizing fixing (on
> accounts on basically no-one using it)

Ok. This would be left alone too then.

Regards,
Dawit A.




More information about the kde-core-devel mailing list