Is it okay to spoof navigator.product by default?

[Allan Sandfeld Jensen]

On Tuesday 21 September 2004 15:55, David Faure wrote:
> On Tuesday 21 September 2004 15:51, Allan Sandfeld Jensen wrote:
> > Hi khtml people!
> >
> > I have the following patch which "fixes" problems with old versions of
> > opencube's quickmenu (see bug: http://bugs.kde.org/show_bug.cgi?id=74162
> > ) The patch is adapted from the Safari-tree.
> >
> > QuickMenu is used on large number of sites, and the progress to update to
> > newer versions that actually detects and supports konqueror seems pretty
> > non-existant.
> >
> > The question is if anyone would object if I commit the attached patch. Is
> > it okay to pretend to be Gecko, yet another place?
> > Since it is a value IE doesn't provide and one Safari already spoofs, I
> > cannot see the harm.
>
> Well this is going to break every other site that actually has
> konqueror-specific code (and that uses product to detect khtml), doesn't
> it?
> I'm okay with spoofing when the useragent was explicitely set to something
> else, but IMHO by default the product should remain "Konqueror/khtml".

This case basically boils down to whether we want our users to be able to use 
a number of websites by default. 

We know the patch "fixes" some websites, but can you show me any it breaks or 
confuses? The only reason I can see for a website to use Product over 
UserAgent to detect konqueror, with would be to do so reliable even when 
konqueror is spoofing. If we decide to spoof Product to make Mozilla spoofing 
better, we are already breaking the intentions of these websites. 
(I've seen one webdeveloper-site recomming checking for KDE in Vendor as the 
reliable way to detect konqueror)

Personally I only see one problem with this patch; in that it would be a 
better  solution to fix KHTML, so it again supports the Netscape4 ecma-code 
like it  did in the 3.1.x series (this bug is regression). I just dont have 
the skills, and I can't  see anyone else caring.

`Allan