Determing whether a bug is a konqueror bug or a site bug.

[Maksim Orlovich]

> I've been trying to help out with the konqueror bug triaging of late.

Thanks for your effort.

> So my question is how to determine whether to close as being poor code on
> the
> site's end, or confirming as being a valid case of a konqueror bug?

The only way to really know is to fully analyze the bug: find out what
difference in behavior is causing the problem (and then make judgement
calls on what's proper behavior, and then on whether broken behavior is
worth emulating). And yes, this requires HTML and JS knowledge, but you
can actually pick it up. Basically, one thing that's often useful is to
save all of the webpage to disk (Opera and Mozilla usually do a good job
of saving full page), and then to try to remove major pieces, trying to
get the page as small as possible in a way that's still showing the
problem.

There is one common class of problematic pages, however: web pages or
scripts checking the browser by name, and often not doing anything useful
for browsers not named "IE" or "Mozilla", and may be "Opera". To test
whether it's likely go to Tools -> Change Browser Identification, and
reload the page a couple times, and see if it helps. Usually faking as
IE6, recent Firefox and Safari 2.0 are worth trying. If it helps, there is
a very high chance the page is doing something that can't be solved other
than by renaming the browser, so you should note it down in the bug.. (One
still has to read the code to figure out whether that's the case, or
whether we don't support enough proprietary extensions, but this is a very
important clue of what to look for)

Hope this helps,
Maks