|
If you drive into an under-ground parking garage in
a truck, all the while saying "it's really a sport's car" and you get your truck
stuck, is it the fault of the garage designers? Or your own damned fault for
trying to pass your truck off a a sports car?
----- Original Message -----
Sent: Tuesday, March 13, 2001 6:36
AM
Subject: Re: [Dynapi-Dev] [
dynapi-Patches-407882 ] Browser.js Recommended
The reason that the code was submitted in the first place
is because I found
that when using the API on a Windows 98 machine after
installing Opera 5.02 I
was able to spoof my way past the browser sniffer
that is used by the API.
If I build a site and want that site to
completely and totally ignore Opera
browsers because they don't support
the API then I don't them to slip by due
to a crappy browser sniffer.
The code that was submitted didn't test for
anything new. It
only improved the way in which the testing was done in
order to get a more
accurate reading. The code that is used has
COMPATIBILITY issues and
should be fixed. Not doing so will most likely
cause a number of
bugs due to the browser not being detected properly in the
first place.
So it is not a waste of time, it is not too much code, it is
not the
Netscape ultimate client sniffer because we don't need the ultimate
client
sniffer, we just need to know if the browser we are dealing with is NS
or
IE and if it's on Mac or Win32 which is what the original submission will
determine. So in review, submitting a fix to a major bug in the core
of the
API is being rejected based on the fact that it does better
detection of the
browsers that we do support. We are in turn
refusing to ensure COMPATIBILITY
of the code and are ensuring failure of
the code.
Once again, just my point of view.
Michael Pemberton
wrote:
>I'm afraid I have to agree with Doug here.
>
>The
reason for detecting the various browsers is for COMPATIBILITY. It is
not
>intended to be used as a giant test to see which variant of a
spoofed AOL
>browser they are using.
>
>If you can find
an example of the API working on a specific platform that
>requires a
few browser specific fixes, send em on in. THEN, and ONLY THEN is
>there any point in detecting the browser. Otherwise we are just
simply
wasting
>our time detecting which non-compliant browser they
are using and they won't
be
>able to see your page anyway.
|
