Closed GoogleCodeExporter closed 9 years ago
I cannot reproduce this in Firefox 3.5 or 3.51 (Windows), nor in Firefox 3.0.12
under GNU/Linux, either 32-bit or 64-bit.
Possibly you have the browser's user-agent string configured to mask its
identity as
firefox? That could provoke this problem, and it raises an interesting issue
about
the way the browser-detect code works at present.
Original comment by c.1%smit...@gtempaccount.com
on 24 Jul 2009 at 6:17
That is an interesting point. I have checked, and the version of Firefox I have
is
reporting:
Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.1) Gecko/20090716 Ubuntu/9.04
(jaunty) Shiretoko/3.5.1
This is because I installed it from the Ubuntu repository. According to Ubuntu
"we only use official branding [Firefox] for our default browser (default for
jaunty
is 3.0) also we explicitly want both to be installable side by side"
I knew the name was different in the GUI, didn't even consider the user-agent
etc !
It might be better to remove the error message anyway and just exit gracefully ?
Original comment by system.a...@strategicbookings.com
on 24 Jul 2009 at 6:39
In r127 I have (marginally) improved the firefox detection code, which should
obviate this error message (although not, obviously, on the curvycorners.net
home
page until the next release).
(As you say, the default Ubuntu Firefox is 3.0.12 and I find it really strange
that
Canonical would tinker with the user agent string as if it were a question of
"official branding". (I would think that was a question for the Mozilla people
anyway.) Oh well, we live and learn. (And whilst on the subject of
bleeding-edge
Linux browsers, Opera 10 is looking extremely good in its 64-bit Gnu/Linux
incarnation.)
You say that it might be better to remove the error message and exit
gracefully.
From your standpoint, as a client programmer, I can well appreciate what a
nuisance
it is to have these noisy error messages popping up and being a pain. However,
from
my standpoint, I much prefer to have early warnings of anything potentially
dangerous in the script. As it happens, the error message that you were picking
up
was benign, but merely served to show up a non-fatal shortcoming in the script.
(You
can suppress these messages anyway, without altering the script, simply by
defining
curvyCornersVerbose as false before you include the script.) I hope that this
policy
has paid off, in this instance, by contributing to a slightly more robust
browser
detection which, had it gone unnoticed, might have caused some extremely
peculiar
misbehaviour.
Thank you for reporting this!
Original comment by c.1%smit...@gtempaccount.com
on 24 Jul 2009 at 11:45
I didn't know about the verbose setting. Perfect.
Thanks
Original comment by system.a...@strategicbookings.com
on 24 Jul 2009 at 11:53
Original issue reported on code.google.com by
st...@sgperkins.org
on 20 Jul 2009 at 3:51