[OpenLayers-Dev] select-feature.html on FF 2.0.0.3

[Christopher Schmidt]

On Fri, May 04, 2007 at 06:58:33AM +1000, Cameron Shorter wrote:
> Eric Lemoine wrote:
> >On 5/3/07, Christopher Schmidt <crschmidt at metacarta.com> wrote:
> >>On Thu, May 03, 2007 at 03:23:20PM +1000, Cameron Shorter wrote:
> >>> Eric,
> >>> Good that you have tracked down the Ubuntu/Firefox issue, but we
> >>> probably should be tracking in the OL issue tracker too.
> >>>
> >>> Ideally we should introduce a work around in OL as we don't want to 
> >>tell
> >>> all Ubuntu users to go install Windows and wait until the next Ubuntu
> >>> release.
> >>
> >>I can't imagine what kind of workaround we can introduce for "SVG is
> >>broken in Feisty." I think the answer is simply "Download a real
> >>FireFox.", since the one in Feisty is apparently broken in ways that the
> >>official release is not.
> >
> >Agreed. That is why I submitted a bug to Ubuntu in the first place.

> I haven't looked into how broken Feisty SVG is broken. You might be 
> right about it not being fixable. By having the bug recorded, we at 
> least have something we can point users at so we don't have to keep 
> answering the same user questions.

Agreed. I've added it (and a workaround -- download a copy of firefox 
from moz.org)  to the VectorFAQ wiki page. It's my plan to merge this
with the main FAQ at release time. 

> One work around that comes to mind is to use the Walter Zorn vector 
> rendering which is also available for old browsers and would give us 
> vector support in old browsers.
> http://www.walterzorn.com/jsgraphics/jsgraphics_e.htm

No, you're missing the point of 'workaround' here -- that's a new
feature, which solves a host of problems (and introduces a host of other
problems). I believe that the better solution than *that* is to create a
Canvas layer, which Firefox supports and would add Safari support to our
base. However, neither of these are features which make sense in 2.4. 

> Alternatively, we could offer an alert message telling users why the 
> functionality is broken and what they should do about it.

Except we can't detect that it's broken. (And, since switching tabs
'fixes' the display, it's not clear that we're likely to have a simple
test case.)  

Documenting it and explaining workarounds are great. However, I strongly
believe this is not something that we should delay a 2.4 release for -- 
that's all I'm really saying.

Regards,
-- 
Christopher Schmidt
MetaCarta