[OpenLayers-Dev] need IE help with new popups (*almost* finished)

Erik Uzureau euzuro at gmail.com
Fri Mar 28 05:49:28 EDT 2008


Dear Dev,
So as you can see here:
http://dev.openlayers.org/sandbox/euzuro/pop/examples/popupMatrix.html
the new popups are almost done and ready. though I am sure a more thorough
testing will show more little issues, there are two that stand out
right now. They
are both problems I'm having in IE6, and I'm hoping maybe someone out there
would be kind enough to help me figure out what the problem is.
1) Divs don't Hide:
See for yourself:
 * open a framed popup
 * close framed popup
 * pan map so that marker is in different quadrant
 * open popup
You will note that the divs from the old popup have not gone away. No good.
The code is in the updateFrames() function at the end of the Framed.js file.
Basically, we loop through a collection of all the DOMElements for the popup
and if they are not corresponding to the current this.relativePosition, then we
call OpenLayers.Element.hide() on them, essentially hiding the div.
{{{
            if (pos != this.relativePosition) {
                if (frame) {
                    OpenLayers.Element.hide(frame.div);
                }
            } else {
                if (!frame) {
                    frame = this.frames[pos] = this.createFrame(pos);
                }
                OpenLayers.Element.show(frame.div);
}}}

...but for whatever reason, in ie6, the div does not hide. I have
tried replacing the call to OpenLayers.Element.hide(frame.div) with
frame.div.style.display = "none"; but I get the same results.

The only way I have been able to get a satisfactory performance is by
actually adding and removing the div from its parent node. That is a
possible solution but doesn't seem like the best one.

If anyone with ie6 and a mind for debugging could take a crack at
that, I would be grateful.


2) Memory Usage

In a much more unpleasant twist of fate, it appears that ie6 is
horrendous with these framedcloud poups. each position it opens adds
20mb to the running memory. If anyone could take a look at that, maybe
run it through drip, that would be awesome. It is a major performance
problem at this point. :-(

Thank you to anyone who can help.

Yours Truly,
Erik



More information about the Dev mailing list