Animated pan calls moveTo with wrong parameters

[openlayers]

The tween in introduced in [6111] seems to call moveTo of OpenLayers.Map with this.moveTo(lonlat, this.zoom, true);. This causes the moveend event to be triggered in each step of the animation.

To prevent this the call of moveTo should be changed to something similar to this.moveTo(lonlat, this.zoom, {'dragging':true, 'noEvent':true});.

[crschmidt]

Previous patch doesn't go quite far enough: try panning on a layer with a vector feature drawn across the edge, and you'll see that it doesn't redraw at the end. Instead of triggering the events ourselves, we should just call a 'standard' moveTo, so that it passes through to the moveTo of the layers with dragging not set to true so that they redraw themselves.

[pgiraud]

Hum, I would agree with your patch but there's still something that doesn't sound good to me. Correct me if I'm wrong (I didn't test it) but it seems like the way you wrote the 'done' callback, 'movestart' will be triggered twice.

[crschmidt]

You're probably right. I guess we could use moveTo with noEvent set to true, and then just trigger the moveend event?

[pgiraud]

This sounds really better. And this matches the first way I wanted this code to be written and the reason why I wanted the Eric's noEvent parameter to go in. I'll check that tomorrow morning.

[pgiraud]

1348-r6196-A0.diff includes my last comments This worked for me in an example registering the movestart, move and moveend events.

[crschmidt]

I can confirm that this works for the use cases I can think of. Thanks for the patch, please commit.

[pgiraud]

(In [6203]) It looked like movestart, move and moveend events weren't triggered in a good way while using panTo (with panTween effect), this is fixed by calling moveTo with 'dragging' and 'noEvent' options, and trigering 'moveend' internaly, r=crschmidt (Closes #1348)