WFS markers are drawn when layer is not in range...

Thoughts on this patch?

so i just looked at this and i think the real problem here is in Map.js

in the setCenter() function is the core place where we're doing tests on layers' inRange-ness. As we cycle through the map's overlays, we do this:

                    var inRange = layer.calculateInRange();
                    if (layer.inRange != inRange) {
                        // Layer property has changed. We are going 
                        // to call moveLayer so that the layer can be turned
                        // off or on.   
                        layer.inRange = inRange;
                        moveLayer = true;
                        this.events.triggerEvent("changelayer");
                    } else {
                        // If nothing has changed, then we only move the layer
                        // if it is visible and inrange.
                        moveLayer = (layer.visibility && layer.inRange);
                    }

Which seems wise but really it's the reason for this problem. we check if the layer's inRange-ness *changes*, and if it does, then we automatically call moveTo() on it. Really, we should check if it changes, and if the new inRange is true, then we should moveTo. If it changes and the new inRange is *false*, then we should simply call display(false) on the layer.

no? thoughts? sorry i dont have time to make a patch for this.

moveTo is used to control the visibility. Without calling moveTo, the layer will remain displayed. Changing that is a significant change at the base level of the layer class: I think that the actual right solution is:

• If inRange changes and is now false, we should call setVisibility(false) from within Map.moveTo
• We should change the moveTo functions to no longer be the one to call display(false) on the layer -- this needs to be done in the Layer.js class, I think.

If we do this, so far as I can tell, it should work... ... however, I feel uncomfortable doing this change this late in a release cycle.

So, I propose that we:

• Open a ticket to do the above for 2.6 immediately after we close out 2.5 dev
• Make this minor change to the WFS layer for 2.5
• Fix trunk right after the branch.

This way, we hopefully won't be fighting with tons of churn over a tiny bugfix, and we can get more debugging on it later on.

agreed. even though i think the patch needed would be small and simple (and actually correct this problem with *every* layer, not just wfs)... you're probably right and it's probably a can of worms and if this is good enough to fix the problem for the person who opened the ticket, then so be it.

i approve of committing your patch conditional on the creation of a new ticket for 2.6 with all the above detailed.

(In [4069]) Stop WFS layer from requesting data when layer is not in range. (Closes #685) Note that this is not a complete fix, but instead a patch for 2.5: after 2.5, this should be changed/removed when map.setCenter() is fixed. (See #937)