On Jul 4, 2010, at 9:53 AM, ext Peter Robins wrote: > I'm rather confused by this discussion. When I first saw Bart's bullet > point, I thought it was referring to map events such as layeradded. > Then when I read the log, I saw much discussion of click, i.e. > browser, events. But surely if the objective of 3.0 is to separate the > ui from the 'core' map functionality, then the map, i.e. the map > object, will be part of the core, whereas browser events like clicks > will be part of a particular (albeit default) ui environment, along > with the various other controls, popups, etc. I don't see how having > map.click.on or whatever is compatible with this. Some ui > environments, such as Tim's melodica :-), won't have clicks, so how > can clicks be core?
Clicks in the 'browser' UI environment map to 'map move left', 'map move right', etc. The map will still have these hooks; right now, for example, doing a 'pan' involves a bunch of hooking into the map information, which a melodica interface wouldn't want to implement. Instead, the map should be able to have something like a: * started moving * shifted right slightly * shifted left slightly * stopped moving Hooks; then the default browser environment hooks up to those interfaces trivially. (Mousedown maps to 'started moving', etc.) You don't want each potential interface to have to reimplement all of the knowledge of what happens to the map when it moves, but you also don't want to hook up to those events at the browser level, like we do now. It's still not clear to me how this actually works in practice, but I think the summary is that right now, map.panTo/setCenter are too 'high level' -- they do a lot of work that is unneccesary if you're writing a dragging control, because they can't be told "turn *off* all that extra calculation until I'm done." By moving that up a level, and having the right hooks in place for callers to refer to, we can improve that situation. Regards, -- Christopher Schmidt Nokia _______________________________________________ Dev mailing list Dev@openlayers.org http://openlayers.org/mailman/listinfo/dev