I don't understand this bug at all.
For example, browser has.. <command id="Browser:BackOrBackDuplicate" oncommand="BrowserBack(event);"/> I want to trigger the command with a specific event such as a mouse click with certain meta/shift/etc keys set true. So if this bug is fixed, one would be able to do something like cmd.doCommand(keyEvent); cmd.doCommand(mouseEvent); instead of doing oncommand = new Function("event", cmd.getAttribute("oncommand")); oncommand.call(cmd, keyEvent); oncommand.call(cmd, mouseEvent); So maybe something like.. void nsIDOMXULElement::doCommand([optional] in nsIDOMEvent evt) ?
.oncommand gets called with command event, not with keyEvent or mouseEvent. So if doCommand is extended, it should take nsIDOMXULCommandEvent, not nsIDOMEvent
> I want to trigger the command with a specific event such as a mouse click with > certain meta/shift/etc keys set true. Ah, what you actually want is to have the modifier state initialized. > So maybe something like.. > void nsIDOMXULElement::doCommand([optional] in nsIDOMEvent evt) ? Yes, and nsIDOMXULCommandEvent should be extended to have the four modifier properties available.
(In reply to comment #4) > Yes, and nsIDOMXULCommandEvent should be extended to have the four modifier > properties available. It already does! You might want to add the button property though?