- From: Anne van Kesteren <annevk@opera.com>
- Date: Tue, 23 Jun 2009 20:06:57 +0200
On Tue, 23 Jun 2009 19:51:18 +0200, Joseph Pecoraro <joepeck02 at gmail.com> wrote: >> Jeremy Orlow: Is it too late? It seems as though Joseph's >> suggestion could be in addition to what's already in the spec. > > This doesn't break anything in the current spec. So it wouldn't break > any existing implementations. I'm also guessing that the groundwork > for implementing a feature like this is already in place due to the > ubiquity of addEventListener. To be frank, it seems like a lot of bloat though to avoid a simple comparison. > - Less Listener Functions Fired - Instead of every registered listener > getting fired on every "storage" event, only those applicable will be > fired. This may mean overall less listeners getting fired, and code > that doesn't have to continually check the affected storageArea, > leading to potential performance improvements. Actually, assuming only one of the two solutions would exist the same amount of events would fire. A change to localStorage causes an event to be dispatched and likewise a change to sessionStorage causes an event to be dispatched. Having said that, with your solutions more events will fire since the legacy event will have to be dispatched too. > Cons: > > - This is similar to implementing the full EventTarget Interface on > Storage areas. However, the only event of importance on a Storage > area is "storage" not "click" or "hover" etc. Invalid event types > should probably throw an error or something... or maybe this is more > in the scope of DOM Events. That is still quite a bit of complexity for little benefit. -- Anne van Kesteren http://annevankesteren.nl/
Received on Tuesday, 23 June 2009 11:06:57 UTC