i usually use the first method you described; a global var maintained by
2 functions, start and stop.
the other way is too complicated imo :)
AdamV wrote:
> I'm writing an AJAXy application and when I start doing an ajax call,
> I want to block the UI and show a loading graphic until the loading is
> done. For various reasons, a number of different functions might
> initiate a load and a number might signal its end, and in a particular
> sequence two such functions might be called during one "ajax load"
> event.
>
> As I was thinking about opimizing JavaScript anyways, I figured I see
> if anyone has any thoughts about optimizing the general situation
> where you have a "start" function that might be called once its
> started and a "stop" function that might be called when nothing is
> currently running.
>
> The trivial way to write this would be:
>
> var running=false;
>
> function do_start() {
> if (running) return;
> running = true;
>
> // Do stuff;
> }
>
> function do_stop() {
> if(!running) return;
> running = false;
>
> // Do stuff;
> }
>
> However, the following would also work:
>
> function nullFunct() {};
> function startFunct() {
> // Do Stuf;
> do_stop = stopFunct;
> do_start = nullFunct;
> }
> function stopFunct() {
> // Do Stuf;
> do_stop = nullFunct;
> do_start = startFunct;
> }
>
> var do_start = startFunct, do_stop=nullFunct;
>
>
> Does anyone have any thoughts about taking either approach? I would
> suspect that the later approach gets more useful as the test for
> "should we allow stuff to happen here" gets ore complicated than I've
> shown.
>
> Adam
>
>
