Detecting When an EDGE Instance is Finished Loading

« Web Commanding Capability | EDGE User’s Guide | Changing Anti-Aliasing Levels »

There are occasions where a script inside of EDGE, or a program external to EDGE, needs to detect when an EDGE instance is up and running. This article seeks to describe a few options to accomplish this.

Inside an EDGE Instance

The simplest scenario is when a user has a script they want to execute once EDGE is finished with init and is up and running. Luckily, this is simple with a simple tcl callback. For example:

doug.callback add timer -delay 1.2 -autoreset 0 { myproc_edge_is_running }

This will call a tcl procedure called "myproc_edge_is_running" 1.2 seconds after EDGE startup. The autoreset 0 parameter ensures that it doesn't get scheduled to be called cyclically - it is only called once.

External to an EDGE Instance

Having an external process detect when EDGE is in the run state could be done many different ways. A few that come to mind:

1. One way is to have a script inside EDGE notify the outside world it is up and running. You could do that by putting a tcl callback similar to the one above (only firing only once and not autoreset) and have that script do something that the outside world would notice. It could connect on a socket, or maybe just drop a file that your external program would see. Of course, the external program or the launcher would probably need to ensure that file is removed when EDGE is launched.

For example:

doug.callback add timer -delay 1.2 -autoreset 0 { myproc_drop_startup_file }

In that case, the proc "myproc_drop_startup_file" would get called 1.2 seconds after EDGE startup.

2. Another way you could do it is to have your outside program watching the log output from the EDGE instance and trigger off of that. For example, with the "-fps" parameter on the command line, it is usually safe to say that by the time the frame rate prints start coming out that EDGE is running. Of course, a tcl script could also just use a "puts" to print your own marker if you used this method.

3. A third way without modifying your EDGE project would be to just connect to the web commanding server and wait till you get a valid response back. Once those web requests respond, then EDGE should be up and running. The resource at the "/status" URI would probably be a good resource to hit and the response would include a heartbeat, the current framerate, and lots of other useful information.

There are probably others, but those are 3 quick ones that come to mind. Options 1 and 3 are probably the easiest to implement.