Runnables!

[ColonelHedgehog]

What is Sponge planning to do in the way of Runnables? BukkitRunnables are very useful in Bukkit, so I think Sponge should make something very similar.

Example:

new SpongeRunnable()
{
@Override
public void run()
{
//Do stuff
if(something)
{
cancel();
}
}
}.runTaskTimer(module, 10L, 10L);


What do you think?

[IDragonfire]

Proper Task-based scheduling in Sponge

I’d like to see proper task-based scheduling in Sponge. Tasks are a super powerful abstractions on top of a scheduler for defining work that needs to be done in the future. Bukkit had primitive support for tasks, but it lacked a continuation mechanism for invoking another task at the completion of a first. Where task continuations become particularly powerful is when switching threads to do work outside of the game loop and then returning control to the game loop to apply the results. Here is a…

Formal threading model for the API

The MC server has basically a single “main” thread and lots of other minor threads. Almost all events occur on the main thread (except those that don’t) and similarly almost all methods in the API had to be called from the main thread (again except those that didn’t). In future versions, worlds might be handled on their own thread. Glowstone already does this and there has been suggestions that MC might go that way too. If we say that calling any API method from inside an event handler is al…

Using an Rx-inspired event system

That’s a pretty cool idea, but it’s not what I was thinking. The API seems a bit clouded. I should explain a bit more. By default the observable is ‘executed’ using the game loop scheduler. This schedules the observable to compute in the same tick, and is a good behavior for most cases. There should be a Scheduler interface. One for the game loop(e.g. server.scheduler.loop), one for ‘asynchronous’ computation server.scheduler.async that is outside the game loop, and finally a delayed game loop…

[teozkr]

http://docs.oracle.com/javase/8/docs/api/java/lang/Runnable.html

No point making your own clone when it’s in the stdlib already.

[coaster3000]

They mean the helper scheduler functions. They don’t mean their own actual interface.

[teozkr]

Ah, sure. That sounds more reasonable, though I think the helper methods should be part of a semi-global context object instead of on the runnable itself.

[coaster3000]

They are more like abstract classes. The stuff they add are gonna be part of the object. However the method public void run(); must still be implemented for their class to compile / run.

Its only an expansion to the api.

It is how OOP (Object Oriented Programming) works.

[teozkr]

I know how OOP works, but I don’t think it’s the right place to implement it here. For one, because since the runnable doesn’t have context currently, it would either need that provided (meaning a context object is required anyway) or the task scheduler would have to be a singleton, which seems to be against what I’ve seen from the rest of the code so far. Second of all, I believe in separating code manipulating data from code doing stuff with data. Look at how Scala’s Futures work, and how every method that would require scheduling another task takes an ExecutionContext that does the hard lifting.