ArisKostakos's Forum Posts

  • 2 posts
  • [quote:18ip3f14]what i'd suggest is something better - way better - behviours and plugins should have event listeners for ticking. once tick happens - each object / each behaviour updates globally, therefore no for loops at all.

    this. kinda

    Polling every single condition every single tick is always going to be much slower than an event-driven architecture.

    Here's a small example:

    CONDITION: Compare two Values-> [PlayerSprite.x] Bigger Than [500]

    ACTION: ..

    By polling, you check every tick if PlayerSprite.x>500. An Event-Driven way to do this would be like that:

    Assume BehaviorsTriggeredArray which is always empty at the beginning of every tick.

    The game runs and the player moves to the right. Somewhere deep in the engine there's a MoveSprite() function. When MoveSprite is called, it will trigger a MoveSprite event to which all Conditions that depend on Sprite Positions listen to. This is when the condition will check the PlayerSprite.x>500 expression, and if true, it will add a [Behavior#15Triggered] Object to the BehaviorsTriggeredArray.

    When the engine needs to update, it goes through the BehaviorsTriggeredArray objects one by one, and acts accordingly.

    If our game only had that one condition above, then with the polling method, the engine checks, 'is player X bigger than 500?' every, single time.. Even if the player never moved.. That's a huge overhead that can be avoided. In the event-driven scenario, the engine does absolutely nothing.. Then, when the player starts to move, it will check if it's X is bigger than 500. Which makes much more sense.

    My two cents

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads
  • I'll give it another shot posting myself. I won't post links this time so hopefully it will work.

    [quote:1dn0fyer]time has passed since my post and I am more curious than implicated so no worries, my opinion is not based on actual precise research but about feelings and my experience (which is not directly tied to this).

    No worries, I know what you mean. Many people promising all sorts of things on the internet . A landing page cannot be held credible by itself. We'll release more when we have more.

    glerikud, thanks for the GM's licensing info. I agree it doesn't make too much sense.

    [quote:1dn0fyer]And if you ok with that, I'd also have a few suggestions:

    • Start your own forum very soon. The current one is a bit off and is hard to find.
    • Post updates to your website. Nowadays the users seem to be interested in the promising project's alpha stage too. It would do very good for your engine to let users know that it's still a living and constantly developing project.

    A website + forum are underway.

    Thanks,

    Aris

  • 2 posts