shaunjs
patreon
Slopes update
🕑 Added 2017-09-14 11:48:31 +0000 UTC So the worst news is I won't be able to finish the slopes video for this week. I'm still not in a position where I have something I'm happy to record let alone edit. I'm not going to promise next Friday either but we'll see how it goes.There will still be "A video" however, not a tutorial most likely.
The good news is I'm making some good progress, and have enroute solved the "issues" with the previous tile system of having a player hitbox, or a movement speed, that exceeds one tile. Now we'll always hit the closest obstacle, and don't run the risk of moving through walls by moving too quickly. Our player can also have as big of a hitbox as you like, as the system dynamically adjusts the rows and columns we check based on the hitbox.
Really the problem with all this is and the reason it's taking a while to provide a strong, "fits as much as possible solution" is that the best solution to "I want slopes" in a tile based platformer is "no actually you don't."
Most platformers that really hinge on having sloped and varied terrain use either a full physics engine, complex polygonal collision like Braid or a bitmask system like Worms. Other platformers get away with only handling certain specific cases they want to employ such as stairs or slides (for which you could even state machine your way through!)
So it goes without saying that no slopes solution for a tile based platformer is going to be perfect, elegant and handle every case. Trying to do that would end up defeating the point of working with tiles, in that keeps a lot of things simple.
One exception to this you could argue is N+ and N++, but that game was coded by wizards whose skills exceed mine and exceed the scope of the kind of tutorials I write!
I think we can at least get something as solid as say, Megaman X and still be able to teach it in a way that works in a video format. That's my goal!
This has steadily turned a bit from "slopes" into an advanced tile based collision system all round, as having that is somewhat ironically the easiest way to implement slopes with as few "what ifs" to deal with as possible.
Will keep you all up to date as I make more progress =)
-S
Comments
Will this update also include partial-tile collision? (half tile, quarter tile, etc)
thanks for the update. I love the transparency. Looking forward to it
