LÖVE

I'm not really interested in making Mario. I'm more interested in doing a take on N the Way of the Ninja. Or Meatboy.

Heh, neither N nor Meatboy use Box2D, but a custom (and probably a much simpler) platformer engine. You generally don't use rigid body physics for platformers unless you plan to manipulate rigid body physics in-game, otherwise I don't think it's really worth it.

Hm, there is a stunning lack of general platformer examples in LOVE, though. Quite odd; platformers are usually the first examples/frameworks people make with game engines. And once the game engines in question get popular, you see ten of them pop up a day. SHOULD I BE THE FIRST TO CREATE ONE FOR LOVE? :

Also FYI, the core game frameworks of Mario, N and Meatboy are probably all very similar.

Skofo wrote:Hm, there is a stunning lack of general platformer examples in LOVE, though. Quite odd; platformers are usually the first examples/frameworks people make with game engines. And once the game engines in question get popular, you see ten of them pop up a day. SHOULD I BE THE FIRST TO CREATE ONE FOR LOVE? :

Are you volunteering? Then yes!

Skofo wrote:SHOULD I BE THE FIRST TO CREATE ONE FOR LOVE? :

DO EET!

Skofo wrote:Heh, neither N nor Meatboy use Box2D, but a custom (and probably a much simpler) platformer engine. You generally don't use rigid body physics for platformers unless you plan to manipulate rigid body physics in-game, otherwise I don't think it's really worth it.

Hm, there is a stunning lack of general platformer examples in LOVE, though. Quite odd; platformers are usually the first examples/frameworks people make with game engines. And once the game engines in question get popular, you see ten of them pop up a day. SHOULD I BE THE FIRST TO CREATE ONE FOR LOVE? :

Also FYI, the core game frameworks of Mario, N and Meatboy are probably all very similar.

An easy-to-use framework with, perhaps, a level editor would benefit the community greatly. I personally believe that you can do a lot of cool things with a physics engine in platformers but also introduces some complications. You are free to take any parts that you want from Lunar Penguin, e.g. the code for computing distances.

Skofo wrote:Mmm, platformers in Box2D are tough to get just right.

I haven't really touched the physics much (or at all), but it seems that if all you're going to do is to make a guy jump and not go through boxes, wouldn't invoking the gods of physics be a little too much?

mike wrote:

Skofo wrote:Mmm, platformers in Box2D are tough to get just right.

I haven't really touched the physics much (or at all), but it seems that if all you're going to do is to make a guy jump and not go through boxes, wouldn't invoking the gods of physics be a little too much?

The issue is the jumping which should be possible only if you are on the ground or similar. Basically, there needs to be a function that tells whether an object is resting on the ground. More specifically, returns a shape/object that a shape/object is touching (is close enough), and the direction. Imagine having a platform and walking past the edge. After you have left the platform, you should not be able to jump. If rude added the callback for when objects stop touching, then there's a fairly simple solution to this, just keep track of all objects that are being touched or was touched some time ago. Today, there seems to be a few solutions:
1. Fake it by keeping track of the x and y-ranges of every platform and go through them to see if you can jump. Each platform needs to be axis parallel to do this and a tile-based engine might easily work this way.
2. Keep track of the last time you landed. This has the unwanted effect that you can only jump if you just landed on a platform and that you can jump some time after falling off.
3. Explicitly compute the distance to the nearest object (this is probably the most expensive and is what I do. it has the added bonus that you get the normal of the surface you jump from. it is not certain that the untouching-callback will be able to achieve this straight of out of the box).
4. Simulate it with physics - e.g. add a leg to the body and have it push out with a high impulse to produce a jump. (the simplest solution but possibly hard to get right)
Any other ways?

There's always the whole idea of checking the physical position against an internal map. Basically just using the physics engine to move the character and the love engine to run the rest.

I know this is late, and this topic has gone a bit wayward. I read this a few weeks ago when I had first come across Love. At first I was a little worried about this thread. I started to think maybe this wasn't for me, as I do believe this engine has superb potential but it will take time for it to mature into the full article. I did look at other engines but I kept coming back to love because there was just something about it. Love is capable of a lot more than people intially think, it just takes quite a bit of work and bit of imagination. I think if the community knuckles down, we could see some games of decent quality appearing, and this can only be good for love. I used to visit moddb.com reguarly, and I noticed that love doesn't seem to be on there. Moddb is good for putting your game/mod out there, the site is also there for freeware games too, and I have seen a lot of promotion for open source engines. There seems to be a pretty decent number of people on there who love the freedom of standalone games. When my game is further along in production I will post it up on there In hope to get a bit of publicity for both my game and Love.

As for the Love vs Flash debate. I can see where Mike was coming from, but I don't really think you should be comparing Love to flash. Over the years I have seen so many flash games. Some have been great, but a lot have been quite poor. While you can do some things really quick in Flash, I would much prefer to use Love, as there is a feeling about it, that Flash doesn't have. Not just in using Love, but here in the community too. I have become quite addicted to checking the forum. Some people have posted some really good work, and I love how everyone is supportive of other members, giving tips and advice to everyone.

Sorry for the long post which is quite disjointed in places too, but I just want to share some thoughts.

I think LÖVE is necessary. Lua is a great language and is really needs a good 2D lib. I'm a Python (wannabe) programmer and immediately started to learn Lua just to try LÖVE and make few small games.