We’ve just released our new game, Crabitron. It’s a game that we’ve been developing since before I ever imagined Codea (it’s taken that long), so it’s not made with Codea.
(@John and I will be making some games with Codea soon, now that it’s so easy to export native apps. Apple will hopefully approve 1.5.2 soon as well. It has moved from “Waiting for Review” to “In Review,” but Apple has told us it will require additional review time.)
One additional thing we’ve done is to make all our sales data for this game public. We’ve started up a site called Crabstarter — think of it as a “reverse Kickstarter” campaign. I know some of you here are interested in how well (or poorly) App Store games do. So on this website you can see our sales figures — pulled automatically from daily iTunes reports:
Some of you have asked us to add extra in-app purchases or donation buttons into Codea so that you can support us, while we are very flattered by your comments, we would love it if instead you took a look at Crabitron. If you like the look of it, then back our project to support us.
Out of curiosity, could Codea handle something like Crabitron? If you guys could post some native vs. Codea speed performance results sometime, that’d be great. Thanks!
@Zoyt we believe so — now that Codea has shaders, a lot of the effects Crabitron uses could be recreated.
The largest potential slow down would be processing the abundance of physics contacts and joints in Lua (not the rendering). So when you have tons of objects colliding on screen, I imagine that’s where things would start to slow down.
It would be a pretty huge thing to manage in one Codea project, though.
Regarding Stack3r, that is something @John and I would like to export using the new Codea feature.
Congratulations @Simeon and gang, insta-purchase for me
Wrt physics simulation and slowdown…have you guys ever taken a look at Chipmunk? I understand the pro version (while a little pricey) has some major optimizations that take advantage of iPad hardware, and it also has several great features that Box2d does not. I know at this point it’s probably not realistic to swap out the physics backend of Codea, but I thought it might be worth mentioning, and am interested in hearing your thoughts on it if you have evaluated it (I’ve only use Box2d myself thus far).
@toadkick it’s not so much about the solver speed — that is very fast. Just simulating the physics is fine in Codea. But when the solver hands you a whole lot of contacts every frame, and you iterate through them in Lua and analyse joint forces (to detect breakage) and other things, then you start to push the limits of running an interpreted language on an iPad. It’s similar to how image processing (iterating and mutating every pixel in an image) can be quite slow in Lua, but quite fast to do on the GPU or in native code.
I believe Crabitron could be made in Codea, but my biggest uncertainty would be the processing of physics results when a lot of objects are on screen.
That makes perfect sense. I wouldn’t ever consider developing a physics based game in Codea/Lua for the reasons you state, and actually I think it would be quite risky from a development standpoint, because of the probability of running into a performance wall that can’t be overcome…native would definitely be the way to go. Fortunately there are a plethora of other types of games that without a doubt could be made in Codea
Clearly a lot of time spent making the balance of the game right and the artwork is terrific. Combine that with the unique use of the touch screen, and this has all the makings of a hit.
I agree, it’s fun and very humorous, I like it a lot. Best of luck to you guys, hopefully it will rise above the cruft in the App Store and make you guys enough to work on Codea full time
@wave you can use all the features of Codea 1.5 when you export your project to Xcode with 1.5.2. You can even write native addons in a clean way using the addon API.
