Powerful enough for larger projects?

[AXavierB]

So, I'm new to Lua--actually, I'm new to programming as a whole. After a lot of consideration I decided that this would be a nice language to start out with. Since you have to walk before you can run, I plan on starting out making really small games. However, I'd like to make larger ones down the road. So I know this is probably an incredibly stupid question, but I might as well ask anyway: is LÖVE suitable for considerably large games? By which I mean, games of a scale similar to, say, Cave Story, VVVVVV or LIMBO. I just want to have a good idea of its capabilities before I settle on using it. Not that there are many other 2D game engines for Lua anyway, but I digress. Honestly, I was going to be lazy and just use Construct 2, but that's not even finished yet so there's not much I could do with it.

But yeah. Any answers to my silly question would be much appreciated.

[Robin]

Yeah, no reason why not. Some people here have made really awesome stuff.

[miko]

But yeah. Any answers to my silly question would be much appreciated.

I would say it depends more on your architecture design skills than on lua/love2d itself. Check out this thread:
viewtopic.php?f=4&t=2894
And good luck with your love projects!

[thelinx]

I don't know, you tell me.

[ivan]

CaveStory and VVVVVV wouldn't be a problem at all for Lua/Love but I think a game Limbo might not be feasible without modifying the C++ code of Box2D. Limbo, in my opinion, is way more advanced technologically than either CaveStory or VVVVVV.

I would say it depends more on your architecture design skills than on lua/love2d itself

Couldn't agree with you more.

[Nixola]

I don't know, you tell me.

I simply löve this answer.

[YellowAfterlife]

It's not about LÖVE being powerful enough or not, it's about you, the developer, being good enough (commonly defined by motives) to code a large game in Lua. Writing larger project would require implementing content loading systems, organizing code, and other things that normally require understanding at some point, which is normally gained from working with high-level language... and often afterwards sticking with that for development.

[Jasoco]

LIMBO could be possible as it's mainly physics. It's just the graphical end that would be harder. Most of it is just large PNG files, but things like lighting effects would be the hardest.

Braid however would be pretty much impossible on the graphical side. There's so many particles and warping effects. Even the audio side when manipulating time by slowing or reversing or fast-forwarding. And then the time manipulation which requires you to record the location of every moving object in a level every frame with a constant framerate. I experimented with recording and the longer you record the more memory used up by the table.

[tentus]

... And then the time manipulation which requires you to record the location of every moving object in a level every frame with a constant framerate. I experimented with recording and the longer you record the more memory used up by the table.

Something you could do is use multiple tables and run a compression thread on the oldest one. Say every 10 seconds gets its own table, automatically created and filled as time passes. Once you're in second 30+, start a thread that looks at seconds 1 through 10, and compresses them, probably by means of looking for linear relationships... if an object doesn't move, you don't need much data on them, or if they're moving in an easily interpolated fashion you can cut out data. The point is, the current table is still at full resolution, and the previous ten seconds as well, it's only when you get to the relatively old stuff that you have to start making tradeoffs between memory and fidelity.

[AXavierB]

LIMBO could be possible as it's mainly physics. It's just the graphical end that would be harder. Most of it is just large PNG files, but things like lighting effects would be the hardest.

Braid however would be pretty much impossible on the graphical side. There's so many particles and warping effects. Even the audio side when manipulating time by slowing or reversing or fast-forwarding. And then the time manipulation which requires you to record the location of every moving object in a level every frame with a constant framerate. I experimented with recording and the longer you record the more memory used up by the table.

Like I said, I'm completely uneducated when it comes to programming, but isn't it possible to use a separate graphics engine if the main game engine isn't powerful enough?