I'm crowdfunding my newest game: reTux (new Mario-inspired platformer)

[levi]

FWIW, I'm not so sure you need a strong basis in CS theory before you code your first significant project; Big O notation wasn't something that was taught to me at university (or, if it was I completely didn't see the point of it at the time), but after coding a few project professionally, the idea was introduced to me, and it seemed like a useful way to describe the sort on inefficiencies I encountered in my own and others designs.

But if you're coding an engine you hope others are going to use, having the sort of understanding that comes with understanding and using these terms is pretty essential IMO.  On the other hand, for a one-off like ReTux as long as you get get it to run at a decent speed on your target hardware without hogging the system, that's generally good enough for me.

 

[fusion_power]

I appreciate the fact alone that onpon4 is willing to make such an project  and asking for assistance  of the crowd. It may be technicaly a long way but I already see progress. He wrote he already doubled the FPS, that may be a good sign.

Of course for an 2D sidescroller, the performance should be much higher, but this does not come from nothing. If he keep going and continue to optimize his engine, it may turn out well and even add some valuable knowledge. I had also coding education but I never could do an game engine at all. I have more talent in everything graphical.

So I find it even more impressive, how well games can be optimized if needed. When I look at all the old Computers and Consoles and their CPU power it's a miracle how great the  games run there. Look at classic Sidescrollers on 8/16 Bit machines with maybe 3-4 MHz CPUs or so. They ran at 50/60FPS, including multi layer parallax scrolling, alot of onscreen effects, sometimes huge sprites (NeoGeo) and so on. How the hell did they do that - "nobody knows". 

So  there is room for improvements  here for sure, if such games run on the  Pandora without any issues and even through "emulation layers". But at least we know its possible, thats good I think. 

 

[_jr_]

I don't agree that one needs any part of CS to be a good programmer. I do like CS (and have got the grades to prove it), but there are just as many crappy programmers with CS grades than without. Sometimes when hiring it seems to me that higher formal degrees imply lower programming skills. CS presents lots of ways of thinking about problems (and I'll bet at least one of them is close to how onpon4 thinks about software). While these are all valid and valuable, knowledge of none is strictly required to programmatically solve a problem.


And again: someone presents a mostly functional prototype and asks for funds to work fulltime on finishing and polishing it. Where is the problem? Because onpon4 admits issues? Really?

 

[bzar]

So I find it even more impressive, how well games can be optimized if needed. When I look at all the old Computers and Consoles and their CPU power it's a miracle how great the games run there. Look at classic Sidescrollers on 8/16 Bit machines with maybe 3-4 MHz CPUs or so. They ran at 50/60FPS, including multi layer parallax scrolling, alot of onscreen effects, sometimes huge sprites (NeoGeo) and so on. How the hell did they do that - "nobody knows".

They did that with 1) dedicated hardware for this purpose, 2) lower resolutions and bitdepths meaning orders smaller pixel data size, and 3) tricks that work in specific situations, are bug-prone and are not suitable for any kind of engine.