Re: [Super-tux-devel] SuperTux

[Ingo R. <gr...@gm...>]

Dennis Wagelaar <den...@vu...> writes:

> I don't see why poorer people have to be locked out (yes, there are
> people with old hardware who *don't* have 50 euros to spare).

If they can't affort 50EURs for a brand new card they should go with a
used one. A used Geforce2MX or RivaTNT2 will probally sell for 5-10EUR
or even less. Beside from that old hardware won't be able to run
Supertux in software anyway, 800x600 in software consumes a hect of a
lot CPU power, so going OpenGL only might actually LOWER the entrance
barrier, since you need much less CPU and just a very cheap gfx card.

> Also, it's a pity that handheld devices are ignored. It would be
> great to have a game that scales over such a wide variety of
> computing hardware as SuperTux 0.1.2 does.

SuperTux doesn't scale. Just because you can get something to compile,
doesn't mean that its of any use. The controlls can be extremly tricky
with most handhelds, RAM constrains may make actually running it
impossible and last not least a jump'n run at 10fps or less is simply
no fun. If you want SuperTux for handelds, you should write a SuperTux
specifically for handhelds, take the gfx scale them down to suit
320x240 or something like that, kick out all those FX and probally
replace some probally bloaty OOP code with some little hacks where
needed. Since you can reuse levels, code and stuff as needed it should
actually be not that difficult to start such a project. But it really
needs to be a seperate thing, if you try to satisfy both at the same
time, you will end up with something that doesn't look good on either
them.

> That gives you an edge over the mainstream games industry.

No, really not. The mainstream game industry knows what they are
doing, if they want game X on handheld Y, they port it, by
restructuring the game to suit whatever is available, turning a 3D
game into a 2D one and stuff like that is actually quite common
(SplinterCell, TonyHawks). If you just take a game designed for PC and
brute-force compile it for a completly different architecture you
won't get something with which players will be much happy.

> Then there's the gameplay issue: I thought only the physics system
> determines the gameplay? The rendering is just the visual feedback.

Um, well, graphic is of course visual feedback, but how to you expect
the player to react to something which he can't see (sprite rotation
not implement -> no sprite)? Go play some Yoshi's Island (1995, SNES)
to see what kind of effects we are talking about and yep, even that
game already used '3d acceleration' back than in form of the SFX2
chip. You will quickly find out that a lot of those effects are simply
not doable in software in a useable fasion.

> The only problem with the physics system is that the coordinate
> system is currently aligned to the physical amount of pixels used in
> the game (e.g. 800x600). Since coordinate scaling by a software
> rendering engine is not feasible (too many in-game calculations), it
> might be an idea to create a lookup table for all the physics
> constants used in the game.

Aehm, that is quite trivial. Scaling physics down to another CO is
trivial and extremly cheap compared to the amount of work that the CPU
needs to spend on GFX when OpenGL is not involved.

> I hope you will reconsider your plans and keep the architecture of
> SuperTux open.

Nope.

-- 
WWW:      http://pingus.seul.org/~grumbel/ 
JabberID: gr...@ja... 
ICQ:      59461927