Scratch

I don't think anyone. I think it slowly developed bit by bit.

I think a large group of beginning Scratchers all just kind of reasoned that would work.

Basically it just was starting common knowledge, I mean logically it makes sense and doesn't need heightened reasoning.

I know Archmage is at least partially responsible for scrolling games, but I don't know about this. Personally, I reasoned it out on my own, so maybe everyone figured it out by themselves.

Hmm. I don't think it was Tanner-FBI. I'm pretty sure Cube World just got popular because it was a platformer (but not the only one). Since the top loved page counted all projects, no matter the age, it pretty much just stayed there.

archmage wrote:

Tanner-FBI didn't even make cube world, he took it from Jacool.

And Rhy's platformer base has some major flaws,
http://scratch.mit.edu/projects/archmage/101212

This one is older and works better than Rhy's.

Well we'd expect you to say that, but RHY really does have better boundary and slope detection, so...

archmage wrote:

I've worked out the code in Rhy's project.

Collisions work by moving the player character in the opposite direction it is moving by the same amount it moved to collide in to a wall.

For example

p=player w=wall
-P---wwww
Rhy's code will make the player move like this:
----wPwww
Then move out of the wall by an equal amount:
-P--wwww

If it was accurate, it would move the player beside the wall like so:
---Pwwww

Mine does this, Rhy just hates sensor sprites for whatever reason, this collision bug is also the reason why the player character is so floaty, it can't move too quickly in to the ground.

Rhy's code also is all bunched together for player player sprite which is a bad thing.

But yours does not do as you say.

If you collide to the wall as such in your example
-P---wwww
to
----wPwww

Then your sprite will not go to ---Pwwww but instead stay at ----wPwww.

I've tested this.

Thus you can say neither is perfect then.  RHY's does work a bit smoother though.

EDIT:  After looking at your code though it seems it works generally the same way as RHY's, moving away from the wall at .9 times the velocity (which I daresay is not exact).  Oddly though when I test it (I tested it eight times to be exact) the sensing sprite stays stuck in the wall.  What is with your script?

Last edited by Kileymeister (2011-04-22 16:26:14)

Yeah looking over at it it does work the same way with the wall collisions, but differently with others, I code it differently sometimes I think. The sensing sprite detects which side the player is colliding on which helps to make script much shorter.

But one thing about Rhy's is that you can't alter the variables much, especially the falling speed, it has to be floaty or it will cause bugs, mine doesn't do this.

The really professional ways of doing this can't be done in Scratch with its art based detection methods.

To prevent this interesting conversation from being closed, should I edit the topic's name and such?

Wait a second, this conversation is turning into an argument.

TuffGhost wrote:

archmage wrote:

TuffGhost wrote:

hi archmage i've gotten the sprite stuck in the wall like 5 times on that project so i would stop trying to say that you're better

the first person that i had ever heard about platformers from  (like the when touching color [] go up one) was actually chrisg, in 2006
so that pretty much settles that

Yeah, that definitely not biased since you are like his best friend or something. They both work pretty similarly so if the only bonus was that mine is shorter, simpler, and easier to modify I'd say that it makes mine better, even if his is slightly better with wall collisions.

i like how you immediately think that i'm doing this for rhy even when i didn't even say his name

rhys platformer base is actually incredibly easy to modify if you actually know how to, which i guess you don't

and the whole point of not using sensors is because they slow everything down

K, I am not arguing this anymore. Rhy's script is about 3 times longer, includes the animations and everything else in the one loop (creates editor lag/bad coding practice), and takes an average of 22% CPU (mine takes 4%). These are the reasons I prefer mine, but whatever suits your owns needs is fine.

Last edited by archmage (2011-04-22 17:10:08)

Kileymeister wrote:

I think a large group of beginning Scratchers all just kind of reasoned that would work.

agreed