So, one day later we're ON! The team got some trouble to meet on the first day and on the second only partial meetings were possible. 

But, here we are, moving forward. And moving forward is the idea of our (untitled) game. To prevent yourself from being caught by some YTBDVPM (yet to be defined vicious pursuing menace) you must run forward as fast as you can and while you're at it you might as well do your very best parkour moves.

So, I met with santagada and we sketched some of the basic moves for the game:



Later I went home, were leticia__n sketched a neat background for our city landscape:


 
 I think that's a very stylish city. If we keep it that way we'll have a unique style and that's totally awesome :)

Well, time to write some code, this game won't get made by itself. See ya!

I'm doing a city generator for our game, for now it is completely randomic, but the output looks so cool we decided to post about it :)
 

Well, the coding is starting to emerge as the commits in our top secret bitbucket repo piles up. I'm mainly working on the physics of the character and it's interaction with the environment. 


 
Everybody and their brother knows that jumping is an essential part of 2D platformers, so I'm trying hard to get this right. Jumping done right goes way back to the Super Mario series and it's strongly related to how pressing the button controls jump height. A lot of older games have fixed height jumps and they feel unnatural and hard to control. Even though I knew jumping could be done nice or painful, I've never really programmed a 2D platformer and when I tried I never got jumping right. It looks very simple, but I found it very tricky actually. But thanks to this awesome site dissecting the Sonic games physics – the classic ones, not the post 32 bit 3D crap –  jumping is looking good so far! Can't say the same about the game though:


Good ol' programmers art :)

Well, we had some two slow days on this pyweek, but now things are moving, and pretty fast! Thanks to cocos2d we have the scene management completely done. The physics are working well enough for now and we're starting to put our nice graphics. 

The level graphics are based on several drawings by leticia__n very similar to the one you see on the first diary entry.


 
For the character, we decided it would be easier to animate if we used a 3D character and rendered the later. After my very lame attempt of doing a nice 3D stick figure, I decided to look for a ready made rigged char and found this one (the first one. MTI Stickman) and it's perfect for our game! The rigging is totally beyond anything I could've done myself. It's so easy to animate!

 

And just some moments ago we managed to get the running animation inside the game. Here's a still frame as a teaser:

 
Well, I have no idea if the game will turn out good, but I know for sure it will be stylish :) Just wait for those buildings.

Well, it was a wild ride for sure. The game obviously didn't turn out exactly as expected, but the results are pretty nice. Not a lot of fun, unfortunately, but a learning experience for sure. I've never programmed a platformer before and it's more challenging than it seemed at first. A lot of early design mistakes made it much more difficult too. The code for the character physics is a huge mess.

The biggest mistakes I've made were all related to overcomplicating things. Instead of using simple rectangles I've programmed some custom collision classes. Also I've tried to model the physics as it was a complete engine only to find myself stuffing specific character behavior code all over the place later. The one thing that could've used a more complex base design is the main character state management. I've read recently some random forum guy talking about how you should take the state machine concept seriously in your code and I think he's right. It seems very natural to represent the character as a state machine and I should've done this way from the beginning.

Well, overall I'm very satisfied with the results. The graphics are looking awesome to me and the mood is very odd in a good way. Unlike our previously unsubmitted attempt on pyweek 9, this one I feel really motivated to develop further.

Well, the contest ended and the score was even better than I expected. The "bad" comments were along what I expected (it isn't really a game, isn't finished, etc), but what surprised really was the compliments. It seem that due the nature of the challenge, everyone is very willing to give a chance to see the good side of games. We've messed up on the packaging failing to notice that the inclusion of pyglet and cocos weren't working until the last moment, forgot to take the numpy dependence in account (due to the cocos particle systems that we included at the very end), and even then we've got 0% DNW.

People seemed forgiving of the downsides and pointed out how they liked the art style or how the wall jumping was nice (to some, at least :)). It was very motivating to see people commenting that they would play the complete game. So, thanks fellow pyweek participants for all the constructive criticism!

P.S.: It surprised me, though, that nobody complained about the most annoying jumping sound ever :D