I agree with what Laurent said about clojuredev, I've been having
great success using it myself, for my clojure/slick creations.  My
first project was converting Kevin Glass's platformer example over to
clojure which gave me a chance to learn slick and Clojure all in one
go!  Unfortunately, the code I wrote for that is getting to be quite
out of date now and probably doesn't work with the latest clojure, but
I can at least offer some input based on my experience.

The platformer example is based on a tile grid with some fancy phys2d
stuff thrown in for collision detection.  For grid storage I started
with a vector of vectors, and like you, which I too found far too
cumbersome.  So I ditched it and decided to go with a sorted map of
coordinates to tiles (from the suggestion made in your linked thread
actually).  Although I'm still not sure this is going to be a "best
practice" in clojure grids, it worked really well for me, simplifying
access to grid immensely.  In your code, moving over to a map would
probably reduce the number of functions involved in your matrix
generation phase.

Also, in order to reduce passing around refs to collections for all
the functions, I declared things like the player and the screen in the
game's namespace and then later defined them during the game's init
stage.  Still not completely ok with that solution, but I think it's
OK within the context of a slick game.

For me the import count reduced as the game grew and I separated it
into various files, so I wouldn't worry about that too much.
Personally, I've always been a fan of explicit imports so I know
what's going on when I reference a class name :)

-Matt



On Feb 9, 9:14 am, phtrivier <phtriv...@gmail.com> wrote:
> Hi everyone
>
>  first of all, this is my first post, so if I ask FAQs or deserve to
> RTFM, please tell me.
>
> I am learning clojure, and trying to adapt some game code written in a
> Java framework called Slick. After a couple of beginner's mistake (the
> kind you do to learn ;) ), I got a working convestion of a Scoller
> example. Still, there are some things that I don't find very elegant,
> or where I'm pretty sure to be reinventing some wheel.
>
> Code is here :
>
> http://tinyurl.com/dj4l9z
> or
> git://github.com/phtrivier/clj-slick-tank.git
>
> * The game is tile-based. I need to build up a collision map ("can the
> tank walk on cell [3,2] ?"). Java does it with a two-dimensional array
> of booleans. From discussion here 
> :http://groups.google.com/group/clojure/browse_thread/thread/5eb78c620...
> , I choosed to use a vector of vector (i probably should'nt have, but
> that's another point). I wrote this :
>
> (defn blocked?
>   "Is a position blocked in the screen?"
>   [screen x y]
>   (let [i (int x)
>         j (int y)]
>     (true? (get (get (screen :blocked) i) j))))
>
> (defn make-row-generator
>   [cell-generator w]
>   (fn [i]
>     (vec (map (fn [j] (cell-generator i j))
>               (range 0 w)))))
>
> (defn make-matrix
>   [w h cell-generator]
>   (vec (map (make-row-generator cell-generator w)
>             (range 0 h))))
>
> (defn make-collision-map
>   "Builds a double dimensioned array telling
> whether a cell is blocked"
>   [m w h]
>   (make-matrix
>    w h
>    (fn [i j]
>      (let [tileId (.getTileId m i j 0)]
>        (let [res
>              (Boolean/parseBoolean
>               (.getTileProperty m tileId "blocked" "false"))]
>          res
>          )))))
>
> (the 'm' is a slick-specific stucture that holds the tile map).
> Am I duplicating some existing library to build matrices ? Is there a
> cleaner way to do it ?
>
> * In a method I check for collisions :
>
>  ;; Movement
> (defn try-move
>   "Try and move a player in the screen given a direction.
>                 Returns a list with the moved player, and a
>                 boolean indicating whether the move was successfull.
>         "
>   [player screen dx dy]
>   (let [new_x (+ (player :x) dx)
>         new_y (+ (player :y) dy)]
>
>     (let [bxy (blocked? screen new_x new_y)
>           bx (blocked? screen new_x (player :y))
>           by (blocked? screen   (player :x) new_y)]
>       (if bxy
>         (if bx
>           (if by
>             [player, false]
>             [(assoc player :y new_y), true])
>           [(assoc player :x new_x), true])
>         [(assoc player :x new_x :y new_y), true]))))
>
> Will i get used to the nested 'if' blocks someday ? Any way to write
> the equivalent of a switch case here, if only to improve
> readibility ?
>
> * Finally, my code starts with a rather ugly
>
> (ns tank
>   (:import (org.newdawn.slick Animation
>                               AppGameContainer
>                               BasicGame
>                               GameContainer
>                               Graphics
>                               Input
>                               SlickException
>                               SpriteSheet)
>            (org.newdawn.slick.tiled TiledMap)
>            (org.newdawn.slick.util Log)))
>
> I know there is no way to import org.newdawn.slick.* (as discussed
> here :http://groups.google.com/group/clojure/browse_thread/thread/fa00a0ff4...
> ). What do you do in programs that need huge list of imports ? I'm
> kinda spoiled by the Eclipse way of doing this, which is roughly :
> import everything, forget about it, and let you IDE clear up the list.
> And at least for development part it make things easier. If I hadn't
> had the exact list of import to copy paste in this case, I would
> probably have gone bored and depressed by the second class to import
> manually (lazyness is not only for evaluation ;) )
>
> Thanks for any ideas, sorry if post is too long / newbie-like.
>
> Cheers
> PH
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Clojure" group.
To post to this group, send email to clojure@googlegroups.com
To unsubscribe from this group, send email to 
clojure+unsubscr...@googlegroups.com
For more options, visit this group at 
http://groups.google.com/group/clojure?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to