Managing different movement styles

ID:266037
 
Jul 30 2011, 5:17 pm
I was trying to access one movement style from Move() and ended up with a recursion error. [due to the step function in that movement].

    Move(newLoc, moveDir)
        switch(state)
            if(STRUGGLE)
                if(dir == src.dir && src.struggle)
                    src.struggle.push(src)
        ..()


Now if I were to try "spawn() src.struggle.push(src)". I would get "runtime error: Cannot execute null.push()."

Below is part of a sword struggle datum. In struggle mode player movement changes to push/pull movement. It used to work when I had a macro called "tapdir(dir as num)". But i removed this macro due to a recent movement system change.

Struggle
    var
        mob/player
            one = null
            two = null

    proc
        Break(var/mob/player/winner,var/mob/player/loser,var/distance) //had to use Break() since break is reserved
            step(winner, turn(winner.dir,180) )
            winner.endStruggle()
            loser.endStruggle()
            loser.knock_back(winner,0+distance,5)


        push(var/mob/player/m)
            if(m == one)
                var/turf/t = get_step(get_step(m, m.dir), m.dir)
                if(t)
                    if(!t.density)
                        for(var/mob/player/o in t)
                            Break(m, two)
                            o.knock_back(m, 1, 5)
                            return

                        for(var/obj/dense_object/d in t)
                            Break(m,two)
                            return
                            //Object should take damage!

                        flick("push_struggle",one)
                        flick("pushed_struggle",two)
                        step(two, m.dir)
                        step(m, m.dir)
                    else
                        Break(m,two)

                else
                    world << output("BREAK","output2")
                    Break(m,two)

            else
                var/turf/t = get_step(get_step(m,m.dir),m.dir)
                if(t)
                    if(!t.density)
                        for(var/mob/player/o in t)
                            Break(m, one)
                            o.knock_back(m, 1, 5)
                            return
                        for(var/obj/dense_object/d in t)
                            Break(m, one)
                            return
                            //Object should take damage!

                        flick("pushed_struggle",one)
                        flick("push_struggle",two)
                        step(one, m.dir)
                        step(m, m.dir)
                    else
                        Break(m, one)
                else
                    Break(m, one)
#1 Aug 3 2011, 6:38 pm (Edited on Aug 3 2011, 7:04 pm)
<Incorrect info. removed so as not to inject fallacies.>

I don't see how using spawn will remove a problem with infinite recursion. That's usually an issue with logic, which spawning wouldn't prevent.

Anyways, I don't see why this is in Design Philosophy and not somewhere else, like Code Problems.

And it would help if you explained what you were trying to accomplish. Like, instead of posting code, just explain what the push proc does.
I don't think people want to examine your code to figure out what it does. It's much faster if you just tell us.
I don't even know what you mean by "sword struggle." And what is moving (Move proc) even supposed to mean in that case?
#2 Aug 3 2011, 6:42 pm
In response to Complex Robot (#1)
Of course 'src' is defined in spawned sections of code, anything that worked before the spawn() will work after it if the references still exist. The runtime error he was getting was caused by one of the variables belonging to src was null. So calling a proc for that reference variable would spit out a runtime error.
#3 Aug 3 2011, 7:02 pm
In response to Nadrew (#2)
Nadrew wrote:
Of course 'src' is defined in spawned sections of code, anything that worked before the spawn() will work after it if the references still exist. The runtime error he was getting was caused by one of the variables belonging to src was null. So calling a proc for that reference variable would spit out a runtime error.

Showing how dumb I am about spawn, again.
Thank you for correcting me.
I guess I'll re-read his post to see if I can get what he's saying.