Hey there! Welcome! Always happy to see more Axe programmers! I'm not as proficient as Zeda or E37 but I'm also around! If you want any tips I am happy to help as well.

A few more months, a bit more progress. Once again I'd like to say thanks for all the help and suggestions. With that being said, I've run into another issue and this time I'm not even sure what the cause is.

So here is the problem:
I have successfully written an algorithm to generate 4-6 non-intersecting rooms and I'm attempting to connect them. I wrote some code that is supposed to accomplish this task, but it seems to only work once. As in, any attempts to execute this task consecutively in any manner promptly results in a RAM clear. Here is the code in question that I am using to connect rooms:
Data(31,31,3,31,31,3,1,1)->°RegionCheck
Lbl ConnectRooms
For(I,0,7)
If {°RegionArray+I}
{°RegionCheck+I}^10->AdjacentRoom
If {°RegionArray+I+AdjacentRoom}
GetAdjRoom()
°RoomArray+(I*4)->Room
ConnectTo(AdjacentRoom)
End
If {°RegionCheck+I}/10->AdjacentRoom
GetAdjRoom()
°RoomArray+(I*4)->Room
ConnectTo(AdjacentRoom)
End
End
End
Return
Lbl GetAdjRoom
°RoomArray+(I+AdjacentRoom*4)->Room
{°RoomX+Room}->AdjacentRoomX
{°RoomY+Room}->AdjacentRoomY
{°RoomWidth+Room}->AdjRoomWidth
{°RoomHeight+Room}->AdjRoomHeight
Return
Lbl ConnectTo
.../Args
ConnectTo(AdjacentRoom)
...
If [r1]=°Below
{°RoomX+Room}+1+(Rand()^({°RoomWidth+Room}-2))->PathFromX
AdjacentRoomX+1+(Rand()^(AdjRoomWidth-2))->PathToX
If PathFromX=PathToX
Connect(PathFromX,{°RoomY+Room}+{°RoomHeight+Room},PathToX,AdjacentRoomY-1)
Else
If AdjacentRoomY-1-{°RoomY+Room}+{°RoomHeight+Room}>4
{°RoomY+Room}+{°RoomHeight+Room}+1+(Rand()^(AdjacentRoomY-({°RoomY+Room}+{°RoomHeight+Room}-1)))->CorridorY
Connect(min(PathFromX,PathToX),CorridorY,max(PathToX,PathFromX),CorridorY)
Connect(PathFromX,{°RoomY+Room}+{°RoomHeight+Room},PathFromX,CorridorY)
Connect(PathToX,CorridorY,PathToX,AdjacentRoomY-1)
Else
Rand()^1->P
If P
Connect(PathFromX,{°RoomY+Room}+{°RoomHeight+Room},PathFromX,AdjacentRoomY-1)
Else
Connect(PathToX,{°RoomY+Room}+{°RoomHeight+Room},PathToX,AdjacentRoomY-1)
End
End
End
End
If [r1]=°Right
{°RoomY+Room}+1+(Rand()^({°RoomHeight+Room}-2))->PathFromY
AdjacentRoomY+1+(Rand()^(AdjRoomHeight-2))->PathToY
If PathFromY=PathToY
Connect({°RoomX+Room}+{°RoomWidth+Room},PathFromY,AdjacentRoomX-1,PathToY)
Else
If AdjacentRoomX-1-{°RoomX+Room}+{°RoomWidth+Room}>4
{°RoomX+Room}+{°RoomWidth+Room}+1+(Rand()^(AdjacentRoomX-({°RoomX+Room}+{°RoomWidth+Room}-1)))->CorridorX
Connect(CorridorX,min(PathFromY,PathToY),CorridorX,max(PathToY,PathFromY))
Connect({°RoomX+Room}+{°RoomWidth+Room},PathFromY,CorridorX,PathFromY)
Connect(CorridorX,PathToY,AdjacentRoomX-1,PathToY)
Else
Rand()^1->P
If P
Connect({°RoomX+Room}+{°RoomWidth+Room},PathFromY,AdjacentRoomX-1,PathFromY)
Else
Connect({°RoomX+Room}+{°RoomWidth+Room},PathToY,AdjacentRoomX-1,PathToY)
End
End
End
End
Return
Lbl Connect
.../Args
Connect(X1,Y1,X2,Y2)
...
If [r1]=[r3]
For(J,0,[r4]-[r2])
0->{[r2]*°FloorWidth+[r1]+Floor+(J*48)}
End
End
If [r2]=[r4]
Fill([r2]*°FloorWidth+[r1]+Floor,[r3]-[r1]+1,0)
End
ReturnIn addition, any number greater than 2 in place of "I" will also cause a RAM clear.
Xeda was helping me over on the Cemetech discord, but frankly we are both lost on what the issue could be.
Overview of the current problem: inputting numbers greater than 2 as the offset of the current room causes a RAM clear and executing the code to connect rooms consecutively causes a RAM clear regardless of room "validity". As an example, if I were to change the For loop in ConnectRooms() from For(I,0,7) to For(I,0,1) there is no issue. This is because on the seed I'm using to test, there is no 0th room. So it is skipped and instead connects the 1st room. However it crashes once it hits 2 as the doesn't like being executed consecutively. If I were to generate additional floors of the same seed, the first room that is valid can have its connections made, but no further. I've done a lot of different tests but now I need help. Is there something that I overlooked or am not taking into account? I've attached a screenshot to further illustrate.

Thanks for the optimization, however even with that change there wasn't any change at all. Sorry if its hard to follow, that's the unformatted tokenized text. Perhaps this would be a bit more readable:
.G
L4+22->->°RoomStart
L4+24->->°RoomWidth
L4+26->->°RoomHeight
L4+28->->°RoomX
L4+30->->°RoomY
L4+32->->°NewRoomX
L4+34->->°NewRoomY
L4+36->->°Invalid
L4+38->->°UR
L4+40->->°LL
L4+42->->°NewRoomX2
L4+44->->°NewRoomY2
L4+20->->°Rooms
Fill(L1,36,0)
0->Rooms->RoomX->RoomY->RoomHeight->RoomWidth
ClrDraw
GenFloor()
CheckFloor()
Main()
Lbl Main
While 1
DispGraph
Input()
EndIf getKey(15)
Return
Lbl CheckFloor
For(J,1,Rooms)
J*4+L1-4->Z
J*5-4->Y
DrawText(1,Y,"(")
DrawInt(4,Y,{Z})
DrawText(13,Y,",")
DrawInt(16,Y,{Z+1})
DrawText(24,Y,")")
DrawInt(29,Y,{Z+2})
DrawInt(38,Y,{Z+3})
End
Return
Lbl GenFloor
While Rooms<9
GenRoom()
PlaceRoom()
End
Return
Lbl GenRoom
GetRoomStart()
While 1
GetRoomSide()->RoomWidth
End!If ChkRoomWidth()
While 1
GetRoomSide()->RoomHeight
End!If ChkRoomHeight()
Return
Lbl PlaceRoom
If Rooms<9
RoomY+RoomHeight-1->{RoomX+RoomWidth-1->{RoomY->{RoomX->{Rooms+1->Rooms*4+L1-4}+1}+1}+1}
End
Return
Lbl GetRoomStart
6+(rand^(33))->RoomX
4+(rand^(40))->RoomY
Return
Lbl GetRoomSide
4+(rand^(9))
Return
Lbl ChkRoomWidth
(RoomX+RoomWidth-1>41)
Return
Lbl ChkRoomHeight
(RoomY+RoomHeight-1>43)
Return
Lbl DrawText
For(I,0,length([r3])-1)
Copy({[r3]+I}-32*4+°CharSprites,L4+50,4)
Fill(L4+54,4,0)
DrawChar()
End
Return
Lbl DrawInt
[r3]/10+16->{L5}
[r3]^10+16->{L5+1}
For(I,0,1)
Copy({L5+I}*4+°CharSprites,L4+50,4)
Fill(L4+54,4,0)
DrawChar()
End
Return
Lbl DrawChar
Pt-On([r1]+(I*((I>0)*4)),[r2],L4+50)
Return
Lbl Input
If getKey(9)
ClrDraw
0->Rooms
GenFloor()
End
Return
[]->°CharSprites
[0000000080800080]
[A0A00000A0E0E0A0]
[60C060C0A060C0A0]
[40A060E040400000]
[4080804040202040]
[A040A0000040E040]
[000040800000E000]
[0000008000204080]
[E0A0A0E0C04040E0]
[E020C0E0E06020E0]
[A0A0E020E08060E0]
[E080E0E0E0204040]
[E0A0E0E0E0A0E020]
[4000400040004080]
[0020402000E000E0]
[00804080E0200040]
[40A0E04060A0E0A0]
[C0E0A0E0E08080E0]
[C0A0A0E0E0C080E0]
[E0C08080E080A0C0]
[A0E0A0A0E04040E0]
[6020A040A0C0A0A0]
[808080E0E0E0A0A0]
[C0A0A0A0C0A0A060]
[C0A0E080E0A0E040]
[C0A0C0A0E08020E0]
[E0404040A0A0A060]
[A0A0A040A0A0E0E0]
[A040A0A0A0A04040]
[E02080E000000000]Still not sure what's causing the bug let alone where to begin. I've changed the loop from a while loop to a repeat loop and there still was no change. What makes this all the more frustrating is that this exact structure and code works in another program. Somehow this program alone is having issues and I'm not sure why that would even be the case. Maybe Axe is just bugged? I don't feel like that's the case since many people have created vastly more complicated projects than this and I have yet to see anything about loops simply not working/crashing.

@E37 Thank you! Your advice was very helpful in better understanding how Axe works and I was able to get it to work perfectly! I think you were correct in saying that CheckRight was returning L and not the comparison like I assumed it would. Using Return!If inside the for loop is what allowed it to work. This is the result:
If getKey(3) and (I=0) and (J=0) and ({Z+1+GDB0}=1)
    If CheckRight()
      1→I
      32→S
    End
  End
  If getKey(2) and (I=0) and (J=0) and ({Z-1+GDB0}=1)
    If CheckLeft()
      0-1→I
      64→S
    End
  End
  If getKey(1) and (I=0) and (J=0) and ({Z+37+GDB0}=1)
    If CheckDown()
      1→J
      0→S
    End
  End
  If getKey(4) and (I=0) and (J=0) and ({Z-37+GDB0}=1)
    If CheckUp()
      0-1→J
      96→S
    End
  End
Lbl CheckRight
  For(K,1,L)
    {K*4+L₁-2}ʳ→N
    Return!If Z+1≠N
  End
Return
Lbl CheckLeft
  For(K,1,L)
    {K*4+L₁-2}ʳ→N
    Return!If Z-1≠N
  End
Return
Lbl CheckDown
  For(K,1,L)
    {K*4+L₁-2}ʳ→N
    Return!If Z+37≠N
  End
Return
Lbl CheckUp
  For(K,1,L)
    {K*4+L₁-2}ʳ→N
    Return!If Z-37≠N
  End
ReturnI appreciate the help.

Hey Xeda, thanks for replying. Initially I suspected that was the case, so I wrote in the code to display both the values of Z and M, which I commented out later to test other ideas. However the case of Z being ever less than zero should never happen as Z starts at 192 as that corresponds to the player's position at the top leftmost corner of the map, effectively (0,0). Then I thought maybe the method of accessing and checking M was somehow incorrect, so I wrote another test program that should be similar to the conditions of my original program:
.ARRAY
ClrHome
0→L
193→A
AddEntry(2,26)
AddEntry(1,0)
AddEntry(15,9)
AddEntry(2,2)
Repeat getKey(15)
  CheckArray()
End
ClrHome
Lbl AddEntry
  r₂*37+[r1]+192→{[r2]->{[r1]->{L+1->L*4+L1-4}+1}+1}ʳ
Return
Lbl CheckArray
  For(I,1,L)
    {I*4+L₁-2}ʳ→B
    If (B-192≤A) and (B+192≥A)
      Output(0,0,"IN BOUNDS")
      Pause 1800
      ClrHome
      Else
      Output(0,0,"OUT OF BOUNDS")
      Pause 1800
      ClrHome
    End
  End
ReturnThis code also executes without a hitch and perfectly loops through each value and displays whether it is out of bounds or in bounds based on a given value. I'm not entirely sure what I am overlooking at this point. When I was doing some more debugging, it looked like checking the value in the for loop was not playing nicely with drawing to the screen. I'm attaching a screenshot so I can hopefully better explain what is going on.
Funny thing: In the midst of writing this reply, I went to test the code again, this time I have text that shows whether or not a given mob is drawn, and strangely, it works.
AddMob(1,0)
AddMob(2,2)
AddMob(15,9)
AddMob(2,26)
AddMob(7,5)
AddMob(8,5)
Lbl DrawMob
  For(E,1,L)
    E*4+L₁-4→G
    {E*4+L₁-2}ʳ→M
    Text(0,0,M►Dec)
  Text(36,0,L►Dec)
  If (M-192≤Z) and (M+192≥Z)
  Text(0,E*6,"Drawn")
    {G}-P*8→r₁
    {G+1}-Q*8→r₂
    Pt-Off(44+r₁-H,28+r₂-V,Pic2+(F/3^4*8))
  Else
  Text(0,E*6,"Not Drawn")
  End
  End
ReturnThe AddMob code hasn't changed but I added a few more to the map so I can better gauge what's being drawn and should be/what's being drawn and shouldn't be/what's not being drawn and should be, if that makes sense. I added in what mobs are being placed on the map for reference in the screenshot. So the mobs at (1,0),(2,2) and (7,5) are within the bounds of being drawn and therefore should be. I went ahead counted in the map data where a mob should not be drawn based on the character's starting position and lo and behold, the mob at (8,5) should not be drawn and therefore isn't. All the other mobs are also not drawn which is the behavior I desired. Now what doesn't make sense is that this code literally was not working the day before. All I did was rewrite it by hand in a new file which makes this even more confusing. Actually the only difference between the two is this line:
If ((M-155≤Z) and (M+155≥Z))became this line:
If (M-155≤Z) and (M+155≥Z)In an earlier version of this project instead of "M" I used {G+2}ʳ which I used in my test program to verify that these were in fact the same values and could be used in comparing values, and it still wasn't working in the original program. So, to be honest, I'm quite lost. I'm going to continue onward with the idea that I can use the same for loop in the collision check for each direction and go from there. This has been a weird 24 hours.

5349: You realize if this thread goes another page you will get an overflow error.

Quote from: Ki1o on May 21, 2013, 08:07:04 pm

As a generalization, Christians believe that they are the only religion and that all other religions are completely false.  They feel they will be the ones to live while everyone else suffers in H E double hockey sticks.  Zero tolerance...
At Muslims have some sort of respect for Christianity because of the same sort of belief in one supreme being.

It is a real generalization. And I don't know where you find your statistics, but it is completely false. Catholics really respect other religions. And when you speak about "They feel they will be the ones to live while everyone else suffers in H E double hockey sticks", I think you're confounding christians with some sects that use bible to have some credibility.

First off, forgive me if I seem offensive in any way.  I hope that doesn't make you hate me or anything.  With that being said, I think, and maybe it may be due to the fact that this is what I'm exposed and have witnessed as an African-American, Catholics are excluded from my perception of Christians.  My opinion of a Christian is your streotypical white American Protestant who is completely intolerant to all other religions including Catholics and especially Muslims, and is somewhat racist.  That's a generalization and I don't mean that all Christians are like that.  This opinion is also partially due to the influence of my history teacher who is a great teacher and his ideas are interesting and he portrays America in an interesting way.

Oh really? But that still makes it only three though... there are even fewer who are interested in what I want to study in college (Game development).  I wish there were more though.