Aleron
Wanderer
Joined: 01 Aug 2005
Posts: 76
|
 Posted: Sat Dec 20, 2008 12:34 am
[2.37] Odd (ie, buggy) behavior with #STEP and paths
|
I'm experiencing some buggy behavior with #STEP and paths in CMUD. As of recently, when I'm using a particular set of triggers for pathing around an area in the mud, #STEP will sometimes send the entire portion of the current path that remains, rather than only the next single direction. The following quote is output from one such occurrence.
| Quote: |
<772hp 690e 121359425exp 123mv [] NS>dir memory: south|east|east
flee dirs:
debug - before #STEP
2s2wdsds
debug - after #STEP
debug - exiting engine alias
<Type 'help' for help>
<772hp 690e 121359425exp 123mv [] NS>Slow walking aborted |
The code below is the alias that is executed in each room which controls everything, including the execution of #STEP.
| Code: |
#LOCAL $lastDir
#SAY debug - entering engine alias
#WAIT 500
#SAY dir memory: @dirMemory
#SAY flee dirs: @fleeDirs
#IF (@hasFled = "false") {
#IF (%numitems(@mobsInRoom) > 0) {
k %item(@mobsInRoom, 1)
doFocus
#VAR inBattle "true"
} {
#IF (@hitPoints <= @botHPThreshold) {
// Fake a flee. This will cause memory
// to backtrack and then sleep, where
// I should be safe.
#VAR hasFled "true"
#VAR inBattle "false"
engine
} {
#IF (@room != "coin") {
// If there are movements still left in the
// fleeDirs variable then do those first,
// THEN continue STEP'ing on the path
#IF (%numitems(@fleeDirs) > 0) {
$lastDir = %pop(@fleeDirs)
#SWITCH ($lastDir = "north") {n}
($lastDir = "south") {s}
($lastDir = "east") {e}
($lastDir = "west") {w}
($lastDir = "up") {u}
($lastDir = "down") {d}
} {
#SAY debug - before #STEP
#STEP
#SAY debug - after #STEP
}
}
}
}
} {
#IF (%numitems(@dirMemory) > 0) {
$lastDir = %pop(@dirMemory)
#VAR fleeDirs %push($lastDir, @fleeDirs)
#SWITCH ($lastDir = "north") {s}
($lastDir = "south") {n}
($lastDir = "east") {w}
($lastDir = "west") {e}
($lastDir = "up") {d}
($lastDir = "down") {u}
} {
#IF (@hitPoints <= @botHPThreshold) {
gt ~[BOT~] HP is too low. Resting...
#VAR botIsSleeping "true"
eatf
drinkw
#WAIT 2000
sleeph
} {
#VAR hasFled "false"
engine
}
}
}
#SAY debug - exiting engine alias |
For a good week or two, this set of triggers was causing me no problem. Whether it's some corruption of the package file or something, I have no idea. I did force the alias to recompile as a sanity check, and I still see this behavior. I cannot reproduce it in a controlled way, it just happens. Sometimes after a few minutes of the triggers running, sometimes after hours, sometimes not at all.
Anything I can send in that would help diagnosing this bug?
Edit:
When #STEP fails like this, if I execute it (#STEP) again, it goes back to working normally. |
|
