Vanilla 3.4.0 Death Mold Bug - Spoiler

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • Bombadil
    Rookie
    • Feb 2011
    • 13

    Vanilla 3.4.0 Death Mold Bug - Spoiler

    I fought a death mold last night, and the monster memory reports that it is simultaneously hurt by fire and resists fire.

    It seems that it resists fire is correct, as fire bolts did nothing.
  • PowerWyrm
    Prophet
    • Apr 2008
    • 2986

    #2
    IIRC that's already fixed for next version.
    PWMAngband variant maintainer - check https://github.com/draconisPW/PWMAngband (or http://www.mangband.org/forum/viewforum.php?f=9) to learn more about this new variant!

    Comment

    • DaviddesJ
      Swordsman
      • Mar 2008
      • 254

      #3
      The bug still seems to be present in 3.4.1.

      Comment

      • Timo Pietilä
        Prophet
        • Apr 2007
        • 4096

        #4
        Originally posted by DaviddesJ
        The bug still seems to be present in 3.4.1.
        Death mold entry in monster.txt lacks -F:HURT_FIRE. monster_base.txt has F:HURT_FIRE for molds.

        I bet that's not the only one of that kind of bugs, if you edit those by hand it takes time to find all the exceptions to the rule. Anybody good with scripting to find out exceptions?

        [edit] grammar so messy I had hard time figuring out what I was saying.
        Last edited by Timo Pietilä; June 7, 2013, 06:45.

        Comment

        • DaviddesJ
          Swordsman
          • Mar 2008
          • 254

          #5
          Another possibility, rather than looking for conflicts in the monster descriptions, would be to just make the code suppress this inconsistency (i.e., if a monster has the flag for hurt by fire, and also the flag for resists fire, then suppress the former when describing the monster to the player).

          If I'm understanding the situation correctly. Basically, make the text description of the monster always match what the code actually does for that monster, even if the configuration files seem inconsistent.

          Comment

          • Derakon
            Prophet
            • Dec 2009
            • 9022

            #6
            But when there's a conflict, how do you decide which one overrides the other?

            Comment

            • Daniel Fishman
              Adept
              • Apr 2007
              • 131

              #7
              Originally posted by Derakon
              But when there's a conflict, how do you decide which one overrides the other?
              The one that actually happens, as the code is written currently - which in this case is RES_FIRE, according to the OP.

              Comment

              • DaviddesJ
                Swordsman
                • Mar 2008
                • 254

                #8
                Originally posted by Derakon
                But when there's a conflict, how do you decide which one overrides the other?
                The code already has to resolve that conflict one way or the other. The only issue is to make the text description match what the code actually does. That could be built into the system that outputs the descriptions, without changing anything else.

                Ideally, the specification of the algorithm that determines the effect would itself be parseable in a way that it could be used both to determine the in-game effects and to generate the text description, automatically. But that's probably too ambitious and a more reasonable goal would be to simply fix conflicts when they are discovered by changing the text generator.

                Comment

                Working...
                😀
                😂
                🥰
                😘
                🤢
                😎
                😞
                😡
                👍
                👎