Title is not especially descriptive.
Issue discussed in the autoBasement thread.
Symptom is that maximizer will suggest a maximization involving the sword and then fail to equip what it just suggested, when asked to equip.
autoBasement is the "canary" since it does a lot of maximization.
The situation as I understand it is that:
Something, currently undiagnosed, happens that makes KoLmafia believe that there are more than one s'n's swords in inventory. Then a maximizer command is run, the maximizer believes there are two swords available and tries to equip both. There is only one sword so the equip command fails giving the observed symptoms.
In some cases unequipping the sword (if so equipped) and refreshing inventory makes the symptom go away. In some cases logging out of KoLmafia and restarting makes it go away.
There is a hypothesis that having the sword as part of an outfit is a contributing factor to the over counting.
Tangentially, the stickers themselves have a number in parens after the type that is supposed to be a count of turns remaining before the sticker fails. This number has been observed to go negative. Correctly tracking this is probably a separate FR, but the opportunity to address this may occur as part of tracking this down.
Issue discussed in the autoBasement thread.
Symptom is that maximizer will suggest a maximization involving the sword and then fail to equip what it just suggested, when asked to equip.
autoBasement is the "canary" since it does a lot of maximization.
The situation as I understand it is that:
Something, currently undiagnosed, happens that makes KoLmafia believe that there are more than one s'n's swords in inventory. Then a maximizer command is run, the maximizer believes there are two swords available and tries to equip both. There is only one sword so the equip command fails giving the observed symptoms.
In some cases unequipping the sword (if so equipped) and refreshing inventory makes the symptom go away. In some cases logging out of KoLmafia and restarting makes it go away.
There is a hypothesis that having the sword as part of an outfit is a contributing factor to the over counting.
Tangentially, the stickers themselves have a number in parens after the type that is supposed to be a count of turns remaining before the sticker fails. This number has been observed to go negative. Correctly tracking this is probably a separate FR, but the opportunity to address this may occur as part of tracking this down.