Alvana reported that MoP changed SetAlphaGradient, requires -1 instead of 0 to keep gradient. And math.huge crashes the current beta build.

diff --git a/AutoTurnIn.lua b/AutoTurnIn.lua
index f4c7098..6e52302 100644
--- a/AutoTurnIn.lua
+++ b/AutoTurnIn.lua
@@ -216,7 +216,7 @@ end

 function AutoTurnIn:QUEST_DETAIL()
 	if self:AllowedToHandle() and (AutoTurnInCharacterDB.all or L.quests[GetTitleText()]) then
-		QuestInfoDescriptionText:SetAlphaGradient(0, math.huge)
+		QuestInfoDescriptionText:SetAlphaGradient(0, -1)
 		QuestInfoDescriptionText:SetAlpha(1)
 		AcceptQuest()
 	end