One of my heroes got killed recently by a single life drain. He had 16 hearts ( full life ) and 14 crosses, actually he even had 22 crosses because he was accompanied by a super prayermaster hero. How on earth is that possible, according to the wiki site it is not possible to kill a hero with that stats. No things like black prayer or mindstorm were in effect, I'm playin 1,40n insecticide.
Besides he was wearing a regeneration ring and was not coming back after I won the fight, but I consider this part rather a feature than a bug.
Sounds like a bug, but there's no telling what the bug was. Life drain damage does prevent regenerating units from coming back to life after the battle, which might be intended (I've seen it happen with trolls and ghouls, it's something about create undead).
Yes, Life Drain does Create Undead type damage so not regenerating is normal.
With that much resistance, the Life Drain needed at least a -28 resistance modifier to do that, not possible normally.
I don't think this can happen unless there is a bug, or the cost or save modifier of the spell was edited.
Was there an Evil Omens in effect perhaps?
Yeah, for that reason i was thinking of a feature rather than a bug. On the other hand, if I remember correctly, trolls are still able to regenerate after being "black channeled".
I hope I attached the correct save, it happens while attacking purples fortress with draconian and witch hero, maybe its helpful.
Thanks all for ya comments!
(January 12th, 2016, 15:50)Seravy Wrote: Yes, Life Drain does Create Undead type damage so not regenerating is normal.
Is it actually documented anywhere that that is intended behavior? It certainly seems plausible, but I don't recall it ever being mentioned.
The spell says it creates undead from the killed unit, and it does so.
I'll check the save as soon as I can (found a possible severe bug somewhere right now)
(January 12th, 2016, 17:34)Mortu Wrote: Yeah, for that reason i was thinking of a feature rather than a bug. On the other hand, if I remember correctly, trolls are still able to regenerate after being "black channeled".
They are. For that matter, if the side with the create undead damage wins, the trolls will rise from the dead. I interpret that as "they still regenerate after battle, unfortunately they regenerate as zombies you have to kill".
(January 12th, 2016, 17:34)Mortu Wrote: Yeah, for that reason i was thinking of a feature rather than a bug. On the other hand, if I remember correctly, trolls are still able to regenerate after being "black channeled".
I hope I attached the correct save, it happens while attacking purples fortress with draconian and witch hero, maybe its helpful.
Thanks all for ya comments!
Confirmed. The AI pumped roughly 1000 mana into the spell. I assume this is because Kyrub fixed the "30k mana" bug for AI fortress casting...but not the slider part (which btw has a bunch of other bugs related to modifiers, too). In fact, the AI casted Counter Magic for over 2000 mana the first turn of combat, I had to manually overwrite its skill to even let it cast another spell (as it had -2300 skill remaining)
I personally tried to fix this bug myself for the human player side (not the AI side) a few months ago and it made the slider malfunction, either allowing tens of thousands of mana, or spending negative mana.
The most direct solution I know for this is to disable the 0.5 casting multiplier at fortresses completely, which would fix the bug on the player's side too but some people are saying they rather stay unable to cast over 16k than to lose the modifier, so I'm not sure what to do. I'll think about it.
Edit : The root of the problem here is that most procedures dealing with casting costs look at the effective cost of things, in other words, how much mana worth of spells you can cast from your MP.
At a multiplier of 0.5, this means your 30k mana allows casting 60k worth of spells. However, the max value in a word is 32767, any higher will count as a negative number. As such, the game will think there is a negative amount of mana available, and not allow casting.
However, if that barrier is gone by fixing only that part of the bug, then we have effectively, -5535 mana available to use on the slider. Since that is lower than the amount we could use based on our skill, it will be used, meaning a full slider would represent -5535 mana. However, keep in mind that -5535 is also +60000, and the slider is linear. If we only push it halfway, we are only spending +30000 mana into the spell and...there we have the bug. Problem is, to fix this, every single part of code everywhere dealing with casting costs has to be changed (sliders, strategy combat, normal combat, autoing, player and AI casting, the checks for available mana, the actual amount spent, the amount to display on screen for the slider and so on). This is way too much work for a feature I already removed from CoM and have no use of fixing for my primary project.
Edit2 : I have a proposal how to fix this in an easy way that compromises in achieving every goal. What if the 0.5 discount was speficically only active for players having below 16k mana in their reserves? I could most likely add such a check into the function that calculated the modifier. Then you'd still have to pay full price if you have too much mana, but that shouldn't be relevant (you have 16k mana afterall to spend so spending an extra 100 per combat is no big deal) and would mean there are no bugs anywhere.
That sounds like a great compromise! Good work Seravy! This is all great stuff, getting things tested and fixed once and for all. I'm so glad to see it coming together!
