Re: Bug Thread: Discussion
 

Yeah, I got that, but unfortunately it got overshadowed by other stuff and was left gathering dust. I'll take it up with JK and KO when I get home. Thanks for reminding me.

Re: Bug Thread: Discussion
 

I have an exploit, not particularly nasty, that I can't see listed anywhere (but I could have missed it). Version 3.20

Ritual of Rebirth: If you rebirth your prophet he retains his priest level boost. Prophetize him again and he gets another priest level. And so on, as many times as you want to get him killed and rebirthed. This is the kind of bug that's fun to exploit in SP in a big game against LA Ermor ;-)

Re: Bug Thread: Discussion
 

That's not really an exploit since it takes several turns to regain the ability to make a new prophet and then kill off your commader again. It also isn't free as it costs 15 gems every time. Third, in all that wasted time it is very likely that your commader will drop off the hall of fame list.

There are also much easier ways to make high level preists in combat. Communion being the best. Just prophetize any astral or blood mage.

Re: Bug Thread: Discussion
 

KO (the game designer) says that's thematic, not an exploit. Coming back from the dead obviously makes your prophet more holy. :) So, exploit away with no feelings of guilt.

BTW, Power of the Spheres (via Crystal Shield if desired) also boosts your H level, if your priest has magic paths as well as H. I use this to boost e.g. E2H2s to E3H3s for Divine Blessing.

-Max

The AI doesn't know how to attack!
 

SP game.

I've been hunting a Harvester of Sorrows. Numerous Mind Hunt's failed to kill him, I finally nailed him with a patrol.

He's got one piece of equipment, a Bow of War.

What's his first move? To move forward despite the fact that I have a ton of troops within range.

Turn 2 he engages in melee attacks.

Turn 3 he routs.

Turn 4 he dies.

Re: Bug Thread: Discussion
 

Not a bug. It would depend on what the AI assigned as that units orders. It sometimes gives orders that fail to take advantage of equipment its commanders have, such as in this case.

Re: Bug Thread: Discussion
 

That walks a mighty fine line between "WAD", and, "a bug". :o

Re: Bug Thread: Discussion
 

I'd say, bug of a different sort. :)
Another thing I currently got in SP - not sure whether it was in a short list: the Forge Lord consistently refusing to cast on himself Immolation (the spell that gives fire aura but burns non-immune caster as well) even though he had FR 100 or 150 in some cases. He prefered to cast Ironskin above existing Stoneskin (!) (there were lightning-casting enemies on the field) or, if Ironskin was scripted, cast Bladewind or Falling Fires (iirc) instead.

Re: Bug Thread: Discussion
 

Any chance he got the Bow of War from you earlier?
I see this most frequently in cases where the AI has already set the scripting, and later the unit picks up a piece of equipment.

Re: Bug Thread: Discussion
 

Immolation isn't an aura, it's a one-time spell that hurts people next to the caster. If there weren't any enemies right next to him right then, it wouldn't have helped.

Re: Bug Thread: Discussion
 

I've seen this behavior too, when a mage casts ironskin over scripted stoneskin. If I scripted stoneskin, although I have ironskin researched, it was for a reason. I didn't want to get the lightning vulnerability.

Re: Bug Thread: Discussion
 

^^I haven't seen that, it shouldn't do that, and it is completely believable that it did exactly that :p

Re: Bug Thread: Discussion
 

That would be a bug in my mind. You don't give a guy a ranged weapon and then give him melee orders!

Re: Bug Thread: Discussion
 

No. This was my first encounter with the guy. It's possible he fought some other AI earlier but I first became aware of him coming from the corner of the world.

Re: Bug Thread: Discussion
 

Actually, Stoneskin on top of Ironskin isn't a bad idea necessarily--Stoneskin is really low fatigue, and I've considered doing Summon Earthpower/Ironskin/Stoneskin with some thugs instead of Summon Earthpower/Invulnerability. It's understandable that you're aggravated when it happens unintentionally, though.

-Max

Re: Bug Thread: Discussion
 

Why is it a good idea though? More fatigue, more loss of resistances, and you don't go higher Prot than you would have from Ironskin in the first place.

Re: Bug Thread: Discussion
 

It works completely by the rules of the mechanics of the game as they are supposed to work. If scripted to attack, then it will attack. Therefore it is NOT a bug.

Whether the script is optimal considering equipment is irrelevant for those purposes, especially since it is unknown when the attack script was put in place. Other factors that could go into it are whether the AI calculates possible damage potential for the unit with missile vs melee and how it factors the Bow of War into that.

If untoward behavior is not present directly and there are that many unknowns, it is not a good idea to assume there is a bug.

Re: Bug Thread: Discussion
 

Well it really becomes an issue of semantics at a certain point, and I'm not sure that rigorous defense of a flawed position is necessarily the best way to go with it. o.O

I don't know how many people would consider it a bug that the AI would either script melee with a Bow of War, or that it would not rescript a commander who got new equipment. I would, so long as there were any mechanic for rescripting at all, though I fear there may not be, in which case it would technically be WAD, and a design flaw rather than a bug.

(<3 JK)

Re: Bug Thread: Discussion
 

If it is something a player is allowed to do, then it is valid for the AI to do it as well and is not a bug. Can a player equip a bow and script the unit to charge into melee? Yes. Is that a bug? No. Just remember that the AI is just suppose to simulate the actions of a player. Its not a bug if that player isn't especially bright, or decides to play the game in ways you yourself would not.

Re: Bug Thread: Discussion
 

Hmmm. I just tested this, and we're both wrong, Jim. You don't get 22 Prot (Stoneskin is supposed to give you +2 on top of whatever you had before), but you *also* don't get more loss of resistances. I tried both variations:

Ironskin, Stoneskin: Scripted stoneskin gets overridden completely. Thug cast Holy Avenger or Blade Wind (ack!).

Stoneskin, Ironskin: Both spells were cast. Prot was as if only Ironskin had been cast. Cold vulnerability (50) went away (!) when Ironskin was cast, and Shock vulnerability (75) replaced it.

-Max

Re: Bug Thread: Discussion
 

So you wouldn't consider it a bug for the AI to just do nothing? Sit in the starting castle and never recruit, research or try to expand. After all, a human player would be allowed to do that. Or maybe it could destroy it's own starting castle, lab & temple. Why not? A human player could.

The AI has enough trouble competing. It's worth pointing out especially egregious examples. Call it a design flaw rather than a bug, if you want. It's still an error of some kind.

Re: Bug Thread: Discussion
 

Interesting. Thanks for that.
In some of the most heated arguments Im not sure that the sequence of casting was considered. Some of the old "works for me / doesnt work for me" battles on the forum might be worth reconsidering.

Re: Bug Thread: Discussion
 

The only way to fix that is to alter the AI decision trees and related algorithms to take this case into account. The real question is whether doing so will introduce more complications that make it more trouble than its worth to fix than leaving it as it is currently. Many of these more obscure issues are that way, even ones that are real bugs, such as the one cleveland reported by PM.

And believe me when I say you people don't want to get hit with the Stick of Unintended Consequences on stuff like this. With one of the recent patches, a batch of problems that had been around for a long while were fixed, but they produced a side effect that was a much greater headache and had to be fixed separately before the patch was released. If that patch had gone out in its initial form, it would have stopped every MP game in progress in its tracks and ruined many of them.

So if it's something as trivial as this particular scripting issue, think VERY hard before turning a molehill to a real hill. You might just get a Precision 100 Gifts from Heaven as a complimentary bonus if it's tackled.

Re: Bug Thread: Discussion
 

My statement assumes that the AI is capable of making judgments. If it were a case that the AI wasn't capable of action, or if it wasn't allowed to weigh it's options, I would say that would be a bug. This example clearly shows that the AI at least doesn't set all of its commanders to "set battle orders". If the AI is able to make decisions and doesn't do what you would do, that is not a bug.

In fact, I'm not sure I would fire off my bow of war if it were me. It might be 13 arrow shots, but it would be 13 short bow arrows, which frankly isn't that exciting. Compare that to the an 18 strength armor piercing life drain attack on an ethereal unit who flies and has fear. You can very likely deal no damage with the bow, but the melee attack is more then likely to deal damage, net bonus HP, and your looming presence is guaranteed to deal moral damage to 15 squares around you.

Re: Bug Thread: Discussion
 

Freely granted that this issue is fairly trivial and quite probably not worth fixing, even without worrying about unintended consequences.

I was reacting far more to the statements that:
If it is something a player is allowed to do, then it is valid for the AI to do it as well and is not a bug.

and

It works completely by the rules of the mechanics of the game as they are supposed to work. If scripted to attack, then it will attack. Therefore it is NOT a bug.


Both of which seem to not allow for the possibility of bugs in the AI logic as long as that logic doesn't permit impossible orders.

In focusing on those responses, not the original situation, I was also ignoring the possibility that attacking was at least a reasonable option despite the results as AoE suggested. That makes it even less likely to be an issue.

Re: Bug Thread: Discussion
 

I see this entry in the short list:

I think this bug may be different:

In the World in Crisis MP game, TC had an immortal unit cast Arcane Nexus. That unit was killed in melee combat, in positive dominion. But TC's capitol was owned by another nation, so the immortal unit was permanently killed. However, the Arcane Nexus is still up.

Re: Bug Thread: Discussion
 

That looks like a variant of the same bug. Immortality has all kinds of triggers associated with it and it probably just checks that when the unit is killed on the battlefield that immortality is ok, global stays up, but doesn't check whether the capital is in friendly hands.

In both cases the common factor is that the caster of the global is eliminated from the game without being explicitly killed on the battlefield or the global dispelled via overwriting or dispel. It becomes a case of "The Void ate my caster, but my global is still up! Nyanyannyan-naa-naa!"

Re: Bug Thread: Discussion
 

Yep, my low astral pretender has entered that strange little house more than once to avoid assassins, death.

Nothing like lost in space and time to keep a global up.

Re: Bug Thread: Discussion
 

Hmm, I could have said that better. The units of a defeated nation that disappear sort of vanish and can't come back until their ID numbers are recycled, at which point a new unit would have the "This unit has cast a global enchantment" icon and the global stays up because that unit had never "died", even though the nation vanished. That's a lot different from "Lost in Time & Space", but the effect is similar in practice.

Dominion bug
 

Three times now, I have gotten EA-Ctis to Dominion 0 (verified by scores, graphs, etc) - and the god will not die.

He finally dies when I take his citadel. This despite the fact that I have 2 candles on his citadel at the time I take it.

Re: Bug Thread: Discussion
 

Attachment is a semirandomed SP game with CBM1.41. Plz see the battle occured at upper skelde. My S2 commander has reinvigoration 4 via blessing. Everything is ok in the first few turns, but when he start to move forward, he get 20-40 fatigue per round automaticly until it reachs 200. My enemy has no mage other than priests on the BF so I think it can't be result of any spell. Plz check it, thanks.

Re: Bug Thread: Discussion
 

heh that did look like a bug - but the opposing forces include blowpipes shooting paralyzing poison. That's what did him in.

Re: Bug Thread: Discussion
 

Thanks! So it would be a cheap and very effective way to counter solo thugs:)

Re: Bug Thread: Discussion
 

hmmm that seemed very unlucky though... they have damage 0 and he had armor 21 and luck... though I guess once he has fatigue, subsequent darts are more likely to critical and hence possible get through the armor. I'm gonna have QM look at it.

Re: Bug Thread: Discussion
 

I think the poison secondary effect with ranged weapons does not have to cause damage before taking effect. The same problem exists with the Androphag archers aswell.

In the past I have also had a glamoured unit in melee get poisoned even when no damage was done. So there seems to be some bugs with poison.

Re: Bug Thread: Discussion
 

Yeah, the poison effect is known and debated at length.
The blowguns paralyzing poison probably works the same way. That's brutal against thugs. Though shields might help? I can't remember. Does poison resistance work against the paralyzing poison? I'd expect it too, but without checking...

Re: Bug Thread: Discussion
 

if it were to be so, we should surely all have seen many an expansionist pretender in SP play go down to the woodland blowpipes. but this has not been my experience, so I remain puzzled.

CBM did just buff the blowpipes w/ +2 accuracy, but that can hardly be the reason.

Re: Bug Thread: Discussion
 

Because of the way that deviation is calculated in the game, a moderate Precision buff may not have a profound effect in most scenarios, but when there is a singular target, and thus only -1- single square that is worth hitting, every point of Precision increases accuracy a tremendous amount (up to a certain point of perfection).

Re: Bug Thread: Discussion
 

Precision just makes the missile land in the right square.

Blowpipes work the same way as poison arrows, as everyone said. Except there is a MR roll to resist the blowgun poison (Which is actually 50-damage stun damage, not paralyzing poison as the description says. The effect is immediate and does not divide stun across turns the way poison divides damage across turns).

From the debug:

blastsqr: unr17793 x43 y23 aoe0 dmg0 eff109 spc1075838977 as-1 al0^M
blastsqr: unr17793 x43 y23 aoe0 dmg50 eff46 spc1073754240 as-1 al0^M
affectvic vic15643 hv0^M
hitunit 17793 15643 dmg50 spec1073754240 ba4^M
spec_mr pen 12 mr 18 (unr 17793 vic 15643 dmg 46 eff 50)^M
resisted^M

That's a square with unit #15643 getting hit with a dart. It does no damage, and the unit passes the MR check. Later:

blastsqr: unr17798 x43 y23 aoe0 dmg0 eff109 spc1075838977 as-1 al0^M
blastsqr: unr17798 x43 y23 aoe0 dmg50 eff46 spc1073754240 as-1 al0^M
affectvic vic15643 hv0^M
hitunit 17798 15643 dmg50 spec1073754240 ba2^M
spec_mr pen 12 mr 18 (unr 17798 vic 15643 dmg 46 eff 50)^M

Oh, that's bad! No "resisted", so fatigue is increased, from 22 to 40. That stun damage (28) looks about right for a 50-dam attack on 22 protection.


Anyway, nice use of the Blowpipes! I've never managed to get them to do anything useful. :)

Re: Bug Thread: Discussion
 

According to the manual, there is only 11% chance to hit for 12 vs MR18. So I'm somewhat unlucky...

I just did some test. Air shield provides no protection vs the blowpipe, but poison resist 100% works.

Re: Bug Thread: Discussion
 

In the same game you can see an assassination attempt where the assassin manages to poison the mage in melee without even damaging him. So poison secondary effects seem to be activated even in melee if you hit but don't damage.

Strange that air shield does not protect against blowpipes. That technically should stop the hit occurring depending on how much percentage air shield it is.

Re: Bug Thread: Discussion
 

Oh, sorry ... there were many more MR checks your guy passed in the debug data, I just didn't copy them all. :o

It was something around 1 in 10 failure though. But at 50 dam & 25 protection, you only need to fail 4 times to go unconscious.

Re: Bug Thread: Discussion
 

I tested this a while ago and units with poison secondaryeffect weapons couldn't poison a load of ulmish heavy infantry without shields. So I'm not sure what's going on here.

Oh and yeah poison immunity definitely works vs blowpipe poison. It has dt stun and dt poison. Dt poison is just there for immunity checks, it doesn't actually cause the slow poisoning effect seen if you use the special poison weapons.

Re: Bug Thread: Discussion
 

Here's what happened in the assassination:

The assassin missed with his Short Sword.

Then he hit the Adon with his Poison Dagger, doing 11 points of damage, but the Body Ethereal check succeeded, so no physical damage was taken. However, the secondary effect does trigger in this case, so poison damage was done.

Body Ethereal does not protect you from the secondary effects of weapons that hit you, even if it saves you from the initial physical damage. Only if the melee weapon does no damage because it's stopped by armour (or just misses) then the secondary effect does not trigger (like Sombre says).

Debug output:

19573 striking with weapon Short Sword. att21 def25
19573 striking with weapon Poison Dagger. att28 def22
hitloc Assassin strikes Adon wl0 diff3 -> 4
hitunit 19573 15643 dmg2 spec2097153 ba4
damage 11 on Adon, spec0x200001 ba4
Ethereal negated 11 pnts of dmg
hitloc Assassin strikes Adon wl0 diff3 -> 4
hitunit 19573 15643 dmg15 spec8320 ba4

Re: Bug Thread: Discussion
 

A new minor bug I just met in ComfortZone:

Normally you can not raise a new prophet from a castle under siege, but the AI can.

I guess this is so obvious that someone should have reported it, yet found no clue in the pinned thread.

Re: Bug Thread: Discussion
 

The ai can also raise troops and mages in provinces it would be unable to.

Re: Bug Thread: Discussion
 

I wouldn't call those bugs, but AI advantages. Same as the AI can receive more gem, gold, etc. income.

Re: Bug Thread: Discussion
 

Rudra has D3, but 0 undead leadership. So when he raises undead they just stand there and dissolve. Normally that's not a problem as you will probably trust your rudra into melee. But once in a while you want him to stay in the back line and cast spells. And since raise skeletons/undead obviously has a high priority he does it a lot unfortunately.

Re: Bug Thread: Discussion
 

This game is *so* freaking buggy.

I move from B-->A, fighting and successfully taking A. In the process about 50 troops start to route.

The combat ends with about 40 of the 50 off map. The only square to retreat to is B.


Even worse, the units that were routing, and yet were still on the map are ALSO gone.

In direct contradiction of the manual, p 81:

"units in a victorious army.....If the battle ends before the routing units exit the map, at they end of the battle they remain with the victorious army."

Re: Bug Thread: Discussion
 

my reply from the other thread you posted this in:

"etreats are resolved after combats and, as such, the units had no legitimate province left to retreat to. That is not a bug, but how the game works.

If there were units in your army which are not in the province afterwards, two possibilities exist: 1) that they were under the effect of poison or decay or somesuch, and so died, or 2) that the battle result you saw was not the actual battle, due to some incompatibility of OS or versions.

Routed units from a magical phase battle will appear in the retreated to province in time for the normal battle there, I believe."