Unfixed bugs of vanilla game, Send them here if you find any Options

[eripmav]

Don't know if anybody already listed this...

Some beholders(Gauth and the beholder worthing 9000XP, maybe others) can fire their eyestalks even when unconsious. Interestingly, they dont do this when feared or confused...

[Sikret]

Don't know if anybody already listed this...

Some beholders(Gauth and the beholder worthing 9000XP, maybe others) can fire their eyestalks even when unconsious. Interestingly, they dont do this when feared or confused...

Yes, it's a vanilla game bug which has been fixed in IA v6.

[Silencer150]

I saw this mentioned in post #186, but there was no follow up.

Are there any plans to include a Shapeshifter fix in an upcoming version of IA? The readme warns not to install Weimer's version of Shapeshifter Rebalancing because it's overpowered. However the fact still remains that the original Shapeshifter class had some problems. The statistics of the werewolf and greater werewolf forms were nerfed from what you were supposed to get, and the werewolf paws could be dispelled (which also happens if you save & reload). Weimer's version fixed the stats and dispel bug, but introduced what some players felt was an inappropriate way of shapeshifting, and one that came with its own share of exploits.

The Refinements mod (whose HLA tables would likely create incompatibilities with IA) had another idea. It fixed the statistics and made it where the paws, if dispelled, would return at the end of the round. However, for some reason, it also made it where shapeshifting stunned your character, which could very easily get you killed.

On SHS one user created a new variation which got rid of the stun, fixed the animations, and made the paws undispellable. The problem was that you were immune to Dispel Magic. Plus it's still beta, and I don't know if the guy has touched it since. (I can't even figure out how to test this fix, mind you.)


I'd like to see IA address this issue in its own right. It looks like a lot of work and consideration was already put into other classes, especially rangers.

Here are my suggestions. In addition to fixing the stats and the dispel bug, scale the resistances of the werewolf forms so that it's not one huge leap at level 13. Perhaps the STR/DEX/CON bonuses and elemental resistances gradually improve with level, and the werewolf form doesn't grant immunity to normal weapons until level 10. Also, I know it's reminiscent of Weimer's overpowered version, but I'd still suggest an option for spellcasting in the form of a HLA. (Spellcasting shapeshifters weren't in PnP, but then neither were HLA's.) I don't think it would overpower the class if it's a HLA; perhaps it would be the last in the line of a tiered set of abilities. Priest spells do tend to have long casting times regardless (Bless, Chant, Call Lightning, Iron Skins, etc.) and I've found that while my shapeshifter is standing there casting a spell, she's certainly not able to beat up enemies.

I know this is a lot to ask, but I figured it wouldn't hurt to mention it.

[antizyclon]

SPOILER!

An exploit still work in IA v5 with item that have uses per day puting them on bag of holding they get one use can (practicaly unlimited use per day) this removed or is engine hard coded.

[Sikret]

Yes, that is a known bug of the vanilla game. First of all, note that abusing this bug is plain cheat and no serious player should even think of it.

We did consider a possible method to fix it, but it required a huge amount of work and it wouldn't lead to a perfect solution in the end; so, we decided to leave it at least for now.

[darkjeshush]

This isnt a bug, but more of an lame exploit you might want to fix for v6. Idunno if anybody would go through the trouble to do it, but here it is anyways...

You can get free xp in shadow temple by sending out skeleton warriors to various spots of the map covered by fog of war. Then go watch tv or something and the skeleton warriors will farm those little shades respawning infinitely for 8 hours; The skeletons are completely immune to the shades' damage. It would net about a million xp per hour while you're away from your computer.

[Sikret]

This isnt a bug, but more of an lame exploit you might want to fix for v6. Idunno if anybody would go through the trouble to do it, but here it is anyways...

You can get free xp in shadow temple by sending out skeleton warriors to various spots of the map covered by fog of war. Then go watch tv or something and the skeleton warriors will farm those little shades respawning infinitely for 8 hours; The skeletons are completely immune to the shades' damage. It would net about a million xp per hour while you're away from your computer.

All such and similar exploits have already been fixed in v6. See the Progress Report for IA v6. Quote:

- Monsters who spawn to interrupt the party's rest no longer carry any treasure nor do they have any xp value. The same tweak is applied to monsters who can respawn infinitely during the game.

It took us ages to detect, fix, and test the big number of such possible exploits in the game, but we did it.

[Vuki]

I am pretty sure that everybody noticed it already but it is still worth to mention it. Description of Greater Restoration and the effect of it is really different. The spell description says that it affects the target but in reality it affects the whole party. This discrepancy should be resolved imho.

Another real bug in the game is that even if you buy restoration in a temple your character still will be fatigued. The spell causes this effect to the caster and not to the target of the spell (unless it is the same person of course) and in the current case the caster is a priest of the temple.

This post has been edited by Vuki: Jul 29 2009, 12:54 PM

[Vuki]

In the 2nd level of the Maze under Spellhold there is a small bug. You can enter to the room of the djinni (where you need the painting) without opening the door. If you click on the right bottom of the cell then the character can walk to the cell (it is in the following area: from x=1800, y=530 to x=1890, y= 470).

[Kerkes]

I don't know if this a bug however since it is changed for Skullcrusher mace: Equalizer bonus vs Aligments applies to main hand wpn also (if Equalizer is used off hand)

[bulian]

A few discoveries I've made, my apologies if they've been mentioned already:

-Cromwell's upgrade description of Lilarcor/warblade says it gives +2 to strength/dexterity/constitution, which doesn't match the readme or result

-Neutralize poison (4th level cleric spell) cast from a scroll does not remove Neo Otugyh (sp?) disease though it says it will in the description. haven't checked the actual spell.

-script error in Teshal's fight - I tried it in Ch2/3 with my current group, and while I managed to kill the skeleton lords and gated bone golems, everyone in my group eventually succumbed to the barrage of greater commands and were lying unconcious with a group of skeleton warriors (summoned) and teshal standing around not attacking instead of chopping me to pieces.

[Sikret]

-Cromwell's upgrade description of Lilarcor/warblade says it gives +2 to strength/dexterity/constitution, which doesn't match the readme or result

-Neutralize poison (4th level cleric spell) cast from a scroll does not remove Neo Otugyh (sp?) disease though it says it will in the description. haven't checked the actual spell.

I don't have access to v5 files at the moment to check these two problems, but neither of them exist in v6. Perhaps they had been reported and fixed before.

-script error in Teshal's fight - I tried it in Ch2/3 with my current group, and while I managed to kill the skeleton lords and gated bone golems, everyone in my group eventually succumbed to the barrage of greater commands and were lying unconcious with a group of skeleton warriors (summoned) and teshal standing around not attacking instead of chopping me to pieces.

This is an engine limitation (doesn't have anything to do with IA). Sometimes enemies just don't see unconscious targets, sometimes they do.

PS: All three cases were vanilla game issues (not related to IA); so the post and my answer are both moved to the vanilla game bug reports thread.

This post has been edited by Sikret: Nov 22 2009, 02:10 AM

[DavidW]

-script error in Teshal's fight - I tried it in Ch2/3 with my current group, and while I managed to kill the skeleton lords and gated bone golems, everyone in my group eventually succumbed to the barrage of greater commands and were lying unconcious with a group of skeleton warriors (summoned) and teshal standing around not attacking instead of chopping me to pieces.

This is an engine limitation (doesn't have anything to do with IA). Sometimes enemies just don't see unconscious targets, sometimes they do.

I think (dredging up dim memories of having the same problem) it's more a script language problem than an engine problem. You can get around it by telling monsters that if they can't see anything hostile but they can see Player1, attack Player1. (And repeat for Player2-Player6). IF See(Player1) THEN RESPONSE #100 Attack(Player1) END works (iirc) irrespective of Player1's state.

[Sikret]

In my IA v5 game, I was observing this issue. Even with the See(NearestEnemyOf(Myself)), enemies sometimes managed to see unconscious targets and sometimes failed to see them. That's why I thought that it might be an engine problem. Actually, the frequency of cases they managed to see the unconscious targets was more than cases in which they failed to see them. Perhaps, there is a more subtle explanation for the phenomenon. All in all, I prefer to let it pass, as it's not a highly probable or frequent event to have ALL party members lying unconscious in the battlefield.

[lroumen]

I would actually prioritize unconscious characters because they have no defenses... or is that just me.

[Sikret]

I would actually prioritize unconscious characters because they have no defenses... or is that just me.

This is one way of looking at it. The alternate view is that the unconscious character is not an active threat and is not a priority. For example, I will certainly try to attack an active mage who is casting spells on me rather than the mage's friend who is napping on the ground.

Admittedly, it depends on the situation. For example, attacking a 20th level unconscious target has higher priority than attacking a 1st level active target. There are a lot of factors involved and this is an example of where a human mind is superior to artificial intelligence (A.I.).

[DavidW]

In my IA v5 game, I was observing this issue. Even with the See(NearestEnemyOf(Myself)), enemies sometimes managed to see unconscious targets and sometimes failed to see them. That's why I thought that it might be an engine problem. Actually, the frequency of cases they managed to see the unconscious targets was more than cases in which they failed to see them. Perhaps, there is a more subtle explanation for the phenomenon.

To be honest, I'm not sure. I think I did at one point find necessary & sufficient conditions for See(NearestEnemyOf(Myself)) to fire, but I may be misremembering. (Once I found See(PlayerX) always worked, the problem became fairly academic.)

All in all, I prefer to let it pass, as it's not a highly probable or frequent event to have ALL party members lying unconscious in the battlefield.

Fair enough. It happens a bit more often in BG1 situations, where spells like Sleep are effective against low-level parties - I think that's where I discovered the problem, in fact.

[Sikret]

In my IA v5 game, I was observing this issue. Even with the See(NearestEnemyOf(Myself)), enemies sometimes managed to see unconscious targets and sometimes failed to see them. That's why I thought that it might be an engine problem. Actually, the frequency of cases they managed to see the unconscious targets was more than cases in which they failed to see them. Perhaps, there is a more subtle explanation for the phenomenon.

To be honest, I'm not sure. I think I did at one point find necessary & sufficient conditions for See(NearestEnemyOf(Myself)) to fire, but I may be misremembering. (Once I found See(PlayerX) always worked, the problem became fairly academic.)

I did some more comprehensive tests on this issue and now I'm quite convinced that there is some engine problem involved in this case and the See(NearestEnemyOf(Myself)) trigger is not to be blamed. When the entire party are unconscious, enemy behavior is quite random; sometimes (actually most of the times) they see unconscious party members and sometimes they don't see them.

In general, the enemies seem to have a better chance to see the protagonist (than other party members) while unconscious (this may be due to the fact that they seek player1 when failing to see anyone). So, it seems that adding See(Playerx) triggers doesn't look necessary or worth the trouble, because player1 is mostly seen and attacked anyway.

I tried to figure out some general rules, but each and every theory was falsified by counter examples in further tests. Whether (or not) the enemy sees an unconscious target looks to be random. Fortunately, this is not a big deal, because as I said before, the cases in which the entire party are lying unconscious on the ground are quite rare and improbable (actually, I had to try hard everytime to prepare that situation for my tests).

[Hoppy]

In my IA v5 game, I was observing this issue. Even with the See(NearestEnemyOf(Myself)), enemies sometimes managed to see unconscious targets and sometimes failed to see them. That's why I thought that it might be an engine problem. Actually, the frequency of cases they managed to see the unconscious targets was more than cases in which they failed to see them. Perhaps, there is a more subtle explanation for the phenomenon.

To be honest, I'm not sure. I think I did at one point find necessary & sufficient conditions for See(NearestEnemyOf(Myself)) to fire, but I may be misremembering. (Once I found See(PlayerX) always worked, the problem became fairly academic.)

I did some more comprehensive tests on this issue and now I'm quite convinced that there is some engine problem involved in this case and the See(NearestEnemyOf(Myself)) trigger is not to be blamed. When the entire party are unconscious, enemy behavior is quite random; sometimes (actually most of the times) they see unconscious party members and sometimes they don't see them.

In general, the enemies seem to have a better chance to see the protagonist (than other party members) while unconscious (this may be due to the fact that they seek player1 when failing to see anyone). So, it seems that adding See(Playerx) triggers doesn't look necessary or worth the trouble, because player1 is mostly seen and attacked anyway.

I tried to figure out some general rules, but each and every theory was falsified by counter examples in further tests. Whether (or not) the enemy sees an unconscious target looks to be random. Fortunately, this is not a big deal, because as I said before, the cases in which the entire party are lying unconscious on the ground are quite rare and improbable (actually, I had to try hard everytime to prepare that situation for my tests).

Hi Sikret!

I think more of engine limitations as hardcoded but I think we are limited in the IDS files to an extent.

StateCheck(Player1,STATE_HELPLESS) which is the state the Emotion spell causes and this is just a simple check. This would mean adding more and more script blocks to your enemies' BCS to account for the multiple disabling states and then the enemies attack reevaluation. Then to make sure they do checks for all six party member. Also the reverse would have to be considered as well as an opposite block !StateCheck(....).

In my opinion, this approach does give an added sense of smarter enemy AI (or in party AI scripting) but it can create very bloated scripts that can cause some instability if things don't work right. On the other hand it may be something you choose for more specific enemy types or characters and not every single combatant. For example, an enemy golem or enemy troll may not be as choosy as an enemy fighter or spellcaster.

Some mods I have seen append to STATE.IDS and make a more cumulative state check of STATE_DISABLED so that cuts down on the multiple state checks for stunned, sleeping, confused, etc. Still it is not limiting in my opinion, just a lot more scripting language to use as DavidW pointed out.

Then again if the whole party is disabled then it is game over

Good luck with the finishing touches on IA6

[Sikret]

Checking states is already comprehensively used in scripts; they do work properly, but it is not relevant to the discussed issue above. The question was to find necessary and sufficient conditions for See(NearestEnemyOf(Myself)) trigger to turn true and this is what seems to be working quite randomly with no general rules. I did many tests and it seems that when the entire party are unconscious, whether (or not) the enemy can see unconscious party members is random.