[35.6] Monster Summoning V crash

[gatperdut]

When cast from either a scroll or from memory, this spell sometimes produces a crash in my install. I think it might depend on what is summoned. 

For example, in one of the casts that worked it summoned 2 Ogre berserkers.

Another thing is that these gentlemen are 6th level each, despite the spell description stating it'll be 2d3 7th-8th level monsters. Maybe there are no suitable monsters of those levels and the engine makes do with what it has at its disposal, but I thought I'd point it out in case it's relevant.

I don't think any other mod in my Weidu.log besides SCS modifies the spells, hence my posting here.

Also, personal rumination besides the point and very likely out of the concerns of SCS: I suppose their berserker rage immunities might be relevant depending on the situation, but it's rather underwhelming for spell slots unlocked around the time one adventures in the Underdark. 

WeiDU (1).log

[jmerry]

All right. What, exactly, is the list of creatures that can be summoned by this spell? That list should be a 2DA file, probably MONSUM05.2DA. Which would be in the override, and it's an easily read plain text format.

The ogre berserkers (resref = ogregrsu) aren't the problem. Likely, one of the other creatures on the list has a missing animation or something, and that's what's causing the crash.

On the nature of the spell ... my feeling is that the niche "Monster Summoning N" spells fill is numerous fodder. If you want individually strong summons, there are other summoning spells. But then ... ogre berserkers are on the list for the vanilla Monster Summoning III (a level 5 spell). Still having them on the list at two spell levels higher feels off, even with the boost SCS gives them by making them actual berserkers.

[gatperdut]

Thanks, that's a good pointer. It's odd that the filename is different though, I couldn't find MONSUM05.2DA.

These are the contents of my MSUMMO5.2da

2DA V1.0
0
     RESREF	Hit 	AreaHitAnimation
1 OGREMASU msumm1h msumm1x
2 OGREGRSU msumm1h msumm1x
3 DW#MS4YT msumm1h msumm1x

I would assume the issue lies in DW#MS4YT? I did a bit of searching in the override folder, these are the files I found:

OGREMASU: .cre, .bcs

OGREGRSU: .cre

msumm1h + msumm1x: .vvc, .bam

DW#MS4YT: nothing, not even searching for MS4YT in case the hash symbol trips the search

 

Assuming DW#MS4YT is somewhat more powerful, I personally find it acceptable and possibly hilarious to roll the dice and accept the chance that something unimpressive may show up when going for such spells. I suppose OGREMASU is an Ogre MAge (does the GR stand for "grrr" in OGREGRSU?), which could be some sort of middle ground.

[jmerry]

Yeah, a creature that's outright missing would break it. From the resource name, I think that's meant to be a yuan-ti of some sort.

OGREMASU is an ogre mage, as you thought. Like OGREGRSU, it's present in vanilla.

I mentioned MONSUM05.2DA because that's what's already there in BG2EE. But the higher monster summonings are taking cues from IWD, which has slightly different conventions. MSUMMO5.2DA is clearly the table being used here (there would be a reference to it in the summoning effect, if you wanted to be absolutely sure).

So, trying to track things down... we start with the copied IWD resources, which provides this table:

2DA V1.0
0
     RESREF HitAnimation AreaHitAnimation
1    MS5gspi	MSumm1H	MSumm1X
2    MS5jzom	MSumm1H	MSumm1X
3    MS5mino	MSumm1H	MSumm1X

Now, those creatures could simply be copied, but the BG series is a different region with different ecology. We'll pick up a completely different set of monsters.

Ignoring a cosmetic change (replace those "MSumm1H" effects with "spmonsum") which didn't seem to happen anyway, the main action is in summoned_monsters_arcane.tpa. That goes through all the lists, cloning new monsters as needed. So what does that have to say about MS5?

First, make a troll. Basic troll01 cloned to dw#ms5tr. Then tweak the copied minotaur (MS5mino) creature a bit. And finally, run a function "spl_make_summoning_2da". Two creatures on it, the troll and the minotaur. File name msummo5.

Wait. That's completely different from what you have. And if I look down a little farther, your table seems to mostly match the one from MS4. That does an ogre mage (ogremasu, refurbishing the unused creature in the base game) and a yuan-ti (dw#ms4yu, cloned from the basic icyuan01). No ogre berserkers; those are only in MS3.

So, there's at least three things wrong with the table for Monster Summoning 5. First, that yuan-ti somehow got its resource name typoed to dw#ms4yt instead of ds#ms4yu. Second, your table merges parts of two different lists. And finally, none of the monsters involved are supposed to be on the MS5 list; they're all from lower spells.

And looking at the code, I have no idea how any of those three things happened.

[gatperdut]

Okay, you lost me a bit there, but probably DavidW understands it. With some other bug he mentioned running "weidu --change-log <file>" so here's the output for that, in case it helps:
 

weidu --change-log MSUMMO5.2da
[weidu] WeiDU version 24900
[.\chitin.key] 191 BIFFs, 62570 resources
[.\lang\en_us\dialog.tlk] 107336 string entries
[stratagems/lang/english/setup.tra] has 262 translation strings



Mods affecting MSUMMO5.2DA:
00000: /* created or unbiffed */ ~STRATAGEMS/SETUP-STRATAGEMS.TP2~ 0 1500 // Include arcane spells from Icewind Dale: Enhanced Edition35.5
00001:  ~STRATAGEMS/SETUP-STRATAGEMS.TP2~ 0 2000 // Install all spell tweaks (if you don't select this, you will be given a chance to choose by category)35.5

 

[jmerry]

Short summary of the state of things: The initial list for Monster Summoning 5, on first installing the IWD spells, is a troll and a minotaur. The ininitial list for Monster Summoning 4 is an ogre mage and a yuan-ti. The initial list for Monster Summoning 3 includes an ogre mage. Your list for Monster Summoning 5 is the ogre mage from the MS4 list, a typoed version of the yuan-ti from the MS4 list, and the ogre mage from the MS3 list. While all of these are wrong, it's the typoed yuan-ti that causes the crashes you've been seeing.

So, that changelog. Nothing outside SCS. I've looked at the IWD spell component, and didn't see anything wrong there. Now, the spell tweaks... that would be spell/adjust_monster_summoning.tpa. And I've spotted it.

That file starts at high levels and goes down. For MS5, it edits the minotaur slightly, creates a greater yuan-ti (dw#ms5yt), and rebuilds the table with the minotaur, troll, and greater yuan-ti. Then for MS4, it edits the yuan-ti (dw#ms4yu), edits the ogre berserker (ogregrsu), and rebuilds the table with the ogre mage, ogre berserker, and yuan-ti. Only there are two typos in this line - it rebuilds the msummo5 table instead of msummo4, and it puts in the nonexistent dw#ms4yt instead of the correct dw#ms4yu.

The problem (line 145):

LAF spl_make_summoning_2da STR_VAR monsters="OGREMASU OGREGRSU DW#MS4YT" filename=msummo5 END

The fixed version:

LAF spl_make_summoning_2da STR_VAR monsters="OGREMASU OGREGRSU DW#MS4YU" filename=msummo4 END

For a hotfix without reinstalling everything, edit the MS4 and MS5 tables as follows:

msummo4:

2DA V1.0
0
     RESREF	Hit 	AreaHitAnimation
1 OGREMASU msumm1h msumm1x
2 OGREGRSU msumm1h msumm1x
3 DW#MS4YU msumm1h msumm1x

msummo5:

2DA V1.0
0
     RESREF	Hit 	AreaHitAnimation
1 MS5MINO  msumm1h msumm1x
2 DW#MS5TR msumm1h msumm1x
3 DW#MS5YT msumm1h msumm1x

 

Edited January 26 by jmerry

[DavidW]

Thanks! Good detective work.

[gatperdut]

Awesome. Going to do those edits later. Thank you!