Qhimm.com Forums

Miscellaneous Forums => Troubleshooting => Solved Problems => Topic started by: titeguy3 on 2009-10-02 22:44:22

Title: Messed up Pagoda battles by editing AI (solved)
Post by: titeguy3 on 2009-10-02 22:44:22
So I'd mapped out new and improved attacks and AI for Gorkii, Shake, Chekhov, and Staniv. I decided to try and implement them today. I changed the attacks in entries 155 and 156 in PrC accordingly, modified the animation/formation data, to include those attacks, and rewrote the AI from scratch.

When I ran the game to test this new AI, and spoke to Gorkii on disc 1, rather than fighting gorkii, I ended up fighting two Garudas in an "Attacked from both sides" battle vs Yuffie at the Da-Chao Rapps Battle area. I thought that was wierd, but beat them and continued to fight Shake, which caused the game to crash.

This is bizarre, I've modified enemy AI countless times and nothing like this has happened.

I will point out that this is the first major modification I've made since using Proud Clod 1.3.0, and a couple of bizarre things that I've noticed...

1) Since upgrading to 1.3.0, whenever I create Scene.bin, I get an alert that the Scene look up in Kernel.bin "appears to be incorrect" and prompts me to attempt to fix it, despite the fact that I've never touched the Scene look-up data and it matches up with a fresh install kernel.bin's Scene look-up. The error happens even if I never click "yes".
2) What does the 16-bit hex data next to an attack's animation index in the "Animations/Formations" window represent in PrC? It's usually "FFFF", but for an attack every now and again, it'll have some value. Gorkii's third attack had the value "0006"...
3) Garuda's appear in entries 153 and 154, whereas Gorkii's is entry 155. Is it possible that I'm getting some kind of bizarre off-by-one error?

EDIT: I can confirm that random battles are still normal...
Title: Re: Messed up Pagoda battles by editing AI (possible PrC 1.3.0 bugs?)
Post by: secondadvent on 2009-10-02 23:13:18
well prc does not update the kernel, even if you click yes (it creates a test kernel), and massively changing the AI will cause the scene lookup to be off, so to correct it, open it up in hojo and update your kernel that way to make sure the lookups are correct.
Title: Re: Messed up Pagoda battles by editing AI (possible PrC 1.3.0 bugs?)
Post by: titeguy3 on 2009-10-02 23:25:24
Hrm... First time it's happened to me, but I'll give it a try.

EDIT: Yeah, that was the problem, thanks!
Title: Re: Messed up Pagoda battles by editing AI (solved)
Post by: secondadvent on 2009-10-02 23:41:25
yeah, if you modify too much of anything it has a chance of doing that... it is because of the compression, so if you make too many things different (or the same for that matter) the size is bound to change at some point :-P.
Title: Re: Messed up Pagoda battles by editing AI (solved)
Post by: Evan Lee on 2009-10-03 03:40:45
can you also modify attack animation sequences like changing battle stance?
Title: Re: Messed up Pagoda battles by editing AI (solved)
Post by: titeguy3 on 2009-10-03 05:28:03
I've tried that through the AI before... it didn't work.
Title: Re: Messed up Pagoda battles by editing AI (possible PrC 1.3.0 bugs?)
Post by: nfitc1 on 2009-10-04 23:04:00
well prc does not update the kernel, even if you click yes (it creates a test kernel)

GAKK! I really have to change that. :(