I've just quickly tried that - the mod has no apparent effect on the behaviour I described previously.
Dreaming the Impossible Dream: An Adventurer's Quest for a Functional Version of Spell Shield in SCS v 30
#26
Posté 14 juillet 2015 - 05:38
- Alesia_BH aime ceci
#27
Posté 14 juillet 2015 - 05:53
Thanks Grond0! You're awesome!
Your tests helped a lot.
Best,
A.
#28
Posté 14 juillet 2015 - 07:38
Btw. As an aside, I've done some further testing with my old SCS (with vanilla Spell Shield) + Fixpack install to get a more accurate estimate of the failure rate of vanilla Spell Shield when used alongside multiple spells protections against hostile NPCs. I'm up to 12 and have yet to observe a failure.
Illustrations.
Kangaxx casts Ruby Ray during Time Stop. Spell Shield comes down.
Arcadia recasts Spell Shield via Trigger.
Kangaxx uses Wish-Breach
Spell Shield comes down.
That's the behavior I'm accustomed to in cases where enemies other than beholders launch attacks on vanilla Spell Shields cast by the PC: clean with infrequent exceptions. Under those circumstance, the data suggest that my 1/8 estimate of the failure rate may have been high, although I'd need a larger sample size to determine that. I can't comment on the failure rate under other circumstances, such as when the PC launches an attack on a hostile NPC's Spell Shield.
As I think we've established now, the failure rate with the new Spell Shield related system in SCSv30 Fixpack v10 is higher when used in conjuction with other spell protections, at least in the case of use by the player. That's why I'm tempted to revert to an older version of SCS if we can't get your fix working.
I hope you'll get it Aasim! Thanks again for all your efforts!
Best,
A.
#29
Posté 14 juillet 2015 - 08:27
Oookay.
Alesia, check this for me.
Open up spwi321.spl file with NI.
It should have two 146 opcodes (Cast Spell).
One of these is spwi321b. This one should have a power level of 4, should have some opcode that removes spell protections max level 5.
This one isn't important for us now, but must be there.
The second one spwi321c. It should contain a single opcode (Remove protections by type, type "unknown, followed by a number - probably 14) This one takes down Spell Shield. After you install my mod, it should have a power level of 0. Does this happen?
#30
Posté 14 juillet 2015 - 08:37
Alesia, check this for me.
Open up spwi321.spl file with NI.
It should have two 146 opcodes (Cast Spell).
One of these is spwi321b. This one should have a power level of 4, should have some opcode that removes spell protections max level 5.
This one isn't important for us now, but must be there.
The second one spwi321c. It should contain a single opcode (Remove protections by type, type "unknown, followed by a number - probably 14) This one takes down Spell Shield. After you install my mod, it should have a power level of 0. Does this happen?
Will do.
I'm glad you're still on the case, Aasim.
Best,
A.
#31
Posté 14 juillet 2015 - 08:43
I am able to replicate your issues now. Checking now what goes wrong
- Alesia_BH aime ceci
#32
Posté 14 juillet 2015 - 08:48
SPWI321b and SPWI321c are both there. SPWI321c has a power level of 0.
Best,
A.
#33
Posté 14 juillet 2015 - 08:56
As another aside, I've done some further experimenting with the SCS v30 Fixpack unpatched system behavior.
The script is substantially better in v30, which makes the Spell Shield problems less of an issue in practice.
The v30 version is only easier in the subset of cases where the AI uses a Spell Thrust based attack strategy.
My initial trials had, by chance, involved Spell Thrust based strategies, thus v30 seemed trivial in comparison to my prior install, but sampling the full space of AI attack strategies in v30, the new system isn't that bad.
V30 is useable if we can't patch the GoI/Spell Shield/Spell Thrust and Spell Protection/Spell Shield/Breach issues, I'd say.
It would still be great to have a patch.
Best,
A.
#34
Posté 14 juillet 2015 - 09:06
Ok, this is crazy. If I run my mod on vanilla BG2, it doesn't work. Reason unknown for now. As if the engine reads some stuff differently.
If I install on EE2, Spell Thrust removes Spell Shield under GoI. (if you can check on EE2, just install SCS Initialize + Consistent Breach) - it should work.
Same BG2 install - doesn't. ![]()
Interestingly enough, I did find out why I got the Spell Shield to get apsorbed by Breach while under Spell Turning - I used a modified (fixed) Breach version in EE2. (it has to be in your override before either SCS or my mod are installed).
I'll look into it further.
#35
Posté 14 juillet 2015 - 09:07
Ok, this is crazy. If I run my mod on vanilla BG2, it doesn't work. If I install on EE2, Spell Thrust removes Spell Shield under GoI.
That was my hypothesis after Grond0's replication in a non-EE2 install.
Best,
A.
#36
Posté 14 juillet 2015 - 10:29
Another aside. I've fought Kangaxx, Lich and Demi-Lich versions, a few more times with my F/M, Arcadia, in unpatched v30 v10. The new fight and the old fight aren't that different, ultimately. I can still overstock protections, counter attacks on magical defenses, and prevent him from debuffing me, just as I could in v19. The same strategy works. For example, in the most recent v30 experiment he tried:
Khleben's
Ruby Ray
Spell Thrust
Breach
Ruby Ray
Secret Word
Breach
Spell Strike
Spell Strike
Spell Thrust
He was never able to debuff Arcadia, due to use of defensive Contingencies, Triggers, and Sequencers, along with refreshes of lost spell protections. She ended the battle with Spell Protections and Specific Protections in reserve.
As I suspected, I'm more likely to gain an advantage vis-à-vis Spell Shield related bugs in v30, since the Spell Thrust maneuvers against GoI/Spell Shield are more common than the random occurrences of vanilla Spell Shield stickiness, but it's not game break-ingly awful, in contrast to what my initial trials had suggested. It's only a glaring issue in a subset of the AI attack strategies. I just happened to see those strategies in my first trials.
I'm now content to continue Aphril's game, using your Wish fix alone, if you can't get your other patches to work. I'd still love to have a patch -and I think it's something that the broader community may appreciate- but don't go crazy over it for my benefit if you're not having fun with it.
Your Wish patch really helps: Thanks. And further patches would be super-great. But only do it if you're independently motivated.
Thanks for everything Aasim! Really. You've been very helpful.
Best,
A.
#37
Posté 14 juillet 2015 - 10:42
Oookay.
Alesia, check this for me.
Open up spwi321.spl file with NI.
It should have two 146 opcodes (Cast Spell).
One of these is spwi321b. This one should have a power level of 4, should have some opcode that removes spell protections max level 5.
This one isn't important for us now, but must be there.
The second one spwi321c. It should contain a single opcode (Remove protections by type, type "unknown, followed by a number - probably 14) This one takes down Spell Shield. After you install my mod, it should have a power level of 0. Does this happen?
If this will be of any use, in my setup (SCS2 v21 + Fixpack v9.01) spwi321b.spl, which removes spell protections max level 5, have power level 0, so it bypass GoI. The spwi321c.spl, which takes down Spell Shield, have power level 0 as well.
- Alesia_BH et Aasim aiment ceci
#38
Posté 14 juillet 2015 - 10:43
Ok, I've got to go. I'll do some more when I get back home. The version of mod uploaded has Wish Spellstrike (I wanted to see how it works)- if you're using that, and don't want to,
1)unistall,
2) replace spwi903b with spwi513b and spw903c with spwi513c, in criticals/components/spellshield.tpa file so the first piece of code looks like this:
3) reinstall.
- Alesia_BH aime ceci
#39
Posté 14 juillet 2015 - 10:45
If this will be of any use, in my setup (SCS2 v21 + Fixpack v9.01) spwi321b.spl, which removes spell protections max level 5, have power level 0, so it bypass GoI. The spwi321c.spl, which takes down Spell Shield, have power level 0 as well.
Yeah, Alesia wants different behaviour (SS removed, but GoI unremovable). This can be done easilly I think, Breach will be more difficult to change.
- Alesia_BH aime ceci
#40
Posté 14 juillet 2015 - 10:54
If this will be of any use, in my setup (SCS2 v21 + Fixpack v9.01) spwi321b.spl, which removes spell protections max level 5, have power level 0, so it bypass GoI. The spwi321c.spl, which takes down Spell Shield, have power level 0 as well.
Yes. In v9 Spell Thrust bypasses GoI entirely. In v10, it does not: GoI foils Spell Thrust, yielding a spell ineffective message. That was a deliberate change between versions.
We're hoping to preserve the v10 ability of GoI to stop Spell Thrust, while allowing Spell Thrust to interact with Spell Shield prior to interaction with GoI. Ideally, the player should be able to use GoI to protect Spell Immunities from Spell Thrust (in the absence of Spell Shield), but he or she should not be able to use GoI to protect Spell Shield from Spell Thrust.
In some of the randomized scripts, the AI leads with Spell Thrust, the assumption being that the Spell Thrust would have removed the Spell Shield. But if GoI is active, with the current behavior, the Spell Shield remains in place, leading the AI to waste spells.
Best,
A.
EDIT: I see Aasim caught that one already.
#41
Posté 14 juillet 2015 - 11:04
Thanks Serg! And thanks Aasim!
Best,
A.
#42
Posté 14 juillet 2015 - 11:14
Ok, I've got to go. I'll do some more when I get back home. The version of mod uploaded has Wish Spellstrike
Noted. That would explain why I Wish for My Enemies' Magical Defences to Be Utterly Destroyed stopped bouncing off of Spell Turning. I was wondering about that.
Best,
A.
#43
Posté 15 juillet 2015 - 12:29
Just for the record. This was a test conducted in SCS v30 with Consistent Breach installed and Fixpack v10. The Wish-Breach patch developed here is not installed. The character is running the old Spell Shield (226).
Spell Shield and Spell Turning are active
Wish-Breach is cast. Wish-Breach is stopped: Spell Ineffective
The Spell Shield comes down. Breach does not hit Spell Turning
So, with v30, Consistent Breach and Fixpack v10, using the old Spell Shield (226), we get the behavior we desire in repsonse to an enemy Wish, without patching, although with a non-zero probability of random stickiness.
Hopefully we can achieve the same with the new Spell Shield, allowing us to combine the old Spell Shield/Wish-Breach spell interaction with the new Spell Shield's lack of random stickiness.
Thank again for everything, Aasim!
Best,
A.
Btw. There were three trial: 100% success rate.
#44
Posté 15 juillet 2015 - 06:21
Also for the record. This was a test conducted in SCS v30 with Consistent Breach installed and Fixpack v10. The character is running the old Spell Shield (226).
Arcadia is running Globe of Invulnerability and the old Spell Shield (226)
The Yuan-Ti Mage casts Spell Thrust
The Spell Shield comes down.
So, with SCSv30 and the Fixpack v10, using the old Spell Shield, we get the behavior we desire in response to an enemy Spell Thrust, without patching, although with a non-zero probability of random stickiness. As I asserted in the other thread, the old Spell Shield, Spell Thrust, and the new Fixpack GoI work together, at least in the context of hostile NPCs targetting Spell Shield/GoI protected PCs.
NW: Three trials. 100% success rate.
In contrast, here are the result with the new Spell Shield, under the same test conditions.
Arcadia is running GoI and the new Spell Shield.
The Yuan-Ti Mage casts Spell Thrust.
Spell Ineffective
The Spell Shield does not come down.
So, with everything else unchanged, but using the new Spell Shield instead of the old one, we lose desired functionality under the conditions delineated above.
NW: Three trials 100% failure rate.
Best,
A.
Btw: Unsurprisingly, results with Secret Word in SCS v30 and Fixpack v10 are the same. If the new Spell Shield is used, Secret Word will fail to take down Spell Shield in the presence of GoI.
#45
Posté 16 juillet 2015 - 04:21
Btw. I just thought I should add a quick note for anyone who might want to replicate my tests with the vanilla Spell Shield.
Note that in each trial I spawned a hostile NPC mage, rather than having an ally cast a Spell Protection remover. That is an essential step.
Oddly, the behavior of the vanilla Spell Shield is asymmetrical: it behaves differently depending on whether a hostile or non-hostile character has launched the attack. When hostile NPCs launch conventional arcane attacks on a PC's Spell Shield, the failure rate is low. In contrast, if a friendly character launches the same attack on a PC's Spell Shield, the failure rate is high.
In my prior posts, I had been assuming familiarity with that aspect of the vanilla Spell Shield's behavior, and describing my results with restrictive statements such as "at least in the context of hostile NPCs targeting protected PCs." Recently, it occurred to me that some might not have been aware of the asymmetries, and might not have appreciated the significance of the restrictives. Now everyone knows.
Best,
A.
An illustration.
Nalia cast Secret Word at Arcadia: no spell ineffective message, the Spell Shield does not come down. NW: 4 trials 100% failure rate.
With the same buffs active, a Yuan Ti mage casts Secret Word at Arcadia: Spell Ineffective message, the Spell Shield does come down. NW: 4 trials 100% success rate.
#46
Posté 16 juillet 2015 - 06:17
@Aasim. I just setup a EE2 install to test your patch. I haven't had a chance to run systematic trials, yet. I'll get back to you with results.
I did notice that your Wish-Breach patch took and worked (Spell Shield Dispelled). I was glad to see that, although we still need to implement a Breach version. The Wand of Spell Striking looks good, too, in preliminary testing. Nice work! We're making progress.
Here's why I'm posting now: I don't have any conclusive findings yet, but I've observed Ruby Ray failures in EE2 v30 patched, with the new Spell Shield. Neither of those things occurred in my non-EE2 install. Can you or anyone else replicate? I'll document my findings regarding Ruby Ray in EE2 when I get the chance.
Best,
A.
EDIT:
Aasim's Wish-Patch (Spellstrike version) working in EE2 SCS v30 n=2
Enemy Ruby Ray failing to take down a PC Spell Shield (new) in that same setup. n=4
#47
Posté 17 juillet 2015 - 11:58
This is a series of trials conducted in EE2 v1.3 SCS v30 with the new Spell Shield, EE2 v1.3 SCS v30 with the old Spell Shield, and vanilla SCS v30 Fixpack v10 with the new Spell Shield.
Summary of results: In EE2 v1.3, with the new Spell Shield, Ruby Ray fails to take down Spell Shield. In the same install, with the old Spell Shield, Ruby Ray successfully removes Spell Shield, at least in the case of hostile NPCs targeting Spell Shield protected PCs. In vanilla BG, with SCS v30 Fixpack v10 and the new Spell Shield, Ruby Ray successfully removes Spell Shield. In EE2 v1.3, the interaction between Ruby Ray of Reversal and the new Spell Shield appears to be bugged. Replication would be appreciated.
In EE2 v1.3 SCS v30 with the new Spell Shield. Kangaxx casts Ruby Ray. The Spell Shield does not come down (note portrait icon).
Also in EE2 v 1.3 SCS v30 with the new Spell Shield. Arcadia cast Ruby Ray on herself via the Rod of Reversal. The Spell Shield does not come down.
In EE2 v 1.3 with SCS v30 with the old Spell Shield. Kangaxx cast Ruby Ray of Reversal. The Spell Shield comes down (note lack of portrait icon).
In vanilla BG with SCS v30 and Fixpack v10, using the new Spell Shield. Kangaxx casts Ruby Ray, the Spell Shield comes down
In vanilla BG with SCS v30 and Fixpack v10 and the new Spell Shield. Arcadia cast Ruby Ray on herself via the Rod of Reversal. The Spell Shield comes down.
Conclusion: The interaction between Ruby Ray of Reversal and the new Spell Shield appears to be bugged in BG-EE2 v1.3, but not in non-EE BG2. This may be a local issue, however. I have no way of knowing at present. Replication would be appreciated.
Best,
A.
#48
Posté 18 juillet 2015 - 01:38
Hey sorry to Jump in awkwardly, but I've been following this bugfix closely (OCD will not let me play without understanding my spell interactions)
I've been replicating your issues as best I can considering I only use my EE2 setup nowadays. The best way I've found to reliably test all these interactions is to; set up your prebuffs beforehand, create a save while buffed, SK your PC into an enemy (In my install casting debuffs on allies seems to have effect at all), then debuff away with a second character.
With SCSv30 I'm able to remove SS with spell thrust while GoI and ST are active, however Wish breach is still blocked even after Aasim's patch. Further testing shows that my spell thrust is fully bypassing GoI though.... (shouldn't EE already incorporate the fixpackv10 change?)
- Alesia_BH aime ceci
#49
Posté 18 juillet 2015 - 01:30
Hello Relay. Welcome to the forum. It's good to meet you!
I've been replicating your issues as best I can considering I only use my EE2 setup nowadays. The best way I've found to reliably test all these interactions is to; set up your prebuffs beforehand, create a save while buffed, SK your PC into an enemy
Makes sense. That's how I've been testing, too, although I've been CLUAing enemies to my PC.
To test the old Spell Shield in SCS v30, I've been buffing a character in a pre Spell Shield fix version of SCS, saving, and then importing the save. The logic here being that casting Spell Shield in another version will apply Spell Shield (226) to the character and that (226) will still be on the character when the save is loaded in a new environment. Each time, I've checked the CRE file to confirm that Spell Shield (226) is active. I can then CLUA in an enemy to see how the old Spell Shield responds in the SCS v30 Fixpack v10 environment.
however Wish breach is still blocked even after Aasim's patch.
Just to clarify. In vanilla, Wish-Breach is blocked by Spell Shield. That's how it should behave, in my opinion. Wish Breach should be blocked and the Spell Shield should come down. With the old Spell Shield, that how it works, when it works.
With the introduction of the new Spell Shield, Wish-Breach started bypassing Spell Shield in SCS. The reason is that the Ardanis Spell Shield file had not been patched for the Wish-Breach spell that David created for SCS. The new Spell Shield works by conferring immunity to a list of spells. If a spell isn't on that list, then it isn't stopped. David's Wish-Breach isn't on that list.
One of the objectives of our patch is to recover the old Wish-Breach/Spell Shield interaction: we want Wish-Breach and Spell Shield to interact as they used to. Aasim tried to achieve that by patching Wish-Breach so that it cast the regular Breach spell. That approach didn't quite work, however, because with Consistent Breach installed the patched Wish-Breach would bounce of Spell Turning and leave the Spell Shield untouched.
In the most recent version of the patch, Aasim changed Wish-Breach so that is now, effectively, Wish-Spellstrike. That works, in the sense that the spell is blocked by Spell Shield and Spell Shield is removed in the process, but it has the disadvantage of applying the wrong effect: Spellstrike intead of Breach.
If you are seeing Spell Shield stop Wish-Breach after installing the patch, then that is probably the patch working- that's what we intend.
With SCSv30 I'm able to remove SS with spell thrust while GoI and ST are active...Further testing shows that my spell thrust is fully bypassing GoI though.... (shouldn't EE already incorporate the fixpackv10 change?)
Interesting. Is this with Aasim's patch installed or without it? I'd be interested in seeing a with and without pairwise comparison with regards to Spell Thrust/GoI/Spell Shield.
I can't comment on what EE does or doesn't implement. I just bought EE. I was, however, under the impression that Fixpack v10 was included.
Have you done any experiments with Ruby Ray in EE? I'd be interested in hearing your results.
Best,
A.
#50
Posté 18 juillet 2015 - 02:09
With SCSv30 I'm able to remove SS with spell thrust while GoI and ST are active, however Wish breach is still blocked even after Aasim's patch. Further testing shows that my spell thrust is fully bypassing GoI though.... (shouldn't EE already incorporate the fixpackv10 change?)
Btw. I just ran a quick test in EE without Aasim's patch. ST did not remove SS in the presence of GoI.





Retour en haut





























