Sword Coast Stratagems v33 now available

[subtledoctor]

3 hours ago, pochesun said:

Not a good service.

lol @ you think you are owed "service"

To be fair, the support for your mod is way, way worse. So, again, maybe try some perspective.

2 hours ago, Lilaina said:

and three months is nothing compared to Beamdog not fixing bugs they caused years ago. 

Now THAT's what I call perspective! 

[Salk]

@pochesun,

I just want to add that DavidW has been checking back on this Forum pretty much regularly despite the long pauses and when he returned he always worked in order to produce a new version of SCS, trying to gather all the existing bugs report.

If this admittedly quite critical bug with the AI was reported in a timely manner I am sure DavidW'd have fixed it before leaving again but that was not the case. 

We should just be happy to have a previous version to fall back on and, most of all, we should be happy if DavidW ever returns here and decides he still cares enough to bother with yet another version of SCS. He's under no obligation to do so.

 

[morpheus562]

Where can v33.4 be downloaded from?

[subtledoctor]

3 hours ago, morpheus562 said:

Where can v33.4 be downloaded from?

You can get v33.4 from here.

Ordinarily I would not be quick to recommend using an old version of someone's mod. So it might be worth consolidating information about the recent updates.

Caedwyr: "SCS version 33.4 fails to install with the weidu version updated to v247. SCS installs just fine with weidu v246."

DavidW: "v33.5 is now out, addressing the WEIDU v247 issues."

Grodrigues: "The error appears first in the Improved Shapeshift component, but is not fatal, and then explodes in the NPC customization. ... I should add that this is a regression as with SCS v33.4 and weidu 246 everything installed fine."

DavidW: "v33.6 is now out, fixing the problem with grodrigues's install."

DavidW: "Updated to version 33.7, mostly to fix a non-English-language problem that 33.5 created."

It seems to me that 33.5, 33.6, and 33.7 don't actually have much in the way of substantive updates, in other words 33.4 might be a fine version to use. The error that seems to have led to it being changed was an install error, not an error in the behavior of the mod. Which means, if you can install it, it might be fine to use 33.4.

Interestingly, I installed 33.4 just fine with Weidu v247 on my Mac... but I don't know if Caedwyr's install failure was with a component I did not try. (I didn't install all of the tactical options.)

I'm not going to tell people what to do. For me, I have installed 33.4, some but not all AI components and some but not all tactical components, in English, with Weidu v247. And I have not seen anything odd in-game. So using v33.4 is a solution that works for me.

Anyone else reading the thread, can take in the information I linked to above and come to your own conclusions about what to do. (My perspective on the reported 33.7 bug: it wouldn't bother me much, because if a mage runs out of spells, they are dead regardless whether they stand around doing nothing or act "smart" by moving and attacking with a quarterstaff.)

[The_Baffled_King]

Quick post about a few things, tagged @DavidW as it contains a bug report

(1) General praise

This mod is amazing, and thank you so much for all the effort you have put into it. I cannot imagine ever playing BG without it again.

(2) Bug report regarding BG1 Tactical Challenge: Improved Doppelgangers

This component misses some of the doppelgangers in the Merchants' League Estate (Aldeth Sashenstar's quest). I poked around in Near Infinity and am fairly sure I can see where the oversight occured. In brief, there are 4 scenarios to trigger the dopplegangers: (1) attack Zorl; (2) attack Irlentree; (3) give Brandilar the incriminating letters taken from Zorl's office; (4) give Aldeth the incriminating letters taken from Zorl's office.

The first three scenarios result in SetGlobal("ZorlDopple","GLOBAL",1), which (a) makes Zorl transform himself on the middle floor; (b) makes Irlentree transform himself and create 4 doppelgangers on the bottom floor; and (c) makes AR0129.BCS create 10 doppelgangers on the top floor. These scripts all result in the use of your new .CRE, dw#dopzo, which is fine.

The problem is with Scenario (4). Giving Aldeth the incriminating letters taken from Zorl's office leads to State 29, Response 47, Action 12 in ALDETH.DLG, which starts aldcut02.BCS. This is the cutscene which puts the party, the 4 named characters, and 14 merchants on the top floor. The 14 merchants created by aldcut02.BCS are dopald.CRE, which use dopald.BCS. This BCS still uses DOPPLEGANGER_CHANGE2, so they end up turning themselves into the weaker versions of the doppelgangers (they even nerf Zorl and Irlentree, too!).

(3) Pre-battle casting

Just wondering if there is scope to amend this component so it (effectively) pauses combat until the pre-battle casting routine is complete? Could it trigger a cutscene, perhaps? The component is still pretty good, and I'm grateful for it, but it is awkward how the party's AI can take chunks out of enemy casters' defences before their pre-buffs are loaded. One can get around this to some extent by attacking other hostiles while waiting for the caster pre-buffs to load, but that uses up actions, so the player is less likely to be able to interrupt the enemy caster once its buffs are loaded up and the party has started to target the caster(s) again.

[xiaoleiwen]

9 hours ago, subtledoctor said:

Anyone else reading the thread, can take in the information I linked to above and come to your own conclusions about what to do. (My perspective on the reported 33.7 bug: it wouldn't bother me much, because if a mage runs out of spells, they are dead regardless whether they stand around doing nothing or act "smart" by moving and attacking with a quarterstaff.)

I have to agree with this, I don't think I notice anything wrong with the mage after finishing bg2 with 33.7. The only time that I noticed something slightly weird was during the mage vs mage duel in Underark, as the enemy mage stand still after a long fight, and I considered it normal that time because he run out of spells, didn't bother and just continue my game. I probably won't remember it if I didn't see the bug report here..

I guess it is more obvious in BG1, but I think don't mind if the mage run around or stand still after run out of spell...since he is useless then..and I always kill mage ASAP 

Edited March 25, 2021 by marcnivar

[tomasz86]

7 hours ago, The_Baffled_King said:

Just wondering if there is scope to amend this component so it (effectively) pauses combat until the pre-battle casting routine is complete? Could it trigger a cutscene, perhaps? The component is still pretty good, and I'm grateful for it, but it is awkward how the party's AI can take chunks out of enemy casters' defences before their pre-buffs are loaded.

Not only that. One can literally one-strike kill most mages with a backstab, including bosses like Davaeorn (in whose case it wouldn't actually be possible in the original BG, since there he starts with some protections as soon as we enter his lair).

Edited March 25, 2021 by tomasz86

[The_Baffled_King]

3 hours ago, tomasz86 said:

Not only that. One can literally one-strike kill most mages with a backstab, including bosses like Davaeorn (in whose case it wouldn't actually be possible in the original BG, since there he starts with some protections as soon as we enter his lair).

If we are talking about using backstab on unsuspecting targets, then I just choose not to use those tactics on mages or priests. It is less than ideal, but in those circumstances it is easy to account for the drawback so that SCS does what it is intended to do and allows enemies to prepare.

I feel that the problem is more pronounced when a caster sees the party and combat begins. If I have autopause for when enemies are sighted then, much like declining to backstab so that SCS works as intended, I can similarly decline to target the mage until their pre-battle casting is finished. The difference here is that, as combat has begun, I cannot control the consequences of my decision to let SCS work as intended, and the end result is that enemy mages benefit more from my decison than SCS intends, and more than I intend.

The problem is that enemy mages don't say 'right chaps, thanks for the pause, but I've got my buffs cast now, so let's roll initiative in 3... 2... 1. They just start casting and, in the absence of supernatural reflexes that let me pause the game at that point, mages end up getting a head-start in what they want to do. This scenario can be worse when the party has used its actions on other hostile creatures.

I just wonder if it is feasible for the game to do something a little more invasive that basically makes the enemies and players play fair, so that combat as a whole effectively begins only after pre-battle casting has finished? Alternatively, if the pre-cast spells triggered the auto-pause in the same way that normal spells can (with that option enabled), then I think that would also be an improvement. Either way, I'm grateful for the component and plan to keep using it.

Edited March 25, 2021 by The_Baffled_King

[Lilaina]

Mages also don't seem to attack with Melf's minute meteors on higher difficulties, at least in BG1 EE, so the bug does make mage encounters easier before they run out of spells as well.

Edited March 25, 2021 by Lilaina

[KaleNich]

I got issues with Mulaheys Elite Kobolds. They are just standing there doing nothing. It annoys me because in an older version when it worked it was a beautiful and chaotic fight. I tried it with and without the improved Kobolds component. Using the 33.7 version and Enhanced Edition.

My mods:

Unfinished Business v16.0

NPC project v1.25

Rogue Rebalancing v492

Item Revision v4b10+IR Revised 1.3.4

Spell Revision v4b18+SR Revised

Tweak Anthology v9

Atweaks v453

SCS 

[Greenhorn]

Well, after roughly half an hour of installation newest SCS 33.7 yet again refused to install on my easy TUTU. Initialise AI components (required for all tactical and AI components) failed with following error:

Quote

ERROR locating resource for 'COPY'
Resource [ELAIRSU4.cre] not found in KEY file:
    [./chitin.key]
Stopping installation because of error.

ERROR Installing [Initialise AI components (required for all tactical and AI components)], rolling back to previous state
Unable to Unlink [weidu_external/backup/stratagems/5900/OTHER.5900]: Unix.Unix_error(1, "unlink", "weidu_external/backup/stratagems/5900/OTHER.5900")
Will uninstall 5761 files for [STRATAGEMS/SETUP-STRATAGEMS.TP2] component 5900.

 

[GrimJim]

I think I figured out the issue with mage AI in 33.5 and up. It looks like it has to do with the file stratagems/mage/ssl/meleeblocks/standard_mage.ssl. For whatever reason, 33.5 completely deleted the contents of this file which seems to handle the usage of Melf's Meteors and other attacks. I think it can be fixed by replacing the contents of this file with the last good version. You then just need to change line 250 from

INCLUDE FILE(%scsroot%/genai/ssl/shoot.ssl)

to

INCLUDE FILE(%MOD_FOLDER%/genai/ssl/shoot.ssl)

and then reinstall the Smarter Mages component. 

[Bartimaeus]

Man, for how many files I looked through and compared, somehow I missed the contents of something literally right in the stratagems\mage\ folder being straight up missing. Good job, me...and actually good job you for spotting it!

Edited April 5, 2021 by Bartimaeus

[Salk]

Good job, @GrimJim!

Thanks for the investigation and for offering us a solution.

Edited April 7, 2021 by Salk

[The_Baffled_King]

@GrimJim, +1

Thanks!

I still have v33.4 installed, so I was following this with interest.