DavidW

I'm not so worried about players becoming more effective. But I think this will be seriously useful to enemy wizards. 9th level spell slots are a fairly significant constraint on my scripting. (That's only an issue as and when I take SR properly into account, but doing so has got most of the way to the top of my to-do list.)

In NWN epic spells weren't consuming any slots. Same here - you pick an epic spell, it appears in your innate bar, while all 9th slots are used for memorizing 9th only spells. Careful with this change. Wizards could end up with a lot more high-level spells, and they're arguably powerful enough as it is.

Re Sunfire: the fact that the caster is immune is a significant advantage over Fireball when used by enemy wizards. Even non-Globed wizards can use it. (Of course, that's somewhat less crucial when it's used by a PC: as usual, party-unfriendly or caster-unfriendly spells are significantly more useful in the hands of the PC than of the computer.)

Actually I have reduced their arrows from an initial 20 to 10 for a reason. The Hobgoblin Shaman you get at 12th lvl goes in melee after casting a bunch of buffs, and I wanted his allies to go with him instead of leaving him alone and continue to fire arrows from afar. Perhaps I can script them to follow him in melee even if they don't run out of arrows. I'll try to test them a little when I get back to SR. Get the hobgoblin shaman to Shout(39) when he goes to melee. (I'm pretty sure SCS doesn't use that number.) Get the hobgoblins to set some local variable to 1 if they hear that shout from someone with the same allegiance as them. (You need the allegiance check in case there are two versions of MS1 running concurrently.) If the variable is set to 1, get them to EquipMostDamagingMelee and get in there.

This is a bit moot, as I'm not keen to have Breach not affect Stoneskin.

Agreed (this is on my to-do list too). But actually, the awkwardness in both cases is that ideally the animation should be visible when the player casts it but not when enemies do. But implementing that in SR is probably impossible. (I can do it in SCS because enemies can use modified versions of the spell.) If you like, leave the invisibility off Sanctuary and the graphic on, and I'll offer it as a tweak, for enemies only, in SCS. I don't think enemy spellcasters in vanilla use Sanctuary anyway. Fine with me, but my point was that this way players would see AI clerics vanish in an instant (as per Invisibility spell) which is clearly not what you'd expect from Sanctuary. Well, it is what you'd expect from sanctuary if you've installed a tweak that says that. Party-unfriendliness is much less of a problem for the player, because they can just get out of the way. My own experience playing against Sanctuarying clerics says that iterated Fireballs (and the like) are a pretty effective counter. Although I'm answering a particular question, I think the point is going to apply more broadly. In some sense, I'm just not interesting in greatly enhancing my possibilities. What I find interesting about tactical coding for BG2 is taking a framework over which I don't have control, and trying to find the most interesting and optimal strategies and encounters within that framework. Now, I'm not religious about it - occasionally I make a localised change because I think it drastically improves the puzzle, or because something is just broken or boring without a tweak, or occasionally just because it looks fun, but overall I'm really serious about the minimalism thing. It's very core to SCS's flavour, and quite core to my own interest in writing it. Having said all of which, it's also interesting to work in other peoples' frameworks, which is why I'm generally quite keen (in principle) to write AI that makes allowance for SR. That probably goes for potions too, in principle, but it would always be as some additional step, not as part of the core project. Feedback really. (Plus, if you can see your clothes, why can't you see your Fire Shield?) I don't think it's a big thing though. I think I've given the wrong impression in some of this discussion. I am interested in improving priests' effectiveness, but I intended to convey that as an interesting puzzle and to seek insight into what makes it difficult, possibly with a view to very small shifts that might make things more interesting, possibly just in the hope of getting good suggestions. (The Sanctuary thing falls into the former category, especially now I realise that it already has an Invisibility effect when the player casts it.) It's not the kind of thing that makes me want to make big changes in vanilla's status quo. The point about fighter/clerics is well taken. I do play with cleric/mages a bit, but they're quite hard work and also I'm not really comfortable making in-game clerics into cleric/mages, given the flavour is so different.

Agreed (this is on my to-do list too). But actually, the awkwardness in both cases is that ideally the animation should be visible when the player casts it but not when enemies do. But implementing that in SR is probably impossible. (I can do it in SCS because enemies can use modified versions of the spell.) If you like, leave the invisibility off Sanctuary and the graphic on, and I'll offer it as a tweak, for enemies only, in SCS. I don't think enemy spellcasters in vanilla use Sanctuary anyway.

Unless I'm horribly and utterly mistaken, IDSs are static, unlike dynamic NearestEnemyOf(). You'd have to go with 2nd param of 0, protected from everyone.If you're right about IDS being static then we have to find a completely different solution because I tested the "protected from everyone" solution and it's not a viable solution, as then even the caster can't target himself. @David, I've looked again at AD&D Sanctuary description and it says "If the saving throw is failed, the opponent loses track of and totally ignores the warded creature for the duration of the spell. Those not attempting to attack the subject remain unaffected. Note that this spell does not prevent the operation of area attacks fireball, ice storm, and so on)." Thus, on one side I was right about it not granting protection from fireball-like spells, on the other your idea about adding an "invisibility" state isn't so far from that "losing track of". Still, I don't particularly like the whole "invisibility" thing as a concept, and on balance I think it's slightly too much considering Invisibility is a 2nd lvl arcane spells and Sanctuary would become pratically the same thing but better. Sanctuary would have faster casting time (which is a big plus), immunity to divinations (which is a huge plus) but shorter duration (not that I ever used Invisibiliy to remain invisible for long). I don't feel strongly. For the sake of debate, though: (i) I at least use invisibility for longish durations (ii) Sanctuary is self-only, which is a fairly crucial disadvantage (in particular, it's not well suited to scouting, unless the party has a cleric/thief I guess) (iii) Most importantly, Sanctuary already functions as an invisibility spell when cast by the player; this is just a matter of evening the score

Is this a bug, or something legit but annoying?

Well, in PnP nothing prevents you from casting a Fireball right next to the protected creature. You can even cast it directly at the sanctuaried target if you win a check. I'm going from memory, but my recollection (of the 2nd edition spell) was that, if you fail a check, you can't attack the sanctuaried creature in any way. That includes casting a fireball with the intention of hitting the creature, though it doesn't preclude casting a fireball at someone else, with the foreseen but unintended consequence of hitting the creature (the Catholic Church would love Sanctuary). Of course, that kind of rule is totally unadjudicable in a CRPG; it's also the kind of delicate thing that tended to get phased out of 3rd edition D&D, so I wouldn't be surprised if 3rd edition Sanctuary is different. Not on my account! I can't at all guarantee I'm going to use it enough to justify that.

From the AI's point of view, undispellable Invisibility and Protection from Creature Type are functionally equivalent. And either way, there's a significant difference between how the player can react and how the AI can react. I can kind of work around it, but it's tricky, and I'm not sure I can afford the script space to do it properly. Up to you about PnP flavour, but I'm not entirely convinced. On my interpretation of PnP sanctuary, it's not okay to dump a fireball on a sanctuaried target, whereas that's straightforward (and the obvious counter to Sanctuary) in BG2 without the invisibility.

I'm ambivalent about whether it's a good idea but at worst I think it's not going to cause major problems; at best I think it's pretty good. As long as the spell doesn't actively hide the graphic, though, it's definitely more useful for players than the enemy, as PCs can see Sanctuaried creatures but the computer can't. In your place I might actually consider slapping on an Invisibility effect. Re prebuffing: I'll give it some thought. You're right that it's a bit of a pain to implement, but not horrendously so.

Can you clarify the motivation? I'd rather not, because I'd rather err on the side of transparency. The enemy wizards know what spells you're using, after all.

Until Ch.7 there're but four places with rakshasas - Sewers (easy to beat at low levels, iirc he doesn't even know PFMW), Ruhk the Transmuter in Windspear (he's almost alone), Ihtafeer's group (optional, as you can kill genies instead to complete the quest), Spellhold (he's almost alone).So I'm largely unpersuaded there's any need to think of letting low-level party to take down a rakshasa. And against Suldanessellar's guys we already have Pierce Shield and Spellstrike. As well as Shapechange into Iron Golem Yeah, I wasn't all that worried about Rakshasas. It was liches that bothered me more. (Plus, even high-level characters have a problem if they can't use Breach.) Returning to the general II issue, I think I'm basically resolved to at least try out the Ardanis solution in the next SCS release. Having thought about the coding a bit, I don't think it would be too tricky to make SCS sensitive to whether that option is installed, rather than just assuming it willy-nilly as I normally do for defensive-magic shifts. It's just a matter of (a) slightly prioritising Truesight when the option is installed, which I can do in SSL via IgnoreBlock trickery; (b) defining a SpellAntimagic option which gets compiled as SpellArea or Spell as appropriate; © giving mages/priests instructions not to actively target creatures with STATE_INVISIBLE active. It's not a hell of a lot of work to try, and I'm very keen on the virtues of experimental test for these things - theory only gets you so far.

Most of these solutions are (or, I suppose, should be: I don't know how you've coded them) blocked by Minor Globe (which in turn can't be dropped by single-target antimagic when II is active).You're right, though Spell Thrust do work against (M)GoI, removing the former and ignoring the latter (thus still dispelling SI even if protected by GoI), because both in vanilla (at least in my vanilla install it looks so) and SR it has 'power' set to 0 (a la Dispel Magic). Only if it has an area effect (and wasn't the whole point of this discussion to avoid there being one?)

Against a standard AI a first level guaranteed invisibility -spell, for a backstabber/Harm'er, whell yeah. The discussion is basically about enemy use of Sanctuary, not PC use. And my point is that area of effect magic still works against the Sanctuaried. (I can't recall whether SCS attacks Sanctuaried PCs with it.)

Indeed. But is Sanctuary really so effective then? What about area-effect magic?

It doesn't bother me, because it's not an effect that works on the protected creature, it's an effect that works on the caster. That is: one is not stripping Invisibility from the protected creature, one is granting the ability to see through invisibility to the caster. (In the same way, Detect Invisibility in PnP would not be blocked by (e.g.) Minor Globe, because it's doing something to the caster, not to the protected creature.) It doesn't bother me That's basically my maths. Old framework: you need TS running and then you need to drop SI:Div. New framework: you need TS running and then you need to drop Spell Shield. Overall it's a wash. Try it and see. That's exactly my point: the AI (player's or enemy's) can see the enemy, but the player can't. I was thinking about the vanilla versions, which (iirc) have zero duration. But not Minor Globe. Most of these solutions are (or, I suppose, should be: I don't know how you've coded them) blocked by Minor Globe (which in turn can't be dropped by single-target antimagic when II is active). That's a good point; I ought to bear it in mind in scripting. (Though again, Ardanis's modification of TS still works.) I'm really not keen on changing the vanilla spell system, or (even more so!) the vanilla powers of creatures, within SCS except minimally and as an a last resort. To do otherwise obviates the basic design principles behind SCS. (And no, I don't have any detailed logic beyond my own personal intuitions as to why I regard any given change as more minimal than any other - the nearest I can do is "what makes things you already see in the vanilla game have to be changed", though that's highly imperfect.) SR, of course, can do as it wishes with the spells, so this is only relevant insofar as compatibility is concerned. That's fine with me, even to the stage of being something I'm amenable to including. (Case in point as far as "what counts as minimal" is concerned.)

I think I'm with you on balance, though it's a neat idea. But what ill consequences it could create? What I have initially suggested was to make a divination spell (either by altering TS or by introducing another one and swapping names with TS) capable of only detecting-invis-by-script, without any kind of dispelling enemy's illusionary protections. Only caster can target the invisible opponent with spells, not their party. I'm reconsidering this. (Not least because my latest attempt to make antimagic spells penetrate II doesn't work.) I don't think it would work for any spell only to grant the detect invisibility opcode, because that opcode doesn't actually reveal creatures that are invisible to the player, it just allows them to target II creatures (and, I think, their AI to target invisible creatures). But I think adding it to one or two anti-invisibility spells might be a nice feature, and a neater way around my mage-targetting problem than the never-ideal antimagic AoE. I'd probably add it to Truesight without otherwise altering the spell; I'd also consider adding it (with, say, a 5-round duration - not as something that could be used in prebuffs) to Detect Invisibility. I'd leave it off Oracle and Invisibility Purge, since they're clearly described as instant magic attacks on an enemy's illusions, not as boosts on the caster. This slightly disadvantages mages relative to the status quo, largely because Truesight can be pre-cast. But I can live with that (especially as I'm going to be able to throw Spell Shield around now). Two problems: 1) (Vanilla) non-detection won't provide protection. I don't think this bothers me, ultimately: Truesight et al wouldn't actually reveal a genuinely invisible creature; it would just make an already-revealed creature targetable. 2) More seriously, the player can't give attack orders on totally-invisible creatures but the player's AI can. (This is a symptom of the fact that the invisibility detection opcode isn't really designed for PCs.) I think the only way to solve this is to replace See(XYZ) with See(XYZ)!StateCheck(XYZ,STATE_INVISIBLE) in PC scripts. That could be automated, though there'd be a problem if someone installed a new script after the relevant mod. But I think we can live with this.

I'm somewhat reluctant to do this, as it's not clear why SI:Div in particular should work this way. It is certainly a matter of taste - a quibble: this is not quite the same thing as being subjective We're talking about elegance in the sense used in maths or physics, yes? If so, take it from me: it's subjective. (I speak as someone who works in theoretical physics, and the philosophy of science, professionally.) That's not to say that there aren't often strong consensus views on given issues, though. ... or perhaps I'm misunderstanding. Is this intended as a point in philosophy of aesthetics? I'm fairly unsympathetic to realism as a philosophy of aesthetics (whatever its virtues in ethics), but I can see that an aesthetic realist would want to argue that matters of taste are objective. (Apologies for wild offtopicness.)

@Amanasleep: Contingency does feel a bit of a bridge too far for priests. I do use Harm and Slay Living, but it's fiddly to use effectively because it's very short range and requires a hit roll. I don't take advantage of that anyway, really - SCS isn't intelligent enough to pick on someone because they have both combat and specific protections running.

If you mean Non-Detection, then altering it doesn't (alone) help, as SI:Div is still out there. So the right comparison is with altering Non-Detection and removing SI. I don't really know what "almost a fix" means. (And I don't think there's any evidence that changing N-D is a fix according to the developer-intent definition of fix.) At some level this just becomes a matter of personal opinion: what seems more low-key (in terms of the feel of playing the game, as well as in terms of the in-game lore). I don't have an active justification. (I do have a relatively straightforward in-game justification: just assume that the relevant spells have a small area of effect. Of course this is what I tried to actually do, but we both know it works imperfectly for various reasons.) 'Fraid so.

The general point is that minimalism is defined - for me - relative to BG2 itself. I'm not particularly bothered by fidelity or lack of fidelity to PnP. And in BG2 you can see II characters. I think I'm with you on balance, though it's a neat idea. I think that's a good idea, sufficiently good that I might consider stealing it. (It does affect my AI scripts a little bit, but I can work around that.) I don't agree, actually. Of course it's seriously bad news for a wizard to be Breached, but contingencies and sequencers let the wizard get his defences back up quickly and keep attacking. No, you're right. Absolutely.

That's not the original state of BG2. "Indeed, my friends, none of you have any weapon that could hurt me" - Gandalf. Maybe we could make a rule that enemy priests always come prebuffed with some of the longest lasting protections, particularly Shield of the Archons, Death Ward, Remove Fear, and Chaotic Commands (maybe even Skeleton Warriors, since they last for hours--I know I'd cast one every morning if I thought there was even the slightest chance I'd meet danger that day!). If their first move was Sanctuary followed by Blade Barrier and Physical Mirror, they could end up pretty well protected when they cast Gate or whatever. I also believe that UnHoly Word doesn't break Sanctuary. In any event, I might suggest that normally prebuffed clerics start with Regeneration active, which could dramatically increase survivability and reduce the necessity for potions (and more importantly, spell actions) during combat. I use most of these, but ultimately it's still the case that one antimagic attack plus breach takes nearly all of it all down. ... I'm not theorising here: it's an experimental observation that enemy clerics (especially solo ones) go down very quickly in SCS, quicker than I'd like. (Though this isn't disastrous.) I don't think their spells casting time is the real issue, as they aren't much different from arcane's spells, My observations are that it matters (again, this is behaviour noted in playthroughs and testing - I'm not theorising). In two ways, in fact: (i) at lower levels, the difference is pretty key: 2 seconds for a wizard's 2nd level attack spell, vs. most of the round for a cleric's; (ii) wizards have certain instant-cast moves (stoneskin, sequencers, contingency) unavailable to clerics. Not for me: it's much too radical a change to the BG2 rule system. (And my comments about clerics are meant to illuminate certain aspects of the system, and also to invite suggestions as to how to use them better; they're not a crie de coeur and I don't feel something game-mechanical needs to change to work around this. In practice I'm resigned to not being able to use clerics solo without special arrangements; I can use them effectively as parts of groups.) Guilty as charged, at least where the antimagic system is concerned. (I'm very unkeen on saves for a spell like Breach, because you end up having to carry silly numbers of them; this is less of an issue with non-detection.) I could try to offer a defences of the scissor-paper-rock debuff system, but ultimately I just see it as quite core to the way BG2 antimagic works, so I don't really want to mess with that within SCS.Yep, but we're not speaking of antimagic attacks here but divination attacks, and Non-detection isn't a spell protection. If you want something that temporary shields II but goes down in a single hit you could make Non-detection go down against the first divination attack without a save. That would make it a sort of anti-divination version of Spell Shield. Perhaps it's worth getting the parameters clearer here. Remember that SCS, unlike SR (legitimately) put's a high value on minimalism: I don't want to change the spell system more than I absolutely have to. (I'm in general amenable to incorporating other people's spell systems, but - as discussed before - mage buffing and debuffing is a special case.) The change I've felt forced to make is the AoE for antimagic, which in turn is a workaround for antimagic spells actually allowing targetting of invisible creatures, which was my ideal. At the time I couldn't see how to do it (actually I think now I might, which changes things, but I haven't had a chance to run tests). Other changes have to be measured against that baseline. (As always I'm perfectly happy to try to play nice, e.g. with SR's proposed silent overwrite of SI, where it doesn't mess with my scripts.) Reasonably sure, yes. The balance between debuffing and attacking got tweaked many, many times in SCS testing; where it is now is largely a function of what seems to work in practice rather than of theory. It's doubtless not perfect (especially as the right level of debuffing is very situation-dependent), is probably a bit skewed towards high-level fights where most PCs will be heavily buffed, and could no doubt be refined. But I'm pretty sure I need access to a single-target buff-remover. (In late-game anti-mage fights, in earlier versions I wasn't using Breach enough, with the result that nearly all the party were immune to nearly all the enemy's attacks.) Pierce magic in particular is underused, but 6th level spell slots aren't (Death, PMW, Truesight, etc). 5th level is quite a convenient place for the relevant breaching spell to be: I think it makes really quite a big difference to have it at 6th level. I'm also reluctant to have to load up with two different lots of spells where previously I only needed one. (& yes, I could get around this by using scrolls. But to do that for this one situation seems - to borrow a term from earlier - inelegant, and also to load the PC up with useful antimagic scrolls.) On exploration of the issue, I'm starting to feel less confident that we really need to weaken Breach. Having said that, if you're really sold on breaking specific protections into Pierce Magic and leaving Breach for combat protections only, I can probably semi-incorporate that into SCS ("semi" because I'd cheat, and hot-swap Breach for Pierce Magic at casting time.) It would cost me two or three blocks, but that's tolerable. Well, the original opcode 221 cannot be used to take down only a limited amount of spells, but in theory we might be able to find a workaround if we really think such solution would be better than any other one. I'm persuaded by Ardanis's objection. Also, elegance can be subjective...

Sanctuary isn't terrible (and I probably underuse it). But the thing is, priests only have about one (at most) really good defensive buff they can put up anyway. They don't have the layered defences available to mages.