Some comments after completing IA (some spoilers) Options

[coaster]

Sikret

I've now completed IA (+Ascension) with my F/M & party (Jaheira as Ranger/Cleric, Keldorn, Imoen as Swashbuckler, Nalia & Sarevok). Altogether, it is an excellent mod (perhaps "overhaul" is more accurate than "mod" given how comprehensive it is), and thank you for all the time/effort you have spent on it.

I particularly liked:

-Elemental golems (the gating in of new golems was a real incentive to kill off the elemental golem quickly)
-Most of the improved mage encounters (caveat below) - the deadly Chain Contingencies caught me out a few times
-the added content (new encounter with Memory of the Apprenti in the Docks caught me out too but was a lot of fun!)

Some specific points I was less keen on/possible improvements:

-Been mentioned elsewhere numerous times, but the reliance on Ruby Ray of Reversal as the main (and often only) way of dispelling spell protections got to me after a while. It seems to reduce diversity in encounters and makes many of the higher level spell removals (eg. Spellstrike) of rather lesser importance. Could some mages take alternative approaches - eg. some with "uber aggressive" scripts, like fire chain contingencies on sight, rather than slamming up the defensive shields all the time?

-Reduce fire/cold resistances. I can understand why some enemies had the ability to reduce resistances, otherwise you could just cast protections from fire/cold+Immunity:Abjuration etc and chop the enemy to pieces. But some enemies seemed to fire off multiple instances of these every few seconds, which I found quite annoying for various reasons:
1/ Sometimes my fire/cold resistances ended up in the minuses - how does the engine cope with negative resistances? I think there should at least be a lower cap of zero.
2/ The number of times these "lower resistances" were cast on my party seemed excessive. I had little opportunity to counteract them with spells/scrolls/potions and had to rely on brute force to kill off the enemy quickly.

-Is it possible to put an upper limit on the total number of spawns which are on the map at the same time? In the golem encounter in WK, the snake encounter in Windspear etc, the game engine started chugging badly because of the number of enemies (my computer is fairly high spec so I assume it is the engine which is struggling). If the total number of enemies spawned at a single time could be capped at (say) 20 (if you killed one of the 20, a new one could be free to spawn) this might help. Not sure if it is feasible or not.

-Again, sometime similar has been mentioned elsewhere, but is there any possibility of creating some new higher level spells (either "proper" spells or cast via items) which give mass buffing, eg. mass improved haste, mass chaotic commands etc? This is less to do with game balance than convenience, and reducing the time consuming process of casting individual spells on most or all party members.

Anyway, thanks again for a great mod & I look forward to future versions!

coaster

[Baronius]

It's true that excessive number of enemies on the map at the same time causes slow-downs even if you have the best computer and I think that it's possible for Sikret to limit this number, though personally I don't agree.

If the player doesn't experience the limit's effects, it would virtually be equivalent to the current solution. More concretely:

To my knowledge, most of the spawned creatures are usually created out of the party's sight. These enemies have (big) role in the slowdown, while the player doesn't even know about them if they haven't been encountered before. Theoretically, dynamic spawning could be a solution:

IA would allow a maximum of N spawned creature objects in the area, and if this value is reached due to newer spawns, IA would only count the number of additional spawned creatures over the limit (but wouldn't spawn them physically). Why is this good? It could be ensured that as soon as the current number of creature objects gets under the limit AND "virtual" spawns (i.e. the value over limit, which we count) is non-zero, IA would always compensate it by automatically created the virtual creatures. This is a possible solution, because only those enemies count that are encountered by the player at the moment: the situation of encountering K spawned creatures when there aren't any more is almost identical to the case when there are L (L is non-zero) other enemies out of the party's sight (so K+L were spawned altogether). This is correct as long as the player isn't intended to encounter more than K creatures at a time.

For example:

The limit is 50 creatures. There are 44 creatures in the area at the moment. In the next few minutes, the player slays 20 creatures but 40 new are spawned meanwhile. The limit is obviously exceeded meanwhile, because 44-20+40 = 64. During all this, IA would ensure that at every moment when there would be more than 50 creatures, there are actually only 50 creature objects -- and the counter would store the additional ones. For example, the counter would be 14 when the 20th creature (killed by the player) dies.

How is this different from simply limiting the number? The virtual creatures are physically spawned as soon as possible, the player wouldn't get away with it. For example, if the counter is 13 and there are 50 creature objects at the beginning, and the party slays 50/50 creatures while only 40 other are spawned, there wouldn't be 40 creatures at the end. The counter's value would be 3 and there would still be 50 creatures! (Why is this possible? Virtual spawn is basically instantaneous, unlike the regular spawn points which have a given speed. Just imagine the difference between the simple limit and this solution when a SunFire hits 42 kobolds when the spawn limit is 50.)

Advantages:
* Spawns over the limit aren't thrown out: they are stored and virtual creatures are spawned "on-the-fly" (this is similar to a receiver-side FIFO)
* Player should practically experience the same as without any limit, but with smaller performance problems

I wasn't thinking on its exact implementation though.

P.S.: As far as I can remember, reports of many players reflected that slowdowns related to visual effects (e.g. in tremendous IA battles) happen more often and are much worse than those caused by too many objects. Its opposite was more typical to BG1.

This post has been edited by Baronius: Jun 3 2007, 04:30 AM

[thetruth]

P.S.: As far as I can remember, reports of many players reflected that slowdowns related to visual effects (e.g. in tremendous IA battles) happen more often and are much worse than those caused by too many objects. Its opposite was more typical to BG1.

Probably yes, maybe with the exception of the Improved Oasis and YagaShura's battle. In these 2 battles I have noticed that disabling Static Animations related effects helps a lot.

I like the idea of dynamic spawning though.