24 réponses à ce sujet

[Groove Widdit]

Something I've always wondered about is why are the casting times for all the spells the same in NW2? I hope it's not because they did that in DD3.5 - that would be a horrible sterilization. I've never played 3.5, so I don't know what they did to the greatest game ever created in that version.

[metatheurgist]

Never noticed, I assume it's got to do with the "actionization" of the game. In PnP 3.x the system changed. Spells have a casting time of immediate, free action, standard action, or 1 round or more rounds.

[Tchos]

The casting time implementation is the biggest problem in my mind with realtime-with-pause gameplay on a turn-based system.  To answer your question, no, 3.5 does not do that.  If you cast a fireball in 3.5 pen & paper, for instance, then you have pinpoint precision for exactly where the splash radius is, and who is caught in it, because no one is moving out of their turn.

[Groove Widdit]

I love NW2, but the casting time thing is huckin lame.

[Dann-J]

Some spells do cast noticably faster than others. You can usually get 'Mirror Image' up almost instantly.

 

The problem with long casting times in this game is that the AI tends to interupt the spell if you're not controlling the caster. The 'True Name' scrolls in the OC can be particularly annoying in that regard.

[Groove Widdit]

I've never noticed that. My character is mage, though, so I'm usually controlling him. I only use NPCs if they are fighters.

[Lance Botelle]

Hi,

Well, according to the spells.2da there are a couple of times applied to spells. Note the following:-

1) CONJ TIME: Time in ms and varies from 0 - 2000, but most spells are 1500.
2) CAST TIME: Time in ms again and varies from 0 - 3000.

This page (http://nwn2.wikia.com/wiki/Spells.2da) says the CAST TIME value equates to "Time between the casting of the spell and when you can move the caster." And this page ("http://nwn.wikia.com/wiki/Spells.2da") says, "These fire when the Conj section completes. The spell script triggers when this starts. Effects from the spell script should be applied as close as possible to when this time finishes.

Therefore, I would assume, the game engine "cast spells" from almost instantly to around 5 seconds into the six second round, which seems a fair timing to me, especially when the caster can take feats to help improve the reliability of casting said spell during combat. i.e. To help it being uninterrupted.

Personally, I believe most spells should fall into this "within a round" category. Spells that take longer than a round can be role-played or scripted by the builder, and should be plot related type events. This would make sense as well really, as how often is a spell caster going to rely on a spell that takes longer than a round for a combat situation? i.e. Most spells may appear to be "similar" but that is due to the nature of them that they are designed to be used within a combat situation.

Lance.

[I_Raps]

Well, according to the spells.2da there are a couple of times applied to spells. Note the following:-

1) CONJ TIME: Time in ms and varies from 0 - 2000, but most spells are 1500.
2) CAST TIME: Time in ms again and varies from 0 - 3000.
 

 

 

 

As long as this topic is up - can anyone say reliably that Quicken Spell does anything?  I mean actually, not on paper.

 

[kevL]

i just did a test

 

1. Mage Armor

2. Light (quicken)

3. Stoneskin

 

The action indicator for Light counted down twice as fast ... ie. 3 second round. Or so it seemed,

 

console:

givefeat 29

[I_Raps]

My computer is on the fritz*, but I've thought of a good experiment.  Two wizards, come to complete rest for a round or two to clear their actions, pause, queue up the same several spells but Wizard A uses the normal ones, Wizard B uses the Quickened ones, unpause and watch what happens.  Sounds like a job for Sand...

 

 

* I'm using an eleven year old laptop (with XP) and I can tell you this - websites don't even pretend to be efficient anymore.  Give 'em an inch ...

[I_Raps]

I can also see why it fails (if in fact it does fail, which I strongly suspect).  

 

First, to clarify what I mean by failing - the advantage of Quicken in p&p is that it allows two spells in a round.  This is why it costs so much in terms of higher spell level, but it's still worth it.  And this is what I don't think happens in NWN2.

 

From what Lance posted above, I would bet a Quickened spell foregoes the CAST TIME - i.e. it only uses the CONJ TIME.  But they didn't think it through - that saves a half-round, but because of the real-time nature of NWN2 they're still trying to stuff three half-rounds into one round.  To make it work, they should drop the CAST TIME from the next spell as well.

 

One more for the hard-code wish list I suspect.

[Lance Botelle]

As long as this topic is up - can anyone say reliably that Quicken Spell does anything?  I mean actually, not on paper.

Hi I_Raps,

It looks like KevL tested that one out ... but I do like you test idea. Maybe something the original poster might want to try.

With reference to casting another spell, this is how the Quicken feat is described (copied directly from the tlk file). I personally never use metamagic of any sorts and so cannot comment. Maybe somebody with experience can comment on this:-

Type of Feat: Metamagic
Prerequisite: Ability to cast 4th-level spells.
Specifics: Quickened spells can be cast instantaneously, making them invulnerable to interruption. A quickened spell is cast as a free action, and another action can be attempted within the same round, including the casting of another spell. Only one quickened spell can be cast per round.
Use: Quickened spells occupy spell slots four levels higher than normal. Saving throws and opposed rolls, such as those made when <i>dispel magic</i> is cast, are not affected.

Lance.

[kevL]

here's something a bit more comprehensive: 2 wizards, 4 light spells ea.

start w/ everything flat, rested, etc.

the Quickened spells go off twice as fast. That is, 'precisely' when the 4 quickened action rounds are *completed* the slow caster was just beginning to cast his 3rd Light.


re. animations. There seems to be a wind-up animation, then the cast animation. For the quickened casts, the wind-up would get interrupted by the cast animation before it nearly completes. For non-Quickened, the caster holds the wind-up animation for a few moments before beginning the cast animation. Only after the cast animation is under way does the spell go off.

By changing ConjTime to 0, the wind-up no longer plays. The spell casts (almost) instantly (the cast animation still seems to play before the spell goes off, and takes a fraction of a second). This is very similar to Quickened behavior.

So ConjTime controls how long before a spell gets cast. Additional tinkerings show that CastTime begins *after* ConjTime is done, and then immobilizes the caster for its duration.

If ConjTime is set above 6 seconds, the Action indicator 'sticks' until the cast is ready to pop.


I'm checking this out because sooner or later am going to change my Spells.2da to rule: 1 second casttime per spellLevel.

/cheers

[Groove Widdit]

&nbsp;

Hi I_Raps,

It looks like KevL tested that one out ... but I do like you test idea. Maybe something the original poster might want to try.

With reference to casting another spell, this is how the Quicken feat is described (copied directly from the tlk file). I personally never use metamagic of any sorts and so cannot comment. Maybe somebody with experience can comment on this:-

Type of Feat: Metamagic
Prerequisite: Ability to cast 4th-level spells.
Specifics: Quickened spells can be cast instantaneously, making them invulnerable to interruption. A quickened spell is cast as a free action, and another action can be attempted within the same round, including the casting of another spell. Only one quickened spell can be cast per round.
Use: Quickened spells occupy spell slots four levels higher than normal. Saving throws and opposed rolls, such as those made when &lt;i&gt;dispel magic&lt;/i&gt; is cast, are not affected.

Lance.

&nbsp;
I actually was planning on taking quicken for my next feat. I don't want it if it doesn't do anything, though.

[Lance Botelle]

here's something a bit more comprehensive: 2 wizards, 4 light spells ea.

start w/ everything flat, rested, etc.

the Quickened spells go off twice as fast. That is, 'precisely' when the 4 quickened action rounds are *completed* the slow caster was just beginning to cast his 3rd Light.


re. animations. There seems to be a wind-up animation, then the cast animation. For the quickened casts, the wind-up would get interrupted by the cast animation before it nearly completes. For non-Quickened, the caster holds the wind-up animation for a few moments before beginning the cast animation. Only after the cast animation is under way does the spell go off.

By changing ConjTime to 0, the wind-up no longer plays. The spell casts (almost) instantly (the cast animation still seems to play before the spell goes off, and takes a fraction of a second). This is very similar to Quickened behavior.

So ConjTime controls how long before a spell gets cast. Additional tinkerings show that CastTime begins *after* ConjTime is done, and then immobilizes the caster for its duration.

If ConjTime is set above 6 seconds, the Action indicator 'sticks' until the cast is ready to pop.


I'm checking this out because sooner or later am going to change my Spells.2da to rule: 1 second casttime per spellLevel.

/cheers

Hi KevL,

I do have a interest in your results, so do please keep posting them.

How do you intend to handle 7-9 level spells if (1 sec per spell level) the action "sticks" ... i.e. Do they still work after said times?

Thanks, Lance.

[kevL]

hey Lance,

that's all i intend for now. The 'stick' was only graphical, and the spell went off as the action-icon graphic went unstuck and back to normal operation. I used a ConjTime of 10 seconds. Pleased with the results, i'll likely make the changes to Spells.2da before my next play-session


ps. Quickness is fine. Or at least fine enough -- there's a short delay, that wasn't in the spellscript.

[I_Raps]

Perhaps it's just mulishness, but I'm afraid I'm not convinced on Quicken.  If my speculation in #11 is correct (or the opposite case where Quicken drops the CONJ TIME, or any variation thereof), then my experiment is useless.  Its design is fundamentally flawed.  It goes back to this line:  "To make it work, they should drop the CAST TIME from the next spell as well."

 

Casting consecutive Quickened spells, the next spell does have its CAST TIME dropped or whatever - due to its OWN Quickening though.  

 

In the immortal words of the philosopher - Ack.  Thptpth!!

[kevL]

you're like me _Raps : DO THE TEST.

I had Stangal bring up his alter, Stangal, via Universal Companions. Two 15th levels, 4 light spells apiece, all quickened on one of the wizards. The timing ratio over the course of all four Lights, paused to start, was a near exact 2:1


/but, yeh

[kevL]

Hi KevL,

I do have a interest in your results, so do please keep posting them.

How do you intend to handle 7-9 level spells if (1 sec per spell level) the action "sticks" ... i.e. Do they still work after said times?

Thanks, Lance.

hey Lance,
I have this implemented now. Only one fight so far, against a Witch and some Warlocks, levels 5+

Plenty of spells and metamagic-blasts flying around and no noted borks! I expect it to work fine but the curmudgeon is ofc in the details:
- imported Spells.2da into OpenOffice Calc spreadsheet
- made a function to automagically change the ConjTime column to 1000 times the innate spell level
- used NwN_2daEditor to line all the columns up, then did a compare with WinMerge and went down the list ...
- realized I don't know enough about all the spells and spellabilities to do it right the first time.

I intend to keep the spellabilities ( feats ) faster than spellcasts (at least if their rows are different) -- at most 2 or 3 seconds, and generally only a half to 1 second, if not instantaneous.


/note: This definitely requires a fair bit of thought, research, and playtesting to get it right.

[Lance Botelle]

hey Lance,
I have this implemented now. Only one fight so far, against a Witch and some Warlocks, levels 5+

Plenty of spells and metamagic-blasts flying around and no noted borks! I expect it to work fine but the curmudgeon is ofc in the details:
- imported Spells.2da into OpenOffice Calc spreadsheet
- made a function to automagically change the ConjTime column to 1000 times the innate spell level
- used NwN_2daEditor to line all the columns up, then did a compare with WinMerge and went down the list ...
- realized I don't know enough about all the spells and spellabilities to do it right the first time.

I intend to keep the spellabilities ( feats ) faster than spellcasts (at least if their rows are different) -- at most 2 or 3 seconds, and generally only a half to 1 second, if not instantaneous.


/note: This definitely requires a fair bit of thought, research, and playtesting to get it right.

Hi KevL,

"A lot of thought" .... Does not surprise me.

Like you, I have also added some new elements to some of the spell casting stuff. Mine is different, but I have designed what I call "Arcaene Lore Scrolls" (ALS) .... E.g. Magic Missile ALS ...The PC can use it as many times as they have a spell of the ALS available to them. So, if the PC has learned three 1st level spells, then they can cast the Magic Missile ALS three times, and the "power" comes from one of their other learned spells. It's a bit like giving magic users a kind of spontaneous casting ability through using these scrolls.

Lance.

[kevL]

I think it was Kaedrin who introduced a similar feature: Reserve Spells Feats.

it's a bit simpler in that, It checks if PC has say a 'cold' spell memorized & gtg (these are defined explicitly as spells SPELL_X, SPELL_Y, SPELL_Z, etc.), and if so the caster can cast use Feat SPELLABILITY_A (the so-called Reserve Spell FEAT) as many times as desired.


i like the scroll idea

Modifié par kevL, 10 avril 2014 - 09:09 .

[Lance Botelle]

I think it was Kaedrin who introduced a similar feature: Reserve Spells.

it's a bit simpler in that, It checks if PC has say a 'cold' spell memorized & gtg (these are defined explicitly as spells SPELL_X, SPELL_Y, SPELL_Z, etc.), and if so the caster can cast SPELL_A (the so-called Reserve Spell) as many times as desired.


i like the scroll idea

Hi KevL,

I had not heard about Kaedrin's idea.

I just did a search on my blog and found my first post related to 6th July 2007! Nearly seven years ago! Can you believe that! Here is what I said then (http://worldofalthea...-spell-use.html) :-

NEW SPELL USE (ARCAENE LORE)

Along with the new time-scale, I propose to introduce a new spell use concept, called "Arcaene Lore". This will be in reference to the ways spells were cast prior to "The Darkening", which is what the event is called after the PCs return from the demi-plane. In this case, Arcaene Lore is rediscovered knowledge from the "old ways" of casting magik spells. In game terms, this simply means the PCs will discover scrolls and pages with knowledge of how to cast a spell like they did prior to The Darkening. This means being able to repeat cast the same spell without having to learn it first.

How does this work in game terms?

Basically, as the PCs explore the changed world, they may come across scrolls (or simple pages) of magik written in such a way that its powers have been embedded into the page itself. Then, a PC with the right ability can mix their own powers with that of the scroll and release the magik contained within its pages. The page is *not* expired with use, because the main power to unleash the magik within the page is transferred from the caster themselves (if they are strong enough to cast the magik in the first place).

E.g. To start off with, a wizard will gain new spells in the usual D&D 3.5 rules way. If they learn Magic Missile, then they will have to specify how many times a day they wish to have it memorized to use and be governed by their choice. However, if the same wizard then finds an Arcaene Lore scroll with Magic Missile inscribed upon it, then they will be able to use the scroll to focus the power of the spell as many times as they use the scroll. However, they are still limited by their own powers, which means with each use of the scroll, the number of first level spells they can cast in 8 hours is reduced by one. The only side effect of using a scroll in this way, is that the spell lost to give power to the scroll is not known until they check their remaining spells. (There may be a pattern as to which spells are lost first, but I leave this to the player to discover.) This new system is still fresh in my mind, but please not the following:-

1) As the PC acquires more of these Arcaene Lore scrolls, so they will be able to rely less on original spells learned and arrange their pages as they see best.
2) Arcaene Lore pages can be given to any PC that has the ability to cast them, which means spells can be more easily managed between the party.
3) Arcaene Lore is not copyable by PCs, as it is lore and magik from a time past. Only Vol judges when and what pages may be found.

0 ---- 0

All I can add to this, is that it was a pain to implement.

Lance.

[kevL]

that sure sounds neat. I guess a player has to get in there and dirty to see

 

And i'm sure you could do it with a lot less 'frustration' today, if it had to be done again.

 

 

me? Back to Fanglewood, and see if i can figure out why ACP (armor check penalty) is hitting Concentration instead of Set Trap ... /sigh *

 

 

* seems that once upon a time i felt ACP should affect Concentration, in Skills.2da lulz

[Lance Botelle]

that sure sounds neat. I guess a player has to get in there and dirty to see
 
And i'm sure you could do it with a lot less 'frustration' today, if it had to be done again.
 
 
me? Back to Fanglewood, and see if i can figure out why ACP (armor check penalty) is hitting Concentration instead of Set Trap ... /sigh *
 
 
* seems that once upon a time i felt ACP should affect Concentration, in Skills.2da lulz

Hi KevL,

I will have to take a look at that module myself. Catch you later!

Lance.

[kevL]

arr, Matey