Base Mod Content Thread: Creatures

Development of standard ALFA palettes (ABR)

Moderators: ALFA Administrators, Staff - Technical

User avatar
ç i p h é r
Retired
Posts: 2904
Joined: Fri Oct 21, 2005 4:12 pm
Location: US Central (GMT - 6)

Post by ç i p h é r »

Or a floating text string that says...."You Spotted an Orc"...so it doesn't get lost in the combat log.

I like the idea though. An alternative would be to use a targeting item. I find that the OnPerception scripts are not entirely reliable so a fool proof means of determining the monster would be good and especially nice for scouts. Maybe this could tie in with the "size up" feat or somehow offer more useful information with higher ranks in the relavent knowledge skill.
User avatar
Fionn
Ancient Red Dragon
Posts: 2942
Joined: Sun Jan 04, 2004 7:07 am
Location: Seattle, WA

Post by Fionn »

As long as you can only 'see' a mob just barely farther than its perception range (and sometimes not that), I'm not sure such tool would be useful unless said scout could litterally walk up to them. I certainly wouldn't be using anything that locks me into the convo window if I thought there was danger (and the fact that the convo window kills stealth last I checked), so it would just be floaty text in any event.

Now, if said tool allowed you to click terrain you couldn't 'see' mobs on and return info of any mob in a Large SHAPE_SPHERE, that would be usable. This would also allow FB from range, so we'd need to make sure our AI could handle this tactic.
PC: Bot (WD)

Code: Select all

     -----          -----          -----          -----
    /     \        /     \        /     \        /     \
   /  RIP  \      /  RIP  \      /  RIP  \      /  RIP  \      /
   |       |      |       |      |       |      |       |      |
  *| *  *  |*    *| *  *  |*    *| *  *  |*    *| *  *  |*    *|
_)/\\_//(/|_)(__)/\\_//(/|_)(__)/\\_//(/|_)(__)/\\_//(/|_)(__)/\\_(
Thangorn
Haste Bear
Posts: 2081
Joined: Fri Oct 01, 2004 1:00 pm
Location: Queenstown, New Zealand

Post by Thangorn »

I'm working on a scouting static that works with the random encounter system in NC that uses an item that "counts" critters, once you've "counted" three you can run back and report to the Purple Dragons.. sounds really similar to this item...

so you use it to click on a critter and it reports back information about it after a roll? could easily be quick slotted and use sendmessagetoPC to report information, I'd hate to have a convo on it too, having to exit from the convo while copping a beating would be nassty.

heh.. after using it, the next thing the PC using it says could be tracked and pasted as the name of the critter.. imagine an orc running around with "One ugly bastard" on its head.. anyway.. I think its a nice idea but I'm just getting silly now :)
On indefinite real life hiatus

[22:52] <Veilan> obviously something sinister must be afoot if a DM does not have his social security number in his avatar name!
Dorn
Haste Bear
Posts: 2196
Joined: Sat Apr 17, 2004 5:00 pm
Location: Australia (West - GMT+8)

Post by Dorn »

It would be good to see in the chooser some 'standard' classes at different level. Ie -

Priest level 2
Priest Level 5
Priest Level 9
Mage level 2
Mage Level 5
Mage Level 9
Warrior Level 2
Warrior Level 5
Warrior Level 9
Thief Level 2
Thief Level 5
Thief Level 9

All neutral and all Faction Commoner so able to be changed/equipped by the DM.

Would be useful for impromtu DM sessions when you cant wait to update the mod.
playing Nathaniel Ward - Paladin of the Morninglord and devout of Torm (cookie cutter and proud of it)
Thangorn
Haste Bear
Posts: 2081
Joined: Fri Oct 01, 2004 1:00 pm
Location: Queenstown, New Zealand

Post by Thangorn »

Just had another idea..

why dont we build in a functionality into the AI scripts that allows builders to set behaviour by just attaching a number of variables to a creature on creation?

Its pretty straight forward and all the functionality of AJAI could just be accessed through local variables on the critter..

That way we dont need reams and reams of scripts in the pallette for AI.

we can have good AI that we default to if no variables get entered on the critter.

Is this possible mad scriptz0rs?
On indefinite real life hiatus

[22:52] <Veilan> obviously something sinister must be afoot if a DM does not have his social security number in his avatar name!
Ronan
Dungeon Master
Posts: 4611
Joined: Sun Feb 20, 2005 9:48 am

Post by Ronan »

Yes Thang, though such variables are usually created on the creature by ajai's on spawn script. We could move them to the creature's blueprint itself, of course. However, until we know what sort of AI NWN2 ships with, I'm not sure how much planning we can do. AJAI is resource-hungry and really isn't very good IMO.
User avatar
indio
Ancient Red Dragon
Posts: 2810
Joined: Sat Jan 03, 2004 10:40 am

Post by indio »

Agreed. AJAI is something we can hopefully do without.

Surely NWN2 will have improved AI. We can't count on it, but it's a long way from an unreasonable expectation.
Image
User avatar
Fionn
Ancient Red Dragon
Posts: 2942
Joined: Sun Jan 04, 2004 7:07 am
Location: Seattle, WA

Post by Fionn »

AI is usually a drain on CPU. New specs need better CPU, so hopefully...

Dorn - that many CR choices for *each* race/breed would kill my toolset. We should have them availalbe for integration as an ERF (or several ERFs), but not ever server will have Kua Toa.
PC: Bot (WD)

Code: Select all

     -----          -----          -----          -----
    /     \        /     \        /     \        /     \
   /  RIP  \      /  RIP  \      /  RIP  \      /  RIP  \      /
   |       |      |       |      |       |      |       |      |
  *| *  *  |*    *| *  *  |*    *| *  *  |*    *| *  *  |*    *|
_)/\\_//(/|_)(__)/\\_//(/|_)(__)/\\_//(/|_)(__)/\\_//(/|_)(__)/\\_(
User avatar
ç i p h é r
Retired
Posts: 2904
Joined: Fri Oct 21, 2005 4:12 pm
Location: US Central (GMT - 6)

Post by ç i p h é r »

I wouldn't bet on NWN2 shipping with AI that'll exceed what we have now, unless there is a dramatic change in the API which would yield significant performance improvements. Nevertheless, it will certainly be interesting to evaluate what they produce, but I'd wager we'd get a whole lot more of what we wanted if we spent the next 6 months improving the AI we have now.

And Fionn is exactly right. Any *good* AI will always be resource hungry, which is why it is often designed to be scalable to suit the supporting hardware. Starting over won't necessarily ensure a better result vs spending an equal amount of time improving the existing system, especially if it's well written. From reading some of the comments the developers have made in interviews and chats with the community, they clearly expect NWN2 to perform a LOT better than NWN1 simply because it's designed with the intention of running much more advanced hardware. With GPU's becoming standard and bringing with them a lot of horse power, and with multi threading support in NWN2, the performance ceiling should be much much higher than it was 5 years ago.

Having said that, I don't consider AI to be a priority for us as we'll obviously have something that will be adequate, be it JAI or NWN2 stock AI.
User avatar
indio
Ancient Red Dragon
Posts: 2810
Joined: Sat Jan 03, 2004 10:40 am

Post by indio »

I love your reasoning, cipher. Very refreshing.
Image
User avatar
Overfilled Cup
Orc Champion
Posts: 437
Joined: Tue Apr 05, 2005 6:45 am

Post by Overfilled Cup »

How about this. can we currently say erase the drown effect from water elementals..

1.then add a package feat/attack that say takes gust of wind for KD using a cone of cold vfx/adds dmg say cold for water dmg requiring x save roll to be "released" from this effect. Would the vfx then be turned off? Can we get more than one pc trapped in this effect?

2.can we mimic a drench spell..is there scripting available that would mimic this ability. The script would need to extinguish all nonmagical light within x distance of the water elemental. would fall under an elementals attack per round and would be kick ass in an underground setting.

Just trying to gauge what we can and cannot do and this question will come up 1000+ more times as we recreate each individual creature. My real question was in "packaging" groups of fx to make an attack.

And yes there is plenty of work left for both blueprinting and scripting abilities for creatures to be passed out if anyone wants to volunteer.
User avatar
Overfilled Cup
Orc Champion
Posts: 437
Joined: Tue Apr 05, 2005 6:45 am

Post by Overfilled Cup »

Yes I may try to organize into several erfs after we do the basic creatures. Will take some thought to think of the best way to organize that but say for sake of argument.

MM1
MM11
Libris Mordis

OR

If we have a solid ftp no reason we cant have several erfs that way mod owners can import the creatures they want but that the creatures themselves will be universal throughout ALFA. *shrugs*
Thangorn
Haste Bear
Posts: 2081
Joined: Fri Oct 01, 2004 1:00 pm
Location: Queenstown, New Zealand

Post by Thangorn »

cipher wrote: I don't consider AI to be a priority for us
Could you perhaps expand a little on your reasoning for not wanting AI as a priority for the base mod?

How do we make decisions about what we want base mod critters to be able to do if we dont have AI on our base mod christmas list?

anyways, I personally dont care what AI we use, but I'd prefer to eventually code something that can work off variables stored on creature templates. Now that be nice to do later rather than sooner, but who wants to go back and trick up the base mod creature AI /again/ after you've already built 200+ creatures?

I'd prefer to get our basemod AI reasonably close to right, first time and in the context of making critters for teh base mod, rather than in a few months go back and have to completely redo it all again, or worse still having all different wierd, wonderful and diverse AI systems on every mod. We need at least the basics specced out I'd reckon...

I think there are a lot of AI issues to consider in terms of creating creatures and spawns for the base mod so I am happy for brainstorming and discussion on it now, rather than waiting until the last minute to know what people want /and/ trying to trick out the system.

And for the record.. I think it would be 10x easier to code new AI from scratch adding on to the NWN2 packaged AI, rather than go through every piece of code and try and fix AJAI in NWN2. The testing on that would be a huge pain in the @55. I'm no programmer and hell, tell me if I'm wrongheaded here, but trying to fix scripts seems to me to always be worse then starting from scratch IMO.

*shrugs*
On indefinite real life hiatus

[22:52] <Veilan> obviously something sinister must be afoot if a DM does not have his social security number in his avatar name!
User avatar
Overfilled Cup
Orc Champion
Posts: 437
Joined: Tue Apr 05, 2005 6:45 am

Post by Overfilled Cup »

Anyone tried bribing Ronan yet?
User avatar
ç i p h é r
Retired
Posts: 2904
Joined: Fri Oct 21, 2005 4:12 pm
Location: US Central (GMT - 6)

Post by ç i p h é r »

Thangorn wrote:Could you perhaps expand a little on your reasoning for not wanting AI as a priority for the base mod?
Sure. The required features for the base mod include the ACR, mob AI, and mob spawning systems. Developing the ACR to suit our needs is of higher priority than the work involving AI. This is because game rules are fundamental to making ALFA the Hard Core environment we promote it to be and there is sufficient evidence to suggest the NWN sequel will also be designed with the general gamer in mind. So, this is not work we can expect to be done by Obsidian.

I did not imply, or mean to imply, that I did not think AI was important. You know that I have certain ambitions regarding AI from our conversations after all. In the event we need to choose what's done first, I simply think AI has to take a back seat to creating our hard core rules. If we have enough people developing these features, this is essentially a NON issue. We'll be able to work on all systems from the word go.
Thangorn wrote:How do we make decisions about what we want base mod critters to be able to do if we dont have AI on our base mod christmas list?
Ah, now I understand your worry. Rest assured Thangorn, we will have AI in the base mod, whether it's JAI, NWN2 stock AI, or something else. This is an entirely separate issue from priority.
Thangorn wrote:I think there are a lot of AI issues to consider in terms of creating creatures and spawns for the base mod so I am happy for brainstorming and discussion on it now, rather than waiting until the last minute to know what people want /and/ trying to trick out the system.
Absolutely. And that's part of planning...and why we are going about defining feature scope. I haven't posted a thread for this yet (I think it would simply be overwhelming if I created all the threads all at once), but it's on the list. I promise. :)
Thangorn wrote:I think it would be 10x easier to code new AI from scratch adding on to the NWN2 packaged AI, rather than go through every piece of code and try and fix AJAI in NWN2. The testing on that would be a huge pain in the @55. I'm no programmer and hell, tell me if I'm wrongheaded here, but trying to fix scripts seems to me to always be worse then starting from scratch IMO.
That depends if you believe it needs fixing. I'm not in a position to critique AJAI as my experience using it is limited, but I have examined some parts of the code and it seems to me to be well designed with extensibility in mind, well written, and well commented so developers can understand the logic.

I think there are two questions we need to ask ourselves in determining which is the right course of action, rewrite vs extend:

1. Is the current AJAI architecture insufficient to do what it is we want our AI to do?

2. Is the current AJAI code heavily dependent on an older API that retrofitting it would result in gutting major parts of it?

If the answer is broadly YES to either question, then a rewrite would be the best approach. However, if we believe the architecture is sound and is based on current API, I think we'd get far more mileage out of optimizing, customizing, and extending what's already there. In the same amount of time it would take to rewrite the AI, we could have added much more capabilities to the existing system. One of the most common perils of rewrites is that you almost always forget something the old version used to do. And seeing as none of us actually wrote the AI, we're likely to repeat the same mistakes Jasper made in his inception of JAI. If we have an AI expert around who writes this sort of thing for a living, I'm happy to defer to their expert opinion.
Locked