Neverwinter Nights Forum News

Remove all ads!

Here are today's Neverwinter Nights forum highlights, collected by NWVault. Please take into account that these are only single parts of various threads and should not be taken out of context. Bear in mind also that the posts presented here are copied as-is, and that any bad spelling and grammar does not get corrected on our end.

Rob Bartel, Senior Designer Live Team

What's the latest word on the WW series?

---

Q: Is it recommended to go ahead and play WW, or wait until the remastered version is completed?

---

I recommend waiting for the remastered version. Not only does it have some new features (like subraces, for instance), but it's also going to be the official version from here on in. The remastered version is saving out a bunch of database variables that will be taken into account as the series develops. The original version doesn't save these out so, when WW2 rolls around, those who played only the non-remastered version will be treated as if they haven't played WW1 at all. They'll still be allowed to proceed with WW2 but will be given only the default values with nothing special added in.

Georg Zoeller, Designer

1.64 Beta 2 available.

---

Well how else can i explain, When i hit someone with the blind fight feat theres no 2nd roll when the target is concealed. Also I hear about people duping all the time but I cant do it, so its upto the developers to do some research that I as a paying customer pay them to do.

---

We don't invest time to investigate vague rumours, sorry, that's not a good use of our time. The fact that you bought the game doesn't change that, we are not in the 'patch on demand' business. For future reference - I'll delete any vague "I heard there is a bug ..." and "My brother's uncle told my father who told me that you can do this..." kind of bugs from this thread. About blindfight - the roll happens in the engine, it's just not exposed.

NPC Pathfinding with PCs no in the Area

---

If you are having CPU load issues, you can gain a lot of relief by effectively disabling pathfinding for NPC's using WalkWaypoints. If you replace the move-to-object commands with move-to-location commands in x0_i0_walkway, no pathfinding will be invoked.

---

Nearly correct.

"Pathfinding" is the process of finding a valid path from point A to B using a snapshot of the current situation in the area. MoveToLocation performs pathfinding. However, the problem people often associate with pathfinding is that creatures don't react well to dynamic obstacles such as other creatures, which has nothing to do with "pathfinding", it's movement actions. When doing MoveToLocation creatures finds a path. Pathfinding is done at this point. It then starts to move along the found path, which has nothing to do with pathfinding anymore, it is simply walking along nodes on a predefined route. Since other creatures also move, they might block the calculated path which will cause the MoveAction to stop. While pathfinding itself can take static objects or not moving creatures into account (by marking blocked cells as not passable and bumping the cost of nodes close to i.e. dynamic objects), it can not take into account dynamic object movement. Example:

- NPC "Fred" wants to move from "A" to "C"
- Pathfinding runs and finds a valid route.
- Movement starts based on the pathfinding data.
- Player enters the area and moves to point "B" which happens to be right in the calculated path between "A" and "B".
- Fred hits the player and can't pass him. Movement stops.

Because the player can not be predicted at the point when pathfinding starts, the calculated path suddenly became invalid. Same goes for NPC movements, on the fly created placeables, etc. It would theoretically be possible to predict and take into account other programmatic object movements schedule by the engine, but this would be a wildly expensive undertaking and still frequently broken by unpredictable things such as player movement, monster's being summoned, etc.

So MoveToObject has an internal "retry" counter. If the move action ends with as ACTION_FAILED instead of ACTION_COMPLETE (because the path was blocked), MoveToObject will start over (pathfind -> move). I think the retry counter is 6 times, not sure about this.
This causes the illusion of creatures following other creatures, reacting to obstacles, etc. ActionForceMoveToObject is the most aggressive of these functions as it will continuously restart the "Pathfind->Move" cycle until the user specified timer expires. This allows you to judge the cost of these operations:

Pathfinding on an empty tileset is fast and painless as all node information needed to calculate a valid path has been embedded in the tiles.

Once you add lots of placeables, movement requests between area transitions, lots of creatures, possbibly walking random or ambient, etc, you bump the cost for performing a single "find path from a to b" call significantly.

Depending on the size of the area, the distance between NPC and target point and possible area transitions in between, the number of nodes and possible routes that gets evaluated can become quite extreme.

Now think about 10 creatures in an area all doing ActionMoveToObject or even worse ActionForceMoveToObject with a high timeout - If the waypoints are far away from each other, the routes collide with each other and there is a fair amount of player movement, this is going to cause A LOT of the cpu time available to object AI to go into pathfinding operations.

So some thing you can do:
- don't do ActionMoveTo* at all while no PC is in the area, instead teleport. Nobody is going to see it anyway!

- use ActionMoveToLocation for movement that you don't care about if it works (i.e. ambient stuff)

- place waypoints in a way there is a clear and open path between them. Avoid placeables.

- don't make different creature path run into each other.

- avoid walkwaypoints on main player trafficways.

- avoid bottleneck spots like doors.

- use ActionForceMoveTo with a low timeout.

Of course it all depends what you are doing, with a smaller module you might get away with a lot of stuff until something breaks or you get AI time overflows (which means that objects with lower AI priority will skip stuff they are supposed to do). On a huge persistent world where a lot of areas can be active at once (worst case scenario: 100 areas, 50 players, 1 player per area = 50 areas with objects running on non reduced AI), this kind of thing is going to seriously bite you. hope that helps as bit.

Auto-DL Hak's DOA?

---

haha, i was thinking of posting that same thing. I think DD killed hak-auto-dl, because if i could hak-auto-dl for free as part of the app, why would i pay for custom content in DD? But hey im just speculating!

---

Ah, another one of those "let's try to provoke information" posts You could just have asked as well or read the Witch's Wake forum... Auto-DL hackpaks is not a high priority right now as it's quite an expensive thing to do with a lot of potential problems that would need to be addressed (such as security concerns, bandwith considerations, etc).

In a way you could say that DD has something to do with auto dl hakpaks not being worked on at this point because the Live Team is spending their hours working on DD, but I can assure you it has nothing to do with your speculated reason. DD will not feature "in game downloading" of pay modules.

The idea behind DD is not to prevent anyone from downloading free available content - it's the heart of NWN and will always be. We are not looking to replace free content and modules or make it hard for people to use those - that would not only be false, it would be completely stupid and against the core idea behind NWN. DD is about BioWare making high quality modules available to the community for a small fee and use the revenue to finance continued Live Team support for NWN, which could include creation of new features, content, etc.

We will talk more about DD when we are ready to do so, in the meantime you can visit our Witch's Wake Forum where the whole DD thing is being discussed for several month now.

 

Oh, that first post kind of admits WW2 is finally in development again. That is good to know.