Interior Area Tileset Creation Builders (Workaround Found)

[Lance_Botelle]

The Workaround:

After messing with the area for a long time, it appeared I had TWO issues that were affecting the usability highlight for the area in a MP game. Firstly, the area required rebuilding (something had obviously gone awry with the original), but more importantly, I also needed to remove every creature from the area for it to work with a player’s larger party as expected. Rebuilding the area allowed me to use more PCs in the party, but it was not until I removed all creatures from the area too, that then allowed the area to work with a party size similar to the OC areas.
…

EDIT (UPDATE): The problem can be more easily identified than this first post describes … See my next post in response to Kamal that shows the problem using the official debug code alone. Instructions to reproduce the problem are included there too.

This is related to the MP concerns I have that I mention in one of my other posts … I am hoping someone who has the skill at building interior area tilesets may be able to help me with this one… Perhaps @4760 or @rjshae or anyone else with such skills?

EDIT: @Kalister68 (Just found you. Hopefully, you are still around?)

Anyway, I am using a custom tileset (Kalister Catacombs) and it appears to be “limiting” some aspects of design when using scripts. Is this possible?

I found the download here. (Catacombs-1008-1-0.zip)

This is what appears to be happening: An area has only two secret door placeable objects within it (the same object so I could test it twice) that have a heartbeat and a used script attached. The idea is that a PC gets close to the object that the HB does a search check for them and makes the item useable to then open and pass through. In a single player game these objects work absolutely fine. In a MP game, however, if I approach with a party of nine, and the HB fires (feedback confirms), but the object, while useable as far as the script is concerned does not turn “useable” to the player’s PCs.

Then, if I dismiss one of the party members (any of them), the highlight to the object then updates and the object can be used as normal. (I could also save the game and reload and the object would become useable too.)

My initial approach was to simply reduce the party size prior the HB triggering, which I thought would be all that was required. However, that did not work, as it is as if the “requirement” to update the usability glow is “dynamically” changing to always require me to have one less member in the party to allow the object usability to trigger and work as expected.

EDIT: Also, I just noticed that the area will also NOT remove the “debug” overlay when deactivated. i.e. Triggers still show in green, etc.

I have never seen anything like this before and I have spent the last few days trying to resolve this issue, but may have to resort to simply destroying all the objects related to this “door” in a MP game, as that does show to work.

I also tested the object itself (in an outdoor area) and that worked fine with a party of any size.

Anyway, if anyone has any answers, you could help save me going insane with this one. Thanks!

EDIT: I just tested on a OC interior area and the detection and make useable worked fine, even with ten in the party…

EDIT: It appears that there are other factors involved, such as area size and other resources in play, but the “broken” area as it stands alone appears far more “compromised” (uses more resources from the start) compared to the OC interior areas.

[kamal]

Need to see the script.

[Lance_Botelle]

Hi @kamal

Thanks for responding.

EDIT (TO ALL): This demonstrates the problem more easily, in that it uses just the official debug code (enable and disable at the console) to show the problem I have… See “How To Reprodue the Problem” below.

If you mean the HB script, I can confirm that that is not the issue, as it works fine in every other test, be it in a SP or MP game, if tested in an OC built area. i.e. The HB does enable the usability of the object in every instance, even the within the broken are tileset. This can be “proven” by a save and reload after detection.

The problem is, the highlights to show the object (and allow it to be clicked on) now refuses to update on the object within this tileset in certain circumstances. As another example, even the debug overlay for triggers will not update in this tileset when tested in a broken condition.

EXAMPLE: MP GAME WITH ONLY FOUR PCS

NB: Sometimes it may work with this few PCs, and if it does, then my useable object also works fine. However, if you up the number of PCs and/or associates, this area is far more likely to fail than the OC ones. For example, I used the same four PCs, and it worked the first time. I then summoned one more creature, so the party had six members (including the two associates) and the highlight then failed to show again until after I unsummoned an associate. (It can take a few seconds to update, but it eventually returns.) If I unsummon all associates, then it comes back quicker.

NOTE: This also affects the speed of creature drops/creations. i.e. They will NOT show until after all associates are reduced or completely removed.

SUMMON A FAMILIAR, THEN ACTIVATE DEBUG… NO DEBUG HIGHLIGHT SHOWS:

UNSUMMON FAMILIAR AND DEBUG RETURNS!

This is exactly the same problem as I am having with USABILITY highlights when made useable from a script.

HOW TO REPRODUCE THE PROBLEM

1. **Download tilset here. (Catacombs-1008-1-0.zip) (*) **
2. Build a small area with the tiles.
3. Paint a trigger in the area to show in debug mode.
4. Start a MP game for this area.
5. Make a party of four or more PCs.
6. Have one or two PCs create an associate or two.
7. Try entering debug mode to show the trigger.
8. Keep unsummoning associates until trigger (hopefully) appears.

(*) I also have the tileset in my hak available from The Scroll download page. It is the Althea_Other (15 JUNE 2023) file.

It is as if the game runs out of resources to update anything within the area, but also appears to vary subject to PCs available at the time.

FURTHER TESTING:

I have now tested with the OC Standard Interior tileset and the issue appears to be the same, but (like I report in my MP post), only when the party count is up to around 9 or 10. So, the “problem” I am encountering in this particular tileset is that it cannot handle very large parties at all, which needs to be carefully managed when it comes to what the player can interact with while within the area in question. As an example, this tileset even limits entry if the party size is larger than six.

So, my question to the tileset experts now is: Is there any way to alter the tileset to increase the size of the party within areas for a MP game? Model sizes? Texture sizes? Are these what impact the overall MP experience?

P.S. I don’t think exterior areas have this issue.

[kamal]

Other than the unusual walkmesh triangles in the two open floor tiles, dont see anything unusual in the catacombs tiles in your screenshot.

[Lance_Botelle]

@kamal

The problem is not “visual” from walking around the tileset… You would need to read my posts and try the tests I mention to duplicate/appreciate the issue.

My only theory is that this particular tileset is too resource heavy for a MP game. Maybe the models used are too complex? I don’t know, but maybe it is something a tileset expert would know?

Did you manage to reproduce the problem as I describe in my previous post to you?

[kamal]

@Lance_Botelle

I have now tested with the OC Standard Interior tileset and the issue appears to be the same, but (like I report in my MP post), only when the party count is up to around 9 or 10. So, the “problem” I am encountering in this particular tileset is that it cannot handle very large parties at all

These two seem to be contradictory, if the issue happens with the OC standard, as you say it does, then it’s not specific to the Catacombs tileset. I have looked into the standard interior models, here aren’t any parameters hiding in the hex code for anything other than what you would expect to define a tile, ie: the standard texture for it, the hooknodes for doors etc.

Catacombs does use more and higher resolution textures than the stock tilesets, and are probably higher poly count as well, but that should have no effect on scripting.

I do not have the game or toolset currently installed.

[Lance_Botelle]

@kamal

My ignorance of how tilesets work alongside scripts is limiting my questions and explaining of the issue as I am trying to express it, sorry.

My difficulty with this tileset is that it is more limiting than OC ones, to the point of even struggling to play an area in a MP game, even with just a party of six.

While the OC ones suffer in a similar manner, they at least allow around 9 in the party compared to approximately half this number in the catacombs tileset. That makes the tileset harder to work with, with scripted events due to objects in the area failing to update even with a small party.

I was hoping there may be someone with the knowledge and skills to alter/fix the catacombs tileset to allow a similar size party as the OC tilesets. Better still, if someone knew how to improve the efficiency of all tilesets for a MP game, but that is likely impossible.

However, if there is an easy way to alter poly counts, I’d be willing to experiment with that if someone with the skills could explain what I need to have and do to try this. In your knowledge, do you know what I need to do this?

I don’t understand why a tileset has this kind of impact, especially when all players are using the same resources. But, I will try to learn if need be.

EDIT: I just found a few links, so I’ll try to pick my way through them. I just hope it does not take too long to find what I need to know. I am going through these…

https://neverwintervault.org/blog/rjshae/2016/aug/building-tilesets-nwn2

I see MDBConfig may be able to help me… checking.

EDIT: I notice that some of the catacomb textures are considerably larger. So, I am hoping that simply resizing these may help.

Current differences:

I also tried comparing equivalent? tiles from OC and catacomb sets, and the catacomb appears to have quite a few more tabs?:

I also notice the “No Cast Shadows” in the OC versions.

EDIT: I managed to reduce the texture sizes and add “No Cast Shadows”, but the problem remains … I will try to learn how to reduce polygons, but that looks a lot more involved.

Sadly, I have no idea what I am doing now … I don’t even know where to look for the current poly count, let alone try to change it. @kamal is there a straight forward way for me to alter the poly count here? EDIT: Ooh, found some videos … watching now.

EDIT: Well, I managed to select the tile in Blender, used MESH to “decimate the geometry” to half its current, went to export and the export failed.

So, now I don’t know where to go unless anyone can help me further?

EDIT: OK, I found the error log:-

Importing RIGD: Cube
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 5
Polygons: 4
ERROR: Polygon is not a triangle.
ERROR: Polygon is not a triangle.
Importing RIGD: TL_CA_WWWX_01_R
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 61
Polygons: 63
Importing RIGD: TL_CA_WWWX_01
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 30
Polygons: 14
Importing COL2|COL3: TL_CA_WWWX_01_C3
Vertices: 12
Polygons: 4
Importing WALK: TL_CA_WWWX_01_W
Importing RIGD: TL_CA_WWWX_01_RF
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 41
Polygons: 64
Importing RIGD: TL_CA_WWWX_01_02
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 88
Polygons: 98
ERROR: MDB not generated due to errors found during the conversion.

EDIT: Does anyone know if any of these options allow me to decrease the polygon count without loosing “triangles”?

EDIT: Even when I just IMPORT/EXPORT the same tile from the set (doing no action to it), I still get the error message …

Importing RIGD: Cube
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 8
Polygons: 6
ERROR: Polygon is not a triangle.
ERROR: Polygon is not a triangle.
ERROR: Polygon is not a triangle.
ERROR: Polygon is not a triangle.
ERROR: Polygon is not a triangle.
ERROR: Polygon is not a triangle.
Importing RIGD: TL_CA_WWWX_01_R
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 101
Polygons: 126
Importing RIGD: TL_CA_WWWX_01
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 44
Polygons: 27
Importing COL2|COL3: TL_CA_WWWX_01_C3
Vertices: 16
Polygons: 8
Importing WALK: TL_CA_WWWX_01_W
Importing RIGD: TL_CA_WWWX_01_RF
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 81
Polygons: 128
Importing RIGD: TL_CA_WWWX_01_02
Layers: 1
UV elements: 1 eByPolygonVertex eIndexToDirect
Normal elements: 1 eByPolygonVertex eDirect
Tangent elements: 1 eByPolygonVertex eDirect
Binormal elements: 1 eByPolygonVertex eDirect
Vertices: 153
Polygons: 196
ERROR: MDB not generated due to errors found during the conversion.

EDIT: I finally managed to export a model that I have edited, but something obviously went wrong …

BEST RESULT SO FAR!

After restoring the mdb files and remaking the area without any creatures in it at all, I can now get eight creatures in the party to work with objects. Yey! I’ll test with the restored textures, and see if the new area (without any creatures) works as well as I need. UPDATE: So, restoring the textures did nothing detrimental to the area. As a result, I am just not going to have any creatures here until required, I guess… I am none the wiser after all this.

That’s it for me with this issue … unless anyone knows exactly why this is happening and they have a better solution.

Thanks @kamal !

[kamal]

@Lance_Botelle

I also tried comparing equivalent? tiles from OC and catacomb sets, and the catacomb appears to have quite a few more tabs?

The catacombs has more tabs as the tile it made up of more pieces than a stock tile is. Unrelated to your issue but this is why the tileset doesn’t like being retextured via the tile properties.

MDBCOnfig shows you all the properties of each piece of the model. As you can see there aren’t any properties that would be relevant to party sizes.

You’ve got the tab for the _w open on the catacombs tile, which is the walkmesh part of the tile model, this is why that tab reports having a very low number of vertices and faces (walkmeshes are generally quite simple). No cast shadows is probably irrelevant to anything for a walkmesh, though I’m surprised there are texture maps assigned to the walkmesh, as they wouldn’t be visible.

“I just noticed that the area will also NOT remove the “debug” overlay when deactivated. i.e. Triggers still show in green, etc.”

If the debug overlay trigger not removing bothers you enough, One thing you could check is if the _w of the stock tile has textures assigned to it. If it does not, you could try using MDBConfig to remove the textures assigned to the _w of the catacombs tiles (at least the tiles where your triggers are, in order to test). I can’t imagine that would possibly affect the script issue you report though.

[Lance_Botelle]

@kamal

In this instance, the inability to have the debug overlay show simply coincides with the “usability” overlay not showing when a placeable object is made useable via scripting. It is the inability of the game to make an object “useable” (*) that concerns me. IE The failed debug overlay appears to correlate with the problem I have with the gameplay. So, it’s not the debug overlay, as such, that concerns me… But I know if that does not show, then I also know a placeable object is going to fail when a script makes it usable. It is merely a good test I can use to show a more concerning issue.

(*) As I say in a previous post, the item does become “useable” from a scripting perspective, but not from a practical purpose. i.e. A player cannot click on this object yet.

It’s a problem I had not expected to see when working on an area. As I say, I have managed to script in a workaround for a MP game now, but am still bemused as to what the problem actually is.

If you ever manage to have access to NWN2 installed and replicate my issue, you’ll see what I mean. It is very frustrating.

Thanks for the additional info, I appreciate all the advice on this issue.

As I say though, since removing every creature from the area, the usability overlay and player access has returned to normal functionality.

The Workaround:

After messing with the area for a long time, it appeared I had TWO issues that were affecting the usability highlight for the area in a MP game. Firstly, the area required rebuilding (something had obviously gone awry with the original), but more importantly, I also needed to remove every creature from the area for it to work with a player’s larger party as expected. Rebuilding the area allowed me to use more PCs in the party, but it was not until I removed all creatures from the area too, that then allowed the area to work with a party size similar to the OC areas.