The Black Scourge of Candle Cove -- Tchos' development diary

[RichardUther]

Man, I haven't read your last update, but I just watched the OM video and there are not enough 'bad words' in the English language to add to the amount of awesome I'm thinking right now. Just brilliant, brilliant stuff.

 

 

As for the lighthouse model. Seeing your mod in action gave me a better idea of how the whole thing looks, so I think I'll go from scratch, rather than using Chaos' one as a base model. If you want something in particular (like for instance, having a space within the top for you to place a SFX for the beacon) let me know. I can also get two versions of the tower so you don't need to have a high poly count model in the OM.

 

I repeat, this is something I would do for my own project anyways, and fortunatelly I have access to some pretty cool tools (at least during weekdays) so I can do a classic version of whatever and then add the steampunk weirdness for my own thing.

 

Of course, if you asked me to do something completely unrelated like a Pikachu creature, I'd probably send you a private message insulting you, your family and your grandchildrend too, just in case my outrage isn't clear. But a lighthouse? I'm game for that!.

 

Just let me know if you need something specific, that way I can keep the idea in my head in order to make it useful for the community in general rather than just for myself

[Tchos]

Glad you liked it. :) It's nice to be able to show something that doesn't require much explanation about how this and that are placeholders (aside from the lighthouse). It'll be nice to start putting my NPCs in the town proper, as well, because I really like the way the town came out, but I haven't shown any of it yet because there's not much point without the NPCs populating it.

 

As for the lighthouse, there is something specific that it would be nice to have, but I'll send you a PM about that, because it involves a little spoiler that I don't want to give away here.

[Tchos]

It occurred to me today how I might get the collision meshes to work properly without the troubles from yesterday, and that was to plop a huge tower over the island and use its collision to cut out of the walkmesh helper what the walkmesh cutter was having trouble with. That worked. Now the overwater map has a fully functional mesh that'll let you go around the island and near the pirate ship, but not let you go through them. While I was at it I also improved the sound effect selection and re-added the description to the town dock that must have been lost in one of yesterday's crashes.

 

Today I worked on putting in place the interactions with the Merchants' League. The player has three options for starting the main quest, and they all offer different rewards. The paths are mutually exclusive, so the player will only get one of the rewards. Each path results in somewhat different confrontations with the villains. That means another complicated conversation. I've never written conversations with this many branches and variations. I hope I don't miss anything important.

 

At this point, I had to do a little research, since it's time to assign XP and gold values for quest rewards. I know it's important to parcel out XP and treasure rewards according to a standardised system, for players who prefer to play series of modules with the same characters, with reasonable advancement between them. XP for fighting the enemies is taken care of for me, of course, but I have non-combat quests to reward, as well.

 

So, I ended up consulting the 3.5 Dungeon Master's Guide for guidance in that regard. I found what I needed near the end of Chapter 2, in the "rewards" section under "Mission goals". Although it says that such rewards are necessarily ad hoc, the general rule of thumb for major quests in a story is that they should be large enough to seem significant compared to the standard awards for the intervening encounters. That seems straightforward enough, assuming that killing standard monsters in NWN2 awards XP according to the CR tables in the DMG.

 

Since this is for a level 10 party, with an assumed party size of 5 members, I think the minimum reward for accomplishing the first of two major tasks should be 600 XP, since that would be how much a single level 10 player would get for a single CR10 encounter (and XP is duplicated for each party member in NWN2 instead of divided up). With that as the baseline, alternate approaches to that main quest will net somewhat larger rewards, all in addition to the standard rewards for combat and such. If that's not reasonable, then someone please let me know.

 

While working on this questline, I noticed that my Dropbox was reaching its capacity alarmingly fast. I checked my revision manager, and found that for two of the directories (the campaign folder and the module folder) I had forgotten to set a limit on how many old revisions to save. So I set both to a maximum of 10 revisions, and that brought my space usage down to 35%.

 

I think the next task after finishing this quest will be to create a full-sized version of the lighthouse island I made on the overland map.

[Tchos]

I got the complicated Merchants' League dialogue nearly finished. I completed the main branch and the after-quest line, and put in placeholders for the quest finisher branch and the after-initial visit branch. I ended up using a global variable to check if the quest is done in this case, since I have two different outcomes with this NPC, which has to coordinate with the state of the mayor's quest, and I thought it would be easier this way than doing an if-and-or check with the journal entries, which I'm not sure how to do with conversation conditions. (Checking if A and B, or B and C are true).

 

I fully wrote out the dialogue for the main section, with all its branches, conditionals, and skill checks. I've learned one lesson -- I should go with the tiered method of dialogue branches for future quests. I still need to create the quest items for this quest. There are 5 quest items associated with the Merchants' League version of the main quest line.

 

It was starting to wear on me, so I switched to the other priority task -- creating the lighthouse island, "Ol' Sea Dog's Rock". Making the small-scale one in the overland map didn't take too long, since it's only viewed from one angle, and at a distance. For the full-scale version, I had an extra handicap beyond the closeup scrutiny, and that was the fact that it would have to look like a big version of the little version I made. So I put a screenshot of the overland map version up on the screen, and started by creating a 24x24 area.

 

This being the ocean, I can't use a smaller area and block off the horizon with placeables, unfortunately, and even with fog on, I always seem to be able to see where the land ends and the sky begins. I threw some sky rings on there, which hides the edge, though the open sea shouldn't have mountains. Fortunately, I called this place Candle Cove, implying that the bay is at least partially surrounded by mountains.

 

I decided to have the top of the island and the mainland be at elevation 0, so I sank everything around them, and picked a height for the water. I dropped a tower and an NPC on the island for scale reference.

 

After I was working on the landscaping for a while, a low disk space warning informed me that I had accidentally activated Fraps' video capturing. I had forgotten to disable the hotkey after capturing the overland map footage, and it was dutifully filling up my hard drive. What I had accidentally captured was a lot of the initial setup and sculpting of the land. I've watched some people broadcasting their artwork in a live stream as they draw it, and I thought this might be of a similar curiosity for some people, so I kept it, edited out some of the repetitive bits and bits where nothing was happening, and doubled its speed for easier viewing.

 

I got back to work on the island, and did my best to match the miniature reasonably well, and also pick textures that still worked at this new scale. I sure wish there were some tintable rocks available in the toolset that aren't standing stones with no bottoms, which seriously limits the ways I can use them on a slope. Being able to rotate in all three axes would help a lot, too.

 

After I painted the textures on, I used the texture swapper to try different combinations, and settled on a different set. I thought I was going to run out texture slots, since I see the tiles can have no more than 6 textures on them, but I actually ended up with one empty slot left over. I tried using areas of texture to make a weathered look to the place, and then added some grass, trees, and bushes.

 

The toolset crashed while I was experimenting with the water settings, and I lost some work. I had to re-add some of the trees, and the grass.

 

I added the pirate ship and the player's ship anchored near the island (apart from each other), and some rowboats by the island's little dock. One of them is the player's.

 

Next, I roughed out the mainland as a backdrop, and then baked it. Next I need to hook it up to the rest of the module.

 

Edit: I've gone ahead and put up the videos and screenshots of the island in a new blog post.

Edited June 20, 2012 by Tchos

[Tchos]

I added the scripts, conversations, and waypoints needed to connect the new lighthouse island to the overland map, and tested the place out in-game for the first time. The fog settings looked just like they did in the toolset, so that was good. The docks were actually walkable, which I didn't expect. But the solid ground wasn't. So now I had to figure that out.

 

http://3.bp.blogspot.com/-qWVhbLuXETc/T-JQv8NPh2I/AAAAAAAADUA/B-PEpxx7FTs/s320/sea%2Bdog%2Breturn%2Bto%2Bship.jpg

 

First I made a quick conversation to give a debug NPC that just lets me choose an area to which to teleport, so I didn't have to use the overland map every time.

 

I placed a walkmesh helper overlapping the dock and the ground, and after that I was able to walk off the dock onto the ground, but it stopped shortly after that, and a red texture was visible at certain angles from the walkmesh helper. Why it's not invisible, I have no idea. I guess you're supposed to delete it after you bake it, or something.

 

I suspected the slope of the hill was too steep, and that was stopping my walking progress, so I tried adjusting that. I figured out that the yellow outline is the one that indicates the walkable area, and not the nice, clean white outline that showed up after I used the walkmesh cutter. One thing that confuses me is that there are two different yellow outlines, depending on whether I have the "baked" button active or not. The non-baked one looks fine, but not the baked one.

 

I baked, adjusted, baked, etc., etc. I tried adding stairs and walkmesh helpers. I discovered one thing that was probably causing a lot of the trouble, and that was that some of those rock placeables were blocking off the walkmesh, so I changed those to environmental objects. The bit between the dock and land continued to be a problem, and I spent a very long time with it. I used stairs to help it out, but it remained more finicky than a spoiled cat. It's like the troubles I was having in the Dragon Age toolset with the shadow maps.

 

"This can be a lengthy process," it tells me. Yeah, but mainly because of how many times I had to retry it. Maybe I'm missing something fundamental.

 

http://2.bp.blogspot.com/-_S2v1A8KRqs/T-JQvi87jPI/AAAAAAAADT0/FOiyMaiYAZ4/s320/sea%2Bdog%2Bharridan.jpg

 

Finally, I got an acceptable walkmesh. Unbelievable. It still has an inexplicable gash in one part, but there's plenty of room to navigate around it, so it won't stop the player from getting past, at least. Not optimal, but I think it would be easier to make it worse at this point, rather than better.

 

http://1.bp.blogspot.com/-Sjd81xV-_Kc/T-JQwfnXqfI/AAAAAAAADUM/9vGMPlf61aM/s320/sea%2Bdog%2Bwalkmesh.jpg

 

I spent nearly the whole day on that task. It was almost a relief to go back to working on conversations.

 

Looking at it now, I don't think I need to actually test for as many conditions in this one as I thought I would. I really only need to check to see if either of the two alternate end entries in the journal for the merchants have been reached. The mayor's quest, which has only one resolution, can do a double-check instead of this one doing a triple-check.

 

I finished the fairly simple "remind me" branch and "I've reconsidered your offer" branch, and that left the "quest reward" branch.

 

Another probably unnecessary complication in addition to the checks from earlier in this conversation is that I decided to offer three possible rewards for following one of the two branches in the merchant quest, but you'll have to be a master at one skill to get one of those options, and a master of two skills (both related) to get both of the options. Everyone else will just get the default. This is part of my overall design goal of including bonus content for non-combat skills. Of course, since you can use any of your party members to do the talking with SoZ dialogue, you're five times more likely to be able to get this bonus than with non-party-chat.

 

Anyway, that means creating 3 "contract" quest items instead of 1. If I had known about custom tokens before, I might have wanted to approach this differently, and create only one contract with a custom token in its description to describe the different rewards, but apparently custom tokens don't work in item descriptions anyway. Besides, having 3 versions of the contract means I can check the inventory later to determine which reward to give, as an alternative to using the journal.

 

Before I do that, I'm going to try installing some more toolset plugins, like an icon viewer and an item creation wizard.

[Tchos]

One of the plugins I installed was the Sin plugin, which does a lot of things that I could already do with the Usability plugin, but which also includes an "item lab". I just hope it doesn't interfere with any Usability keys. The main reason I want to use a plugin instead of just doing it in the item properties window is that I find that properties window rather chaotic and hard to read, and a plugin like this puts just the relevant settings into a nice, neat window.

 

Another one I installed was Lazjen's Icon Viewer, but I couldn't find any version compiled for the final toolset version, and although it worked anyway with the toolset config file, its functionality was either already duplicated by another plugin (maybe Usability again), or was built into the toolset later in the icon dropdown picker.

 

Then I re-read the documentation on item creation, and set about making the quest items.

 

http://4.bp.blogspot.com/-9e5J_o_UGXk/T-OlRaTFu8I/AAAAAAAADUo/7Tqj5cysBCE/s320/item.jpg

 

Does anyone have an opinion regarding the plot cursed flag? Is it a good thing, a convenience, a necessary evil, or an unnecessary annoyance? I generally find it an annoyance as a player. It's mainly an annoyance when I just want to move them to a different character or put them in a bag of holding to unclutter my inventory, and it won't let me do that. It's also annoying on the occasions when they're not removed from the inventory when the quest is finished. Kaldor Silverwand's Context Menu Additions mod lets me get around that as a player, dropping all items, flagged or not, and allowing me to pick them up with a different character if necessary.

 

Should I trust the players to understand that if they drop, destroy, or sell a quest item, they won't be able to finish the quest?

 

I tried out Grinning Fool's Creature Creation Wizard to create the questgiver from the Merchants' League, but it crashed the toolset when I was scrolling through the appearances. When I restarted the toolset, I found that the quest item I had created had lost its name, description and icon choices. This is the third time I've created this item. I thought item properties are saved immediately as you edit them! I'm glad I saved the description externally last time...

 

I installed a couple of new heads for use on NPCs -- just the ones I intend to use immediately. I can add more if I need them. These are actually NPC heads, but using this package allows me to outfit them as I want, rather than being stuck with the NPC's appearance, clothing and all. I think Garius has a very good head, but his outfit is ridiculous, and unfortunately this pack doesn't seem to include his head, but there were several good ones to choose from for an older man. If only there was nearly as much variety for the non-human races!

 

http://1.bp.blogspot.com/-LaZbOpQFEX0/T-OkdBeXgOI/AAAAAAAADUc/ITrAdYFNQ9Q/s320/quest%2Boffer.jpg

 

My modding was cut short due to an unexpected guest to entertain.

Edited June 22, 2012 by Tchos

[Tchos]

Today I debugged the league conversation and created more quest items. I like items. They're great flavour.

 

My experience seems to confirm that items don't get saved unless you specifically save the entire module, whereas with other things like conversations and scripts, there are ways of saving just that part you were working on. I had several unexpected occurrences. It seemed that when I changed an item, it didn't update the item that I had placed in a container, because it still contained the old version of that item.

 

I looked around for a standalone item creator that I could use outside of the toolset to make it more convenient, but couldn't find any.

 

I imported some new icons for some of my items, and spent a while learning the icon format and getting the proper frame, and experimenting to see which formats and modes would work. The tutorial says it needs to be a 40x40 32-bit TGA with alpha channel transparency. TGAs are very large and inefficient, especially in 32-bit. I also don't like the tiny 40-pixel images, and like the idea of building in support for UI mods like the bigger inventory mod which increase the size of the inventory icons 200%.

 

I made 4 icon images with different settings to apply to duplicates of the same item to see which ones would work. One was a control image made according to specifications, another was a PNG, another was a 64x64 TGA, and another was an 8-bit TGA with no alpha channel. The only transparency I see in the official icons is in the frame corners to make them slightly rounded. Considering those icons are usually on a dark background, I don't think the alpha channel is really necessary.

 

When I tested the new icons, I found that PNG doesn't work, but 64x64 does (presumably any size is resized to fit the UI), and 8-bit also works. In fact I used 64-colour indexed TGAs in some cases, and it still worked.

 

I thought that leaving out the alpha channel would make the lack of transparent corners visible when I moved the icons around over light surfaces, but surprisingly enough, they were in fact rounded! Perhaps the alpha transparency for icons is actually handled by some aspect of the UI, and the icons don't need alpha channels at all! In any case, this brought the file size of my icons down to 3.74k, despite the larger dimensions, which is half the size of the vanilla icons.

 

So now I've added a nwn_icons 2da to my hak, and created all the items necessary for the main quest.

 

The past few days have been relatively unproductive. I'll have to get back on track.

[Tchos]

Another P&P day. I spent most of my working time on item icons. I imported Shaezyra's Bonus Blueprints, so there are a lot of them.

 

I also built a hearthstone/stone of recall/divine intervention type system into the module, with arbitrary limitations to prevent it from being used in the main quest, but allow it in any of the side quests. Works over land, but not over water. Pick one up at a temple, and you can use it to return to the temple. Maybe I'll make it so that it can also be set to return you to an inn or tavern of your choice.

[Tchos]

I couldn't find a list of what sounds are available to use for the PlaySound command, but I found the sounds themselves in the Data folder. I previewed some using VLC, and found some passable ones that way. I picked a sound for the hearthstone, and then looked for a visual effect.

 

Since I didn't have a list of visual effects either, I used LilacSoul's script generator to give me a choice of them from a list, but quickly found that it didn't work. Further research indicated that effects aren't done that way anymore, so I opened up the visualeffects.2da and looked through them for a .sef file that I could apply in the same way that I did the quest indicators.

 

I went ahead and scripted the hearthstones so that you could ask any innkeeper or head cleric to make their location your home to teleport to, since it was pretty simple. The conversations with those characters just includes a check to see if you have one, and if you don't, they'll offer you one, and if you do, then they'll offer to make their place your home location, which sets a local variable on the stone to determine which waypoint to teleport to. I tested it out a few times, and it seems to work fine now. Next I just need to program it to not work if you're in combat, and also to make it take a few seconds to cast.

 

http://1.bp.blogspot.com/-snYhLGftV4U/T-eSR-NMMWI/AAAAAAAADVU/-y3CsDhw3FI/s320/hearthstone.jpg

 

But on to other things first. Heed's Boardable Ships happened to be absolutely perfect for what I had in mind for the first of the climaxes. I don't know how I would have handled it if I hadn't stumbled back across that pack while I was reviewing resources a few days ago. So much good material here, but hard to keep in my head what's available.

 

All of the ships in the test module for Heed's Boardable Ships were missing, replaced with placeables from MotB. In the description for the resource, he said that he updated the 2DA to deal with the additional MotB placeables, but apparently he didn't update the test module. What concerns me here is that my understanding had been that a 2DA file in a hak attached to a module should override the base game's 2DA for that module. This doesn't appear to be the case.

 

I'm also starting to tackle the swimming trigger. Apparently I'm going to need to swap "appearances" for each race, swapping to the appropriate line in the 2da. I guess this would be a good time to try out that "case switch" style of scripting, with the appearance number as the integer for the case? I have to assume it won't swap their heads to some default as well. Then another trigger to swap the appearances back.

 

While adding the boardable ships to my own placeables.2DA, which I took from the SoZ data file to use as a base, I also took another look at the problematic walkmesh helpers from Zarathustra. After a bit of investigation, it seems that the problem was not in the 2DA after all, but that the placeables were for some reason set to the wrong appearances. I'm really getting an intensive course in how all of these things relate to each other! One walkmesh helper was set to Column {Bear (X1)}, another was set to Column {Runed (X1)}, and the other two weren't set to an appearance at all. I would guess that this implies that the 2DA for the resource had been updated, but not the ERF that I imported which assigned the appearance to the blueprint.

 

After changing them to the correct appearances, which were in the list properly, and enabling C2 collision display (there's no toggle to turn that on and off? I have to do it in the preferences each time?), the walkmesh helpers appeared as intended.

 

http://2.bp.blogspot.com/-Mcf-s-5jLSc/T-eO1hf2PXI/AAAAAAAADU4/2SofKVZ16Mw/s320/ramp.jpg

 

With this new understanding of how placeables are defined, I was then also able to create new blueprints for the Heed ships, which is a good thing, since there was no ERF included with that resource. It's unfortunate that we can't have walkable ships that move up and down in the water like we can in other games, but then again for a large ship like this, in calm waters, it shouldn't be moving very noticeably anyway.

 

http://2.bp.blogspot.com/-Jo_2QFOwpKs/T-eO16052hI/AAAAAAAADVE/nFFBh9RrgeY/s320/heed1.jpg

[Tchos]

Baking with Heed's Boardable Ships didn't give me a walkable ship. Instead, the whole thing was insubstantial.

 

I expanded the bay for the ship battle, so that the ships are in the middle, and anything outside of its viewing range is removed. I tried out the non-walkable brush for this one, just to eliminate the walkable area around the boats.

 

In testing the new area, I learned that capitalisation counts when it comes to jumping to a waypoint from a script. New standard is all lowercase for all tags.

 

And now the extremely long and frustrating fight to get a usable walkmesh again. It turns out I can't use the nice, clean walkmesh that the SoZ ship already has built in, because there's a wall along the sides preventing me from walking across on a boarding plank. I should have expected it, since from a development point of view, there's no reason to make a passable section there, even if visually there is a gap, since they never intended for players to walk across it. So, that meant I had to use walkmesh helpers to create the whole thing -- tilted floors, stairs, and obstructions. Added to that is the problem that if I put a proper mesh blocker around a spot that you shouldn't be able to walk through, it creates a big square of non-walkability, which cuts a huge gash across the walkmesh across an entire tile cutting off the boarding plank. Even if I just put a crate there! So I leave that little structure walkable. What a mess, what a mesh. It'll have to do.

 

http://4.bp.blogspot.com/-GTHrCaof_uk/T-jYJ_NWcOI/AAAAAAAADV8/tB0LJrz3DjM/s200/broken%2Bwalk%2Bmesh.jpg http://3.bp.blogspot.com/-Kp5HcOqmDFM/T-jYKRWr8QI/AAAAAAAADWI/NCGYSbRxueo/s200/board%2Bship.jpg

 

This was another all-day task. Now that it's done, at least, I can go on to other things.

 

I created another small area to get the swimming working. I made a trigger (with a visible pressure plate placeable to mark the spot) and put kevL's appearance type case switch code into it, and tried it out with a test human. Worked perfectly! It switched her back and forth from a badger with no problems.

 

Perhaps a week ago, I would not have understood what a "constant" is, and that these all-caps names like APPEARANCE_TYPE_BADGER and OBJECT_SELF are just handy substitutes for what's actually an integer. However, I know that now, and so I could alter the code. I edited the code to pick appearance 4006 instead of the badger, which is the "human swimming" line in the new 2da.

 

http://1.bp.blogspot.com/-dKV7jjLBcmE/T-jYJTz56BI/AAAAAAAADVw/K-sBJSLneYU/s200/badger.jpg http://3.bp.blogspot.com/-ST_HxuuIOEU/T-jYIw7d_CI/AAAAAAAADVk/iHr0d83gPU8/s200/swim.jpg

 

I put it in the code, and voilà! Swimming works! Not only that, but my concern was unsubstantiated about it possibly changing the character's head when she's transformed back. Perfect solution. Thank you, kevL!

 

I'll add in the rest of the races to the code and get to work on the underwater area design. This one will involve some walkmesh trickery, but since it's an organic environment, I predict less trouble than with the ships.