[Role Playing Game] Memento Mori

[GunSlinger]

Wow. That is amazing.

 

[Karassu]

Man, the hair! Is this Warcraft!? Really?

 

[bethmachine]

Is that beard growing thing done by animations, cause then I'm sure that would be the most obvious way or is it done by a different technique?

 

[Talavaj]

Yeah, a lengthy animation with multiple frames for each state of the hair. Most people probably won't notice but hey

There still won't be much of a way to change the clothes, because it is really kinda unnecessary, the only items you will be able to wear are carrying capacity related gear and certain utility accessories.
You will however be able to unlock different loadouts which all come with unique outfits, one for each of the four different endings and the rest for completing tasks and secrets in the game.

 

[bethmachine]

I would notice my beard eating up the crumbs of food I've scavenged, and how long can it grow exactly, estimation?
The loadout thing is intresting.
I'm sorry to get oftopic but I just have to ask, back in 2013 did you ever like requim seeing your project and kithryas project was the biggest and most intresting to me, was there any rivalry going on, what started the rivalry? If you disliked it tell me why you did. Sorry to ask but I just find this topic intresting.

 

[Talavaj]

The beard on the screen is about halfway grown, so it can get to about twice the size in about 28 in-game days.
I thought of making the hair grow as well but the pace would be unrealistic if I was to make it noticeable in any way and it looked really weird heh.

The loadouts are your starting equipment, but they also give you a different outfit and the rarer ones can even grant a stat bonus or an ability or two.
They are mostly realistic and practical, but there will be a couple gag ones that are unlocked via easter eggs.

Requiem ? No rivalry to speak of, we just happen to have polarized opinions on content creation. The author of Requiem likes to go the easy route of using content of other people and getting praise for their work.
It is something I frown upon, praise should be deserved. I value effort very highly.

 

[bethmachine]

Ohh I see, is there any other features you've been working on?

 

[Talavaj]

Main challenge is still the level design, alas it still does not look the way I would want.
The visual vibe is very crysis esque but I recently drew some inspiration from vanishing of Ethan Carther which I find to be one of the most stunning games I have ever played.
Nevertheless !

I will list most of the changes in the OP or maybe even make a new thread given that a lot of information posted before is not actual anymore.

Main change is how the level is done, the open world is a grid of streets while the buildings are randomly generated blocks. So one time you may get more residential buildings while other times you will get more offices or malls.
The open street is where most random setpieces spawn, eg a police car or an ambulance that may not be there next time you play. The buildings are instanced (apart from having a ground level in some cases) and will give you a random level layout to scavenge through. Going into a building costs you some in-game time and fitness, so you basically have to go straight home after a run.

Hideouts have two stats, comfort and safety and some may offer extra expansion slots.
The stats come in play when you go to sleep, low comfort hideouts can cause you to not sleep well due to noise or cold forcing you to wake up while low security hideouts can cause you to get attacked in the middle of the night or require you to scout the perimeter due to suspicious sounds.
Better hideouts are less likely to have these events occur, but they are still very much part of the game as you can run into unique events that may offer you things you cannot find anywhere else.

On start you do not have a hideout, and you are given a small number of items depending on your loadout. Crappy hideouts are unlocked right at the start while the better ones may require specific task AND materials.
You can also upgrade the comfort/security of your hideout but only a limited number of times, so you will HAVE to move sooner or later.

When you die, you DIE Game over. I adopted rogue-like elements for higher replay value and thrown out the old easy setting that would allow you to respawn.
You do however carry stats over, you can unlock different loadouts and perks that carry with you to the next level.

You progress the story via doing specific runs in specific places that always generate on the map. There will also be some mild puzzle solving, random setpieces may spawn with a story behind it that you can solve either just to get the picture of what happened there or to obtain extra stuff. Many unlockables are hidden like this and doing certain puzzles may alter the ending.

Think Dayz, with base building features of State of Decay and with extra rogue-like and rpg elements. Actually, it is very much like State of Decay but with randomization and less open world, and there are no zambies in this uhuhu.

 

[bethmachine]

I like that you chose to remove the easy difficulty as I see being able to respawn not something suiting for this map. Talvaj do you want to pop some screenshots of what you've done with the design of that open world?

 

[Karassu]

I really wanna play this.

 

[Gekigengar]

Amazing how you are still determined to go on.

After I finished my project, there is literally nobody to play it with. ;/

So I move on..

 

[Vinz]

The character is amazing. The face is very realistic. I like it that you split up into many textures for even higher quality.

 

[Talavaj]

-

Yeah I know what you mean, I probably won't be doing the multiplayer for the same reason.
But most of the assets will be ported from my unity stuff so it is not just a strictly wc3 project anymore, as such it would be a waste to not continue as I'm gonna work on it either way especially given that I have made much of the design and script already anyway.

In the end even if it gets a couple hundred downloads and bunch of those people will like it it will be a success !
Besides the main portion being single player, it means people may be able to find it and download it even a couple years after when playing anything online will be pointless.

Talvaj do you want to pop some screenshots of what you've done with the design of that open world?

Yes, I'm working on that

 

[tobyfat50]

Very good idea to move to Unity. Think wc3 got us kinda defeat with its limitations, we spent more time faking stuf than making it.

The terrain thing is interesting but is that unity or wc3, seems like model editor? Mesh or shader?

You'll use normals in unity right? Parallax shader isn't to bad as a alternative.

 

[Talavaj]

I had some time on my hands today and did a speed model of the sewer instance, made the entire whitebox in just a couple hours.
Sewers will be a small and claustrophobic instance devoid of all light, so you will have to use your own sources of light to navigate it.
The "dungeon" will be easy to access from the main level and contain generous amounts of loot, mostly materials flushed down from the streets but it will also be very dangerous to scavenge through.
It will be one of the three underground instance types (other two being parking lot and the metro) that will require player-supplied light to search.

A small collapsed portion of the sewer will also serve as one of the earliest available hideout locations for the players.

Old Post

I'll only move the assets to unity, for now at least. I may remake the project in unity as well who knows lol, but not before this version is done.
But yeah I'll have mostly baked normals from mudbox, or crazy bump for the less important surfaces.

The terrain is displacement map applied on tessellated mesh then retopologized to low poly that is slapped onto tileset as a decal, it helps a lot breaking up the terrain because the tiling does really really show and it looks better than separate models because I can blend the models seamlessly into the ground.

I started playing with it after playing through Vanishing of Ethan Carter which is hands down one of the best looking games I have ever seen and it barely has any shaders at all.

 

[GreeN!X]

I'm still figuring out how you're going to fit all those details in a warcraft 3 map. If you're going to use unity, okay, awesome then. IF you're going to turn it as a campaign, it might be possible. But as a single map? I'm not so sure about that with warcraft's limitations. :/

 

[Talavaj]

What limitations ? The playable map size ?
The map space I use is unlimited, every indoor area is loaded in the same portion of the map. As soon as you leave it, the instance is deleted forever. As soon as you enter a new instance its entire level is created there etc.

 

[Talavaj]

How is it going hive ?
I'm too lazy posting updates in this thread, giving it is filled with all sort of outdated shizwaz.
I'll have this topic locked and remake it when I'm deeper down the road.

Just chiming in to report that two products based off Memento Mori's stuff will be hitting the Unity asset store sometime this month.
They will include most of the items, tools and melee weapons from the mod and the entire Sewer level.
With that off my mind after, I'll be moving onto the city itself as the last step in making the alpha mod playable.
From there on I'll make the rest of planned indoor locations and release them both in as playable locations in the mod and as assets on the store simultaneously.

You can get some sneak peek from the images below. Such as the upgradeable baseball bat and some of the assets from the sewers levels, the finalized things from the mockup in the earlier replies.
There is a bunch of do-it-yourself weapons and a bunch of weapons that can be upgraded, some cannot be upgraded at all. I'm trying to streamline it but it is a mess really.
If you have some cool ideas for improvised weapons and tools you can drop them in the thread here.
Peace~

 

[Heinvers]

[shocktag]It's alive!!![/shocktag] *gasps*

Yeah, looks great. I hope you'll get time for it Tal

 

[Noypi]

The overall idea sounds more than just interesting, can't wait to see it uploaded on the hive

 

[Talavaj]

[shocktag]It's alive!!![/shocktag] *gasps*

Yes, it is of course lol.
I'm just working in my pace as always, things take a bit longer than usual because I have to make them on par or better than the existing unity asset store stuff.
Porting it back to wc3 is easy as I just blend the texture maps together and cut them out of the texture atlas then shrink them down to the 512p maximum for wc3 if needed.

I get optimal assets for my mod while somebody else may end up using the full resolution source models somewhere else and I may not only deliver a game I always wanted to make but also make some buck.
A win-win scenario !

 

[Zwiebelchen]

I gotta say the concept of just having a single "baked" model for an entire zone and load just that is amazing.
However, did you actually do some tests on this mechanic ingame or is that just what you plan?

I could imagine getting rendering issues with the crappy WC3 engine, with polys or entire models disappearing if too close to the camera.

Also, how are you going to do the collision checks? Obviously, you can bake an entire landscape into a single model, but you can't bake in the pathing, right?
I mean; you can create a custom pathing texture and you can have a walkable destructable, but these come with some limitations in that you can't define which planes will be walkable on the destructable and which are not.

In case of your sewers instance, if you "load it" into your map as a walkable destructable, then the units will walk "on top" of the tube, not inside.

How are you going to deal with that problem? Or do you intend to split it in half when it's done to seperate the roof from the ground?

 

[Talavaj]

I haven't experienced problems with extents so far, if you export the whole model with proper extents that is. The only thing that refuses to work is a large walkable destructible.
It gets culled no matter how big the extents are, so I have to replace it with an invisible mesh with empty material for pathing on top of a regular doodad.
So yes the sewers models for example have no issues with pathing because it is two meshes. One regular one, and a much simpler one which is the walkable platform.

As for pathing, the "instance" models are already scaled to match the wc3 pathing grid and the pathing is made by killing/resurrecting pathing blockers according to the current layout of the instance.

I have adapted most of the procedurally generated features from a simple test of a minecraft-like biome generation thingy that I have made.
In which I needed water to be non-walkable at any place the "engine" decides to place it.

This is how the procedurally generated grid world would look like, the grids are aligned to the size of the regular large pathing blocker. The pathing in memento mori works the same but the killed/resurrected blockers and displayed tiles are preset not random.

 

[Zwiebelchen]

I see... pretty clever solution, albeit a lot of extra work.

How do you register how each model affects the pathing blockers? Do you have a scanning algorithm in place that moves locations per tile and then detects the Z-difference to adjacent tiles caused by the walkable destructable and blocks them if they have a too high z-difference?

 

[Talavaj]

Because the pathing blockers are the only destructibles in the map, I simply pick all destructibles within region and kill/resurrect them while ordering the tile doodads in the same region to play a specific animation depending on the biome it should be in. The doodad has different texture in different animations.
Biomes are made by simply scattering regions around the playable map area, if different regions overlap each other they attempt to move elsewhere while same biome regions are not forced to move in order to create "clumps" of the same biome type after that tiny regions are scattered around the map and replace existing tiles back to water to make the biomes look less artificial.
The squares occupied by water simply have their pathing blockers left intact and their tile doodad's animation set to "hide". The actual water is just a doodad submerged below the terrain.

It's all in GUI, I literally just cobbled it up in a few minutes.

 

[Zwiebelchen]

Because the pathing blockers are the only destructibles in the map, I simply pick all destructibles within region and kill/resurrect them while ordering the tile doodad to play a specific animation depending on the biome it should be in. The doodad has different texture in different animations.
Biomes are made by simply scattering regions around the playable map area, if different regions overlap each other they attempt to move elsewhere while same biome regions are not forced to move in order to create "clumps" of the same biome type after that tiny regions are scattered around the map to make the biomes look less artificial.
The squares occupied by water simply have their pathing blockers left intact and their tile mesh animation set to "hide". The actual water is just a doodad submerged below the terrain.

It's all in GUI, I literally just cobbled it up in a few minutes.

Yeah, I know how you would do that in a minecraft-esque algorithm. That wasn't my question; I was talking about Memento Mori here, where obviously you don't have randomized biomes.

How are you going to detect which pathing blockers must be killed and which must be alive based on the scene loaded? That's what I meant with moving locations around and getting the z-difference.

 

[Talavaj]

No detecting really, because the layouts are pre-made the trigger that picks a random layout is also triggered to kill specific regions.
There is one region per pathing blocker for detail pathing, and also larger regions for simpler layouts. Things like lootable objects are scattered similarly like the minecraft-like world gen but in a smaller scale within the selected layout.
I guess it would be more convenient for the game to detect the pathing by itself, but given the layouts really aren't that random it isn't really necessary.

The whole idea is about cramping things in to save space, the randomness is a bonus lol. In a way the actual map is just an overworld that allows you to access different locations kinda like overworld in Final Fantasy would.

 

[Zwiebelchen]

No detecting really, because the layouts are pre-made the trigger that picks a random layout is also triggered to kill specific regions.
There is one region per pathing blocker for detail pathing, and also larger regions for simpler layouts. Things like lootable objects are scattered similarly like the minecraft-like world gen but in a smaller scale within the selected layout.
I guess it would be more convenient for the game to detect the pathing by itself, but given the layouts really aren't that random it isn't really necessary.

The whole idea is about cramping things in to save space, the randomness is a bonus lol. In a way the actual map is just an overworld that allows you to access different locations kinda like overworld in Final Fantasy would.

Yeah, considering the models are static, you can do that; however, the point why I'm asking this was: it actually requires A LOT of rects and there actually is a way to automate this which could get rid of all the rects and regions.

Basicly, what you can do is have a single location that you move around your tile-grid; starting at MapMinX+16 and MapMinY+16. This marks the "center" of the leftmost bottom pathing cell. Then you iterate over the map at steps of 32 coordinate units and move the location there and detect the z-coordinate of the location.
Then you can actually compare if the z-coordinate of the scanned location is within a certain treshold to the neighboring cells.
If not --> enable the blocker

Basicly, an algorithm like that scans height differences based on what a normal human could "climb" with a simple step. Staircases would basicly be within the z-treshold, so would remain walkable, whereas walls or fences would be rendered unwalkable.

 

[Talavaj]

Hmm yeah, I could do that. It is indeed around 150 regions the way it is currently if I remember right.
Making walls part of the walkable platform should work just fine then ?

 

[Zwiebelchen]

Hmm yeah, I could do that. It is indeed around 150 regions the way it is currently if I remember right.
Making walls part of the walkable platform should work just fine then ?

Yeah; actually, that is required for this to work, as you basicly need to have large enough Z-difference between adjacent cells to detect a wall.

The cool thing about this approach is, that is also provides an automatic sanity check: if a terrain slope is too steep for someone to climb (for example a cave wall), it will automaticly enable the blocker underneath.

You basicly only have to set the z-treshold to the length of roughly half a leg of a human because this is kind of the maximum height you can elevate in just one step without having to climb or use your hands.

 

[Talavaj]

Easy, got it !
However there is a problem, using GetLocationZ yields the same value in both animations of a walkable elevator for some reason.

EDIT:
Sorry, I'm retarded. I missclicked and was getting location X instead of Z. Works flawlessly !

 

[Zwiebelchen]

Requires some fine-tuning, obviously, especially when your walkable model is not exactly created on a per-cell base:
When a wall element falls exactly between two scan points, it will be undetectable; so any wall has to be at least 1 ingame cell wide in the walkable model.