Incline SHELTER update thread

[Black Bart Charley]

Its phase based, so already better than Fallout.

 

[soggie]

Not exactly phase based. It's roguelike-based, or Pause with Real Time as marketed by TS. Basically, when you spent an AP in the world, everybody else in the world spends 1 AP too.

 

[MasterSmithFandango]

This is looking really good. For some reason I really dig the look of the map. Hurry up and finish it.

 

[Awor Szurkrarz]

Personally I'm interested mostly whenever it tries to emulate the greatest ruleset ever created by the God of Nature for which there were written tons of books about tactics like X-Com/JA2 does or if it's another anti-realistic game where it's impossible to use sensible tactics.

 

[shihonage]

@Black Bart Charley: simulated animal population may be interesting, but while making the game I became painfully aware of the fact that nothing gets made by itself, and therefore I cannot spend time on things that do not directly benefit gameplay.

The only serious benefit of the "realmap"(tm) was that you could dispatch an agent to seek out the Player, but this can be faked too, and the faking can be better controlled.

Also, page 1 now corrected

@Awor Szurkrarz: PwRT plus cover system (next video will have the cover system). Possibly some melee martial arts moves (hopefully).

@soggie: Yes, when Player is animated, the world is alive. In the video the Minigun Dudes had 200 guns skill, so together they generated a chain of "bullet recoil" in the Player that never got interrupted by a miss, so it became an unstoppable sequence.

Would've had a better chance if Player was running in a straight line for a more continuous interval (inertia mechanic - bionic man effect). Still working out the details of it...

@Flying Spaghetti Monster: Development is constantly slowed down by my real job, dry eye syndrome and futile attempts to have something that resembles a life from great distance. But it will be finished.

The next important system is the cover system, which will affect combat and stealth. After that, it will all be about updating the existing systems/balance to work better in unison.

 

[Kosmonaut]

How do you do the sorting of visible entities? I had a lot of problems with this. Specially when a big structure (like a house) was blitted onscreen. When the PC or a NPC walked near the big object, sometimes it would appear in front of it, when it should be behind. Or viceversa.

At first, I had to split the big object in tile-sized chunks to prevent that from happening. And even then sometimes I would still have ordering issues.

At the end I tried to implement the entity ordering using OpenGL depth buffer (my engine was pure OpenGL).

 

[shihonage]

How do you do the sorting of visible entities? I had a lot of problems with this. Specially when a big structure (like a house) was blitted onscreen. When the PC or a NPC walked near the big object, sometimes it would appear in front of it, when it should be behind. Or viceversa.

At first, I had to split the big object in tile-sized chunks to prevent that from happening. And even then sometimes I would still have ordering issues.

At the end I tried to implement the entity ordering using OpenGL depth buffer (my engine was pure OpenGL).

The map is a 2D array. Each element (cell) of the array contains the following:

* a variety of information about static objects on it, its passability, etc etc
* index of the PERSON standing on it

In Shelter, people don't move pixel by pixel. They jump from a map cell to map cell, but the renderer interpolates their movement and creates animation to fake the smoothness.

So, when the map is drawn, it goes through each cell and draws them in a specific order - people and objects alike. I've changed this order a few times, and it's still not final. Right now, I THINK it's:

(for y=far to near)
(for x=left to right)
{
draw object or person
}

This does create some problems with ordering, still. Thanks to movement interpolation, a person's position in some situations can violate its desired draw order.

These issues haven't been fully solved, and they will be a pain in the ass.

 

[soggie]

How do you do the sorting of visible entities? I had a lot of problems with this. Specially when a big structure (like a house) was blitted onscreen. When the PC or a NPC walked near the big object, sometimes it would appear in front of it, when it should be behind. Or viceversa.

At first, I had to split the big object in tile-sized chunks to prevent that from happening. And even then sometimes I would still have ordering issues.

At the end I tried to implement the entity ordering using OpenGL depth buffer (my engine was pure OpenGL).

The map is a 2D array. Each element (cell) of the array contains the following:

* a variety of information about static objects on it, its passability, etc etc
* index of the PERSON standing on it

In Shelter, people don't move pixel by pixel. They jump from a map cell to map cell, but the renderer interpolates their movement and creates animation to fake the smoothness.

So, when the map is drawn, it goes through each cell and draws them in a specific order - people and objects alike. I've changed this order a few times, and it's still not final. Right now, I THINK it's:

(for y=far to near)
(for x=left to right)
{
draw object or person
}

This does create some problems with ordering, still. Thanks to movement interpolation, a person's position in some situations can violate its desired draw order.

These issues haven't been fully solved, and they will be a pain in the ass.

Is it so hard? I'd imagine by drawing back to front, left to right (in an isometric map) and having a weight for each entity type (terrain, object, player, UI, etc) sorted by the weight should logically take care of it.

I did it with a 3d array, where a 2d array stores the isometric tile information in a [y, x] manner and on each tile there's a array of entities that is drawn from lightest to heaviest (in terms of weight).

for ( y = far to near )
for ( x = left to right )
for ( z = lightest to heaviest)
render entity

This is for the renderer though, because I separated the entity list (logical) from the render-ables list (presentation). And yeah, my characters teleport from tile to tile too logically, while the renderer figures out the correct animation queue to call. So this creates interesting situations where if a character is walking halfway across the tile (in tile[x,y] logically but different screen[x,y] every frame), and dies, the character will teleport back to the previous tile (the current tile it is in logically) and perform the death animation there. This happens when a character is killed by reaction fire or overwatches.

 

[Awor Szurkrarz]

@Awor Szurkrarz: PwRT plus cover system (next video will have the cover system). Possibly some melee martial arts moves (hopefully).

How will damage system look like? Will hit results be sensible?
BTW:
Here is a neat idea for a damage system that is based on real combat statistics.

How will accuracy look? What are going to be quantities of ammo in the setting?

 

[shihonage]

@Awor Szurkrarz: PwRT plus cover system (next video will have the cover system). Possibly some melee martial arts moves (hopefully).

How will damage system look like? Will hit results be sensible?
BTW:
Here is a neat idea for a damage system that is based on real combat statistics.

How will accuracy look? What are going to be quantities of ammo in the setting?

These are pretty specific questions. The answers to them will only be valid after we've completed balancing the game.

I'm not sure what "sensible" hit results are, but you should understand that the #1 goal here is to create combat that is fun, rather than combat that is realistic.

I believe that any fictional Universe that is created should abide by a consistent and understandable set of laws, at which point the dead-on authenticity of these laws loses importance - see Starcraft's implementation of "Terrans".

The armor classes and hit mechanics should be easily understandable, upgrading armor and weapons should give visible and easily comprehensible results.

One of the luxuries of having a game on another planet is that I can create 3-chambered revolvers if I so choose. On a more serious note, for a while, I've been considering ripping off those temporary forcefields from "Another World". How would forcefields account for in that gentleman's realism paper?

Really, it's all a question of how a given element will balance against everything else and whether the overall sum of combat will be fun to play.

 

[Kos_Koa]

I remember you originally chose PwRT in order to accommodate the "dynamic world" model, but now that you are moving toward set piece maps, is PwRT still an ideal choice? Not that I have anything against it, just curious about your reasoning to keep it.

Nice videos by the way, I'm interested to see what your cover system will be like.

 

[shihonage]

I remember you originally chose PwRT in order to accommodate the "dynamic world" model, but now that you are moving toward set piece maps, is PwRT still an ideal choice? Not that I have anything against it, just curious about your reasoning to keep it.

The reasoning at this point is devastatingly simple... time investment. I cannot go back and change fundamental parts of the engine anymore, or the game will never be finished.

I did reverse to fake encounters, but out of dire necessity, and in a way that required the least amount of code change. It's wasting a lot of memory on the big map right now, but I'm not going deeper than that, either.

We'll make PwRT work.

Nice videos by the way, I'm interested to see what your cover system will be like.

Thank you. I'm also interested in seeing what it will be like. In practice

 

[MasterSmithFandango]

Cover system? OMG Popamole!

Just kidding. Anyway, I'm kind of curious about this PwRT system. Say a combatant is shooting at someone four panels away, but their action is to move an additional two paces away, which would be, for arguments sake, outside of your weapon range. I'm assuming both actions are resolved simultaneously, and moving targets probably cause negative modifiers to your accuracy, but beyond that, is it being shown as exploitable with you being able to float in and out of maximum weapons range as say you move out of their range, they change their next action to move to put you back into range, and your action is to shoot, then move, etc.

If I'm not making sense, I apologize, because I'm fucking retarded. I guess I'm just curious to see more of the combat in action.

 

[shihonage]

Cover system? OMG Popamole!

Just kidding.

I just realized that's what "cover system" sounds like, damn it

Actually I meant a cover system like in Company of Heroes, only less complex in execution, with obstructions providing inequal chances of blocking weapon fire.

Anyway, I'm kind of curious about this PwRT system. Say a combatant is shooting at someone four panels away, but their action is to move an additional two paces away, which would be, for arguments sake, outside of your weapon range. I'm assuming both actions are resolved simultaneously, and moving targets probably cause negative modifiers to your accuracy, but beyond that, is it being shown as exploitable with you being able to float in and out of maximum weapons range as say you move out of their range, they change their next action to move to put you back into range, and your action is to shoot, then move, etc.

You mean teetering on the weapon range edge, making them waste their burst while you slip out of range, and then open fire while they approach (and possibly pause to reload)?

It is one of the possible exploits. Right now you can exploit the system badly. You can run around, which makes you hard to hit as long as you keep the same direction, and eventually tackle one of them and knock them out. It's not super-easy, but it is doable.

While they're knocked out you can empty a burst or two into them.

Still working on making it balanced.

 

[Awor Szurkrarz]

@Awor Szurkrarz: PwRT plus cover system (next video will have the cover system). Possibly some melee martial arts moves (hopefully).

How will damage system look like? Will hit results be sensible?
BTW:
Here is a neat idea for a damage system that is based on real combat statistics.

How will accuracy look? What are going to be quantities of ammo in the setting?

These are pretty specific questions. The answers to them will only be valid after we've completed balancing the game.

I'm not sure what "sensible" hit results are, but you should understand that the #1 goal here is to create combat that is fun, rather than combat that is realistic.

Well, systems that try to follow reality tend to be pretty fun. JA2 and X-Com, for example. Or Operation Flashpoint. On the other hand, systems that don't follow it are often frustrating or boring.
Also, realism has one important advantages when it comes to balance - everything in reality has it's purpose. There are various weapons and all of them exist because they have their function. Creating an anti-realistic system forces designers to create new functions for everything and a lot of such systems fail to do so in a satisfying way.
In reality, there's stuff that needs to get done and there are ready methods of doing that stuff.
People do all manner of dangerous stuff like being soldiers, security guards, policemen, etc.
Somehow some of them live for a long time without SFLing, stimpak spamming, etc.
Sounds pretty balanced to me.
Then, there are people who actually do these kinds of jobs for the thrill of the action and people who take part in various hooligan fights, etc. for fun.
So, it can be pretty fun too.

It's easy to create frustrating situations when people don't go down when they take damage that would be critical in RL and survive, it's impossible to pump enough rounds into target due to some weird shot modes.
For example, in Fallout, it's easy to have an incredibly skilled characters - but even skilled characters have to rely on SFLing and stimpak spamming, because it's impossible to reliably kill/disable enemies with headshots from sniper rifle, it's impossible to double-tap enemies, perform mozambique drill and do other stuff that allows to avoid getting swamped.
It all can make games very frustrating and much less tactical than they could be.

BTW Reality can be pretty interesting. People can survive a lot but they go down very fast under right conditions - I think that a big part of playing tactical combat should be creating and exploiting these conditions.

I believe that any fictional Universe that is created should abide by a consistent and understandable set of laws, at which point the dead-on authenticity of these laws loses importance - see Starcraft's implementation of "Terrans".

Still, humans are only humans.

One of the luxuries of having a game on another planet is that I can create 3-chambered revolvers if I so choose. On a more serious note, for a while, I've been considering ripping off those temporary forcefields from "Another World". How would forcefields account for in that gentleman's realism paper?

Personally, I was never very fond of forcefields. Not because of realism, but because I have never seen any game where forcefields were actually cool.

Right now you can exploit the system badly. You can run around, which makes you hard to hit as long as you keep the same direction, and eventually tackle one of them and knock them out. It's not super-easy, but it is doable.

While they're knocked out you can empty a burst or two into them.

Still working on making it balanced.

Why would it be an exploit, if it's not super-easy?

 

[shihonage]

I'm not sure how we can abide by reality in a game, if in reality, should you have a gun, you can come up to most people and shoot them in the head, and they will, most likely, die.

Same goes for Player. One-hit deaths from a stray bullet are rarely fun. This throws a lot of the balance right under the bus. What's the point of getting more powerful guns if the majority of the people don't wear armored helmets?

An RPG is a combination of systems meant to work together, and each of those systems is a game of its own. It is one of the most complex, demanding genres to create.

So, I can't really do extensive research on reality of firefighting, and then do extensive research on reconciling it with proven gameplay mechanics. I don't have the time.

The games you used as examples - JA2, X-Com - are tactical combat games. That is their focus, and as such, they can afford to leisure in making it as detailed as it possibly can be.

I cannot really take features from "FBI Miami Firefight" articles or Youtube videos of JA2. If you name a couple of key features from JA2 combat that you really liked, I will look at them.

But creating an ultrarealistic gunfight utopia should be reserved for a fully staffed full-time development team making a dedicated tactical combat game.

 

[MasterSmithFandango]

Probably the things I like most about JA2's combat is the way attacking body parts have significant effects (it is worthwhile to shoot people in the legs at times). I also like the fact that the more you get shot in combat, the lower your capabilities are. Simply put, I always hated that in most games, beyond being one hit away from death, I could take a huge beating and my combat potential is not significantly harmed. Also, being unable to really heal in combat is a plus.

 

[shihonage]

Probably the things I like most about JA2's combat is the way attacking body parts have significant effects (it is worthwhile to shoot people in the legs at times).

Close to beginning of this video Player takes out Giant Rat's legs, and it can no longer run.

Also, crippling arms lower accuracy, and crippling groin starts HP degradation over time (at the moment). Crippling head... actually I intend for it to disable the textbox window on the left and some other parts of game interface, literally making Player less aware of stuff like enemy HP, criticals, remaining bullets, and things like that.

I also like the fact that the more you get shot in combat, the lower your capabilities are. Simply put, I always hated that in most games, beyond being one hit away from death, I could take a huge beating and my combat potential is not significantly harmed.

Well during this beating the Player would suffer from probably at least one crippling blow to a target area as described above... and his armor may get beaten into 0 endurance...still all under construction.

Also, being unable to really heal in combat is a plus.

Right now healing toggles "fiddle" animation so the enemies start firing at you while you heal. Disabling it entirely shouldn't be a problem either.

 

[Awor Szurkrarz]

I'm not sure how we can abide by reality in a game, if in reality, should you have a gun, you can come up to most people and shoot them in the head, and they will, most likely, die.

How is it a bad thing?

Same goes for Player. One-hit deaths from a stray bullet are rarely fun.

In opposition to deaths from getting swamped by several enemies (which happens regularly because it's difficult to quickly dispose of people), hit point attrition deaths and deaths from random criticals?

This throws a lot of the balance right under the bus. What's the point of getting more powerful guns if the majority of the people don't wear armored helmets?

Shooting people through cover, shooting people's vital locations through other body parts or through other people, higher chance for one shot kills, higher chance of breaking bones, shooting off limbs, etc.

So, I can't really do extensive research on reality of firefighting, and then do extensive research on reconciling it with proven gameplay mechanics. I don't have the time.

There's no such a thing as proven gameplay mechanics when it comes to cRPGs as the combat is usually the weakest part of them.
Also, it takes much less time to pick up some observations about how reality works and some than trying to invent and balance new mechanics.

The games you used as examples - JA2, X-Com - are tactical combat games. That is their focus, and as such, they can afford to leisure in making it as detailed as it possibly can be.

What X-com and JA2 did was taking some common sense stuff that one can pick up after a short research. Also, X-com isn't very detailed.

I cannot really take features from "FBI Miami Firefight" articles or Youtube videos of JA2. If you name a couple of key features from JA2 combat that you really liked, I will look at them.

-Vital locations - getting hit in them means double-triple damage (usually ending with death) + a high possibility of getting knocked out.
-Aiming - spending more action points on a single shot allows to increase hit probability. Firing without aiming usually misses.
-Ability to fire quicker follow-up shots at lesser AP cost.
-Getting wounded lowers performance of characters.
-No healing during combat.

But creating an ultrarealistic gunfight utopia should be reserved for a fully staffed full-time development team making a dedicated tactical combat game.

Neither X-Com nor JA2 is ultrarealistic. It's semi-realistic.

 

[shihonage]

I'm not sure how we can abide by reality in a game, if in reality, should you have a gun, you can come up to most people and shoot them in the head, and they will, most likely, die.

How is it a bad thing?

It destroys passability and game flow control. It's axiomic.

Same goes for Player. One-hit deaths from a stray bullet are rarely fun.

In opposition to deaths from getting swamped by several enemies (which happens regularly because it's difficult to quickly dispose of people), hit point attrition deaths and deaths from random criticals?

Yes.

This throws a lot of the balance right under the bus. What's the point of getting more powerful guns if the majority of the people don't wear armored helmets?

Shooting people through cover, shooting people's vital locations through other body parts or through other people, higher chance for one shot kills, higher chance of breaking bones, shooting off limbs, etc.

Yet there's no point of needing all this if you can just shoot them in the head whenever you're in the next cell.

So, I can't really do extensive research on reality of firefighting, and then do extensive research on reconciling it with proven gameplay mechanics. I don't have the time.

There's no such a thing as proven gameplay mechanics when it comes to cRPGs as the combat is usually the weakest part of them.
Also, it takes much less time to pick up some observations about how reality works and some than trying to invent and balance new mechanics.

No, that's putting things on their head.

Balancing is unavoidable no matter what. However making mechanics WORK takes time, and importing ideas from real world takes a LOT of time.

Ideas are a dime a dozen. Ideas for gameplay mechanics and even whole new games. Everyone has ideas.

Implementation is priceless. Implementation is proof that someone took an idea and made it work within framework of a game, without breaking the shit out of that game.

No matter how "weak" you perceive Fallout or Diablo mechanics to be, they WORKED for those games. They controlled passability as intended, and as such, they're worth a thousand of unprocessed "ideas".

This is why I'd much rather steal from an existing game than a detailed account of an FBI firefight.

-Vital locations - getting hit in them means double-triple damage (usually ending with death) + a high possibility of getting knocked out.
-Aiming - spending more action points on a single shot allows to increase hit probability. Firing without aiming usually misses.
-Ability to fire quicker follow-up shots at lesser AP cost.
-Getting wounded lowers performance of characters.
-No healing during combat.

I'm not sure what APs mean in PwRT combat, but I like these ideas and framework for parts of them is already there. Such as, when you stop immediately after running, your inertia is still high and you're likely to miss. We'll see how much of this I can implement.

 

[Kosmonaut]

Well, systems that try to follow reality tend to be pretty fun. JA2 and X-Com, for example. Or Operation Flashpoint. On the other hand, systems that don't follow it are often frustrating or boring.
Also, realism has one important advantages when it comes to balance - everything in reality has it's purpose. There are various weapons and all of them exist because they have their function. Creating an anti-realistic system forces designers to create new functions for everything and a lot of such systems fail to do so in a satisfying way.
In reality, there's stuff that needs to get done and there are ready methods of doing that stuff.
People do all manner of dangerous stuff like being soldiers, security guards, policemen, etc.
Somehow some of them live for a long time without SFLing, stimpak spamming, etc.
Sounds pretty balanced to me.
Then, there are people who actually do these kinds of jobs for the thrill of the action and people who take part in various hooligan fights, etc. for fun.
So, it can be pretty fun too.

It's easy to create frustrating situations when people don't go down when they take damage that would be critical in RL and survive, it's impossible to pump enough rounds into target due to some weird shot modes.
For example, in Fallout, it's easy to have an incredibly skilled characters - but even skilled characters have to rely on SFLing and stimpak spamming, because it's impossible to reliably kill/disable enemies with headshots from sniper rifle, it's impossible to double-tap enemies, perform mozambique drill and do other stuff that allows to avoid getting swamped.
It all can make games very frustrating and much less tactical than they could be.

BTW Reality can be pretty interesting. People can survive a lot but they go down very fast under right conditions - I think that a big part of playing tactical combat should be creating and exploiting these conditions.

What a load of bulshit.

 

[soggie]

What a load of bulshit.

Awesome constructive criticism.

 

[Awor Szurkrarz]

I'm not sure how we can abide by reality in a game, if in reality, should you have a gun, you can come up to most people and shoot them in the head, and they will, most likely, die.

How is it a bad thing?

It destroys passability and game flow control. It's axiomic.

How does it do that?

Same goes for Player. One-hit deaths from a stray bullet are rarely fun.

In opposition to deaths from getting swamped by several enemies (which happens regularly because it's difficult to quickly dispose of people), hit point attrition deaths and deaths from random criticals?

Yes.

How so?

This throws a lot of the balance right under the bus. What's the point of getting more powerful guns if the majority of the people don't wear armored helmets?

Shooting people through cover, shooting people's vital locations through other body parts or through other people, higher chance for one shot kills, higher chance of breaking bones, shooting off limbs, etc.

Yet there's no point of needing all this if you can just shoot them in the head whenever you're in the next cell.

1. He may have some allies, so you kill one guy and you still have to deal with the others. And more firepower is useful when situation gets messy. A man-melting plasma rifle that will kill/disable even when not hitting a vital location is probably more useful than a slug thrower.
2. You need to get to that cell first. He may object to you approaching him with a gun in your hand.

Also, there are critters, which can be more difficult to put down than humans.

They controlled passability as intended, and as such, they're worth a thousand of unprocessed "ideas".

As a player, I'm interested mainly in a good tactical combat when combat is present. Fallout doesn't deliver it. All the passability control in the world won't change it.

There are three basic combat types:
Semi-realistic, abstract-brutal and abstract-hitpoint attrition based.
Semi-realistic would be JA2 and X-Com. They rely on using semi-realistic tactics and common sense.
Abstract-brutal would be chess, checkers and other games done for purely intellectual gameplay. You make a bad move, you lose. Brutal.
Abstract-hitpoint attrition would be another intellectual style, except that this one is for accountants.

Fallout takes a hitpoint attrition based system and then for some weird reason adds tons of potential critical hits and allows neither using realistic tactics, nor a chess-style intellectual challenge, nor does allow playing as an accountant (because of critical hits).
The only solution is SFLing and using exploits. It doesn't sound like something that works to me.

Then there are all these execution-style encounters with gangs of raiders, etc. (diplomatic character being a viable path, good joke)

I'm not sure what APs mean in PwRT combat, but I like these ideas and framework for parts of them is already there. Such as, when you stop immediately after running, your inertia is still high and you're likely to miss. We'll see how much of this I can implement.

Spending more APs would probably mean spending several time segments. For example in GURPS you can aim for more than one turn (it has 1 second turns).

 

[Darth Roxor]

How I fucking hate 'realism' in my video games.

 

[Kosmonaut]

It seems that also the combat has his larpers. And I hate when people talk about JA realism. Yea right, the fucking turn-based combat, fucking red and green squares, interrupts and headshots that don't kill or cause severe brain damage are so fucking realist.

Hurr durr. Raelism is good, abstract bad.

And I fucking love the "realism is balanced and is not complicated".