1955e3 No.15574119
"You will never finish your game" edition
Resources
>>>/agdg/
>>>/vm/
>#8/agdg/ via irc.rizon.net
>Dev resources: http://8agdg.wikidot.com/resources
Links
>Wiki: http://8agdg.wikidot.com/
>Beginner's guide: >>>/agdg/29080
>Previous thread: >>15537059
Announcements
>QUARTERLY DEMO DAY SCHEDULED FOR November 11th, 30 days away
>Please contribute to the wiki if you can
e6822f No.15574132
You're gonna get us in trouble, old thread is only on page 8
12e3be No.15574150
You are only allowed to create a thread at page 13 or at 700 posts. It's in the wiki.
http://8agdg.wikidot.com/wiki:thread-template
1cd679 No.15574315
mark senpai pls no bing bing wahoo these threads like you did lol,waifuwendsday,4am,fridaynite,everythingelseboardculter
553677 No.15574357
553677 No.15574522
8fd1ba No.15574531
>>15574522
>>15574357
Why are you samefagging?
20a659 No.15577381
Is this the new bread or is this dead bread?
7eeef3 No.15577399
>>15577381
this is moldy bread that will soon be recycled into fresh bread when old bread has gracefully expired
just b patient
12e3be No.15579110
ya uln cphlegeth hupadgh hlirgh, ch' nglui ngbug uh'eagl
uaaah
This thread's is now no longer dead
8fd1ba No.15579871
OH SHIT WHAT UP FELLOW GAMEDEVS
f0ff76 No.15580291
I posted this in the other thread, but it was recommended that I post here.
brainlet here, looking for advice on implementing turn-based action system
Some approaches I've thought of:
<1:
> All entities which can take turns have a "Turnable" component which has a boolean "active" as its sole member; relevant systems which cause entities to act will set the "Active" value to 0.
>Another system keeps track of the entity that was active, and when it's no longer active, moves on to the next entity in a list, and sets it active
I'm not really sure what data structure to use for my list of entities. Maybe a circularly linked list? Regular array?
<2:
>list is strictly an array of pointers to entities
>every "Turnable" entity has two components: Speed, which describes how often a given character can move, and Tdelta, which is the amount of time since its last turn.
>instead of doing like a circular iteration through a list for the turns, iterate through list of all entities, adding fastest playable character speed value to tdelta for all characters aside from fastest playable toon (in case there could be multiple playable characters concurrently)
>while waiting for playable toon to act, it will block
>cycle through all characters currently acting before reaching next round
so (horrific) pseudocode is something like this:
std::vector<Entity> entityList;
//TurnSystem
def markActives():
fastestPlayableCharacter.addComponent(activeTurn);
for ent in entityList.filter($0 != fastestPlayableCharacter):
ent.Turnable.Tdelta += fastestPlayableCharacter.Turnable.Speed;
if(ent.Turnable.speed < entity.Turnable.Tdelta)
ent.addComponent(activeTurn);
ent.Turnable.Tdelta = 0;
I think the second one is pretty decent. Any criticisms?
While I was typing my terrible pseudocode in the comment window, I realized that Tdelta could be larger than speed, so maybe I'll have to add another variable representing the overflow for the next turn.
>>15579936
Not every playable character gets a turn every round. Only the fastest character gets a turn every round, and he decides the value that gets added to every character's Tdelta. Whenever a character's tdelta exceeds their speed, they get a turn, regardless of whether or not they're playable.
f0ff76 No.15580342
>>15579936
oh, and in response to the rest:
I don't think I'll allow the turn order to be changed, it'll be playable characters first, and then npcs in whatever order. I don't think that matters.
However, I wasn't thinking about the case in which a monster could have some fraction of the fastest toon's speed. Maybe I can add a field to the activeField component corresponding to how many turns that character gets on this particular round, and then go through all of them at once.
12e3be No.15580893
>>15580291
That makes sense.
>>15580342
Ok. So the highest speed is added to everyone's Tdelta at the beginning of every turn, and the only people who get to play are those who's Tdelta exceeds their own speed. Everyone who played has their Tdelta reset to zero, or as you mentioned >>15580291 maybe it takes the difference and sets it to that.
However with this system it seems like every character would go on every turn.
Player character (PC): sp 10
Goblin (GB): sp 8
Ork (OR): sp 3
Turn 1
>PCTD: 10
>GBTD: 10
>ORTD: 10
Tdelta is greater than speed for all characters, everyone goes and everyone gets reset back to 0.
A way around this is have everyone's Tdelta set to the fastest speed, and subtract everyone else's speed every turn. When they go to zero or negative add the fastest player's speed or reset to the fastest player's speed, whichever you choose. Here it is adding the fastest player's speed.
Player character (PC): sp 10
Goblin (GB): sp 8
Ork (OR): sp 3
Turn 0
>PCTD: 10
>GBTD: 10
>ORTD: 10
Turn 1
>PCTD: 0
>GBTD: 2
>ORTD: 7
the pc goes, the goblin doesn't, the ork doesn't
End of turn 1
>PCTD reset to 10
Turn 2
>PCTD: 0
>GBTD: -6
>ORTD: 4
the PC goes, the goblin goes, the ork does not
End of turn 2
>PCTD reset to 10
>GBTD set to 4
Turn 3
>PCTD: 0
>GBTD: -4
>ORTD: 1
the PC goes, the goblin goes, the ork does not
End of turn 3
>PCTD reset to 10
>GBTD set to 6
and so on.
0ac45e No.15581493
You can now equip items and they affect stats like damage, movement speed and so on. They also appear as 3d models when worn. Weapons influence what attack animation is used.
b04b5a No.15581521
>>15581493
I guess I can quit now
0ac45e No.15581587
>>15581521
You mean quit being a faggot, right?
b04b5a No.15581638
>>15581587
y… yeah
I'm just jelly of the progress rate
5c1e57 No.15581660
>>15581638
He's working on overarching systems while you're working on details, of course the progress rate will appear different. Plus he presumably doesn't have a child to take care of.
0ac45e No.15581950
>>15581660
This. I'm also taking shortcuts everywhere. If something works but has few loose ends (like having ability to equip AK47 to your head) I leave it, I can fix it later. All audiovisuals are temporary free assets or stuff I made long ago because I want a prototype as fast as possible.
I don't have a child, but recently I got my first job and ironically since then I made more progress than ever before, because I just don't have time to get bogged down in detail polishing hell like I used to in the past
>>15581638
If you feel like your progress is too slow, I'd recommend you work more roughly and focus on big things, as if you were during a game jam, especially if you have little time and energy left after being done with your duties Plus, it's more fun to dev this way.
5c1e57 No.15582142
>>15581950
Strapping an AK to your head is a perfectly valid tactic if you have a strong enough neck and telekinesis to pull the trigger.
4dcb86 No.15582247
>QuakeC
>No pointers
>Hacked in arrays that work slowly because no pointers
>Empty string counts also as FALSE with boolean operators, but when they aren't using boolean operators they count as TRUE
>if not(expression) is a thing and works like if(!expression)
> ++ and – operators will cause compiler bugs in some cases
>using function calls and array calls together might cause compiler bugs
What was Carmack thinking?
Even with all this weird shit, it's still easy to mod gameplay stuff. But I'm itching to alter a sourceport to accept lua scripts or some shit.
174594 No.15582287
>>15582247
It was 4D programming like a White man.
174594 No.15582307
Hey, I need an old 3d engine that works on Linux and Windows and won't make a toaster from 2009/2010 lag, what do you recommend?
5c1e57 No.15582386
>>15582307
Source, or maybe an older version of Unreal Engine.
dffc36 No.15582624
b04b5a No.15582628
>>15581660
>He's working on overarching systems while you're working on details
I guess, but I still feel like I'm neglecting the overarching systems for the fun but slow to implement details
>>15581950
>recently I got my first job and ironically since then I made more progress than ever before
that was me in 2016.
>>15581950
> work more roughly and focus on big things
I guess that's what I'm working on now…
you ok, beelze guy?
adead2 No.15582954
>>15582628
That's a big wound. Is it possible to close the wounds smoothly over time? I imagined Beelzebuy using some spell that dissolves the enemy corpses and absorbed them to heal himself. That'd look really fucking cool, sort of like reverse BG death.
1c5f7d No.15582979
Hello. Am newfag looking to get into AGDG. I chucked this together a few days ago in Bosca Ceoil, and I like it. It's supposed to be for a sub-hub area.
I think it's probably way too repetitive and short, but I would like the input of some anons.
Thanks for any response! ^_^
https://drive.google.com/file/d/1xvy1JAgTxHzFFONIch1OXOJTQhrmD_BX/view?usp=sharing
Sorry for the gulag link-my machine is down so I'm stuck using my mother's windows machine and I have no way of easily converting it into an attachable format
553677 No.15582980
>>15581630
pretty nice.
Are you just changing its shape based on the cup rotation?
Looks right enough to me, the only issue is that it does not spill
d4518b No.15583239
Holy shit I became a lot more productive when I started moving things from my "todo" list into a "completed" list instead of just deleting them when they're done.
It's like collecting shit and completing tasks in a videogame.
f5d04a No.15583423
>making a 2D engine from scratch for the hell of it
>just made a basic particle system
>have yet to implement basic sprite rendering
I know I'm doing this out of order, but I didn't realize how easy basic ECS systems could be implemented. I also got more progress done in the last couple hours than in the past couple weeks.
>>15583239
Similarly, I find it easier to physically write down things I need to do on 3x5 cards, then cross things out in pen as they get completed.
>>15581493
Any throwable items yet? If you have only the ammo for the gun and not the gun, do you punch people with a fistful of bullets? If your character has low INT, can you throw the gun or whack people with it instead? Loving the progress so far.
adead2 No.15583937
>>15583423
No throwables yet but there will be. I'm also thinking about adding possibility of throwing non-weapon items that were put into a weapon slot with some minor effects. For example a short stun for items that'd be annoying to get hit by like some sort of dust, damage for heavier items or healing for a healing potion.
>If your character has low INT, can you throw the gun or whack people with it instead?
When I was beginning to test attack animations I made one character hold the the gun as a quick stand in for a mace, now I'm tempted to make it so when you have lowest INT everything becomes a melee blunt weapon.
15f819 No.15584052
Guys my game is finally getting into the final stages, I even opened a Patreon. I'm getting nervous tbh I've spent so much time and money on this.
How do you feel when you see so close if you will triumph or fail miserably?
1955e3 No.15584165
Working on a comfy scene for Godot Wild Jam 2. Theme is "Hidden Within", so I'm making an Inside clone. This is the starting area, then you go down into some kind of secret base. Enemy avoidal and platforming will be the main challenges. The jam ends in seven days so it'll be tough but I'm hoping it will improve my workflow.
>>15584052
>How do you feel when you see so close if you will triumph or fail miserably?
I've never been there. Good luck.
3312de No.15584232
>>15584165
that looks comfy. needs different music though
1955e3 No.15584242
>>15584232
It's just what I was listening to.
b9cb6c No.15584247
>>15584165
But anon, where's the ray tracing?! TURN IT ON!!!
TURN IT ON!!!
12e3be No.15584790
YouTube embed. Click thumbnail to play.
>>15584165
How do you do the snow in godot?
>>15584232
Here's what I was listening to when I opened the webm.
ab0c43 No.15584821
>>15582247
>>No pointers
>implying that's a bad thing
ae6fa5 No.15585093
Thanks to the anon who directed me to the godot docs yesterday and mentioned the extra functions for object ID on detection. I was able to get my object dependent hit detection working really simply with just
func _on_weapon_body_entered(body):
var name = get_path_to(body)
8fbb00 No.15585615
>>15582980
No I'm doing a combination of cheap shader trickery and building the mesh from plane+cylinder intersection math. It supports funnel shapes too so it never clips outside. Im gonna make it so you can spill it and it slowly goes down.
12e3be No.15585633
>>15585615
That's some sweet technology you have there.
ac4921 No.15586035
>>15585615
>I've been spending months of spare time implementing accurate fluid simulation for a game
>someone comes in with some cheap shader trickery and a few collision detection checks and it doesn't look half bad
I mean, it wouldn't work for what I need but that still feels funny in a sad sad way
1955e3 No.15586080
WIP player character, camera, and tonemapping/color correction.
inb4 >MOUSE
>>15584790
Particle effect.
b04b5a No.15586129
>>15582954
For you.
while this idea is fucking metal, beelzebuff will wear armour so the wounds healing won't be visible on him. This is made mainly for enemies, whom you'll be able to brutalize pretty badly
but maybe I need some vampiric enemy that does that
12e3be No.15586148
>>15586080
Looks comfy, and your lighting looks way better than it has any right to.
>Particle effect.
I could have guessed that. I want to know the details. What sort of math are you using to make them blow around in the wind in different directions individually but the same direction generally?
1955e3 No.15586212
>>15586148
It's rather simple, I simply used initial velocity with the z-axis pointing downward. Then I increased the spread value to about 10. Gravity is used on the x and z axes to achieve the overall direction (uncheck 'local coords' or you'll have to use different axes). Linear acceleration is used to add additional randomness to the downward motion. This is a terrible explanation, here's the ParticlesMaterial itself:
https://my.mixtape.moe/lqmdtl.material
>>15586129
Necromancers, man.
12e3be No.15586222
>>15586212
>here's the ParticlesMaterial itself
Thanks man, I'll look it over.
79385c No.15586233
I'm toying around with the idea of making a Goblin Slayer module for Neverwinter Nights and I was wondering how other people went about designing quests. I have no real prior experience and my outlook is to think about what different options would accomplish, why a player would want to do them, and work backwards from there. As an example, early on the player will be given the choice of several quests, one with a low stated reward killing goblin patrols which considering the subject matter would obviously advance the plot and one with a higher stated reward for clearing out a portion of the sewers for a higher reward and good relations with a merchant which a player would take if they wanted to more quickly become stronger.
a99003 No.15586334
Tonight I spent about three hours teaching myself how to calculate the probability of something happening X times out of Y possible times with a Z% chance of happening each time, then slapping together some code to run the calculations for me if I input those three variables. I can honestly say this is something I never thought I would ever need to know how to do.
8665ac No.15586377
YouTube embed. Click thumbnail to play.
>>15586233
Here's a pretty good talk on the subject by a dev who was the lead designer/project director for FO:NV
b04b5a No.15586462
welp, orthographic perspective which I preferred is getting scrapped…
ac4921 No.15586484
>>15586462
Is it unfixable, or just not worth the effort?
b0b026 No.15586491
Are there any statistics showing how many games are there per genre? I'd like to explore underpopulated genres, so far I have a couple of ideas going.
I also noticed that indie games are focusing on tycoons and sims lately.
5f1ca2 No.15586516
>>15586462
Glad to know you're still alive, anon. Hows beezlebub going?
98ac34 No.15586523
>>15586462
>is getting scrapped
No its not you can fix that shit.
b04b5a No.15587307
reposting to fix link
>>15586484
I think the shader just won't work with ortho perspective
>>15586516
it's been better for the past month or so, thanks, my dude. I reposted most progress in my thread
>>>/agdg/29865
>>15586523
>No its not you can fix that shit.
Yeah, there is a way: I can move the camera way back and set the fov to something very small - this would end up with a perspective almost identical to ortho. But that fucks with ambient occlusion, which adds a lot to the game's looks. But yeah, that SHOULD be fixable
f32a44 No.15587393
I'm having trouble implementing a function to handle combat, first, each call to the function advances it by a step, so I changed it into a state machine, this is because need to prompt the player and ask him for a set of actions one action at a time to complete his turn, up to this point it is simple, but, when a character makes an attack, the target needs to be prompted to choose an action to defend, so a turn depends on the actions of multiple characters and not just one.
It is a turn based combat, player vs AI, but I have no clue how and where to put the AI code, and how to properly design the state machine that can return within multiple actions for multiple characters, here it needs to return whenever it prompts the player for an answer, the answer has to be verified before further processing.
It's already confusing for 1vs1 let alone a party vs party.
Example for actions:
Change posture, move, attack (has to ask the target for a reaction), Contest (hast to ask the target for a reaction)
aa5bff No.15587417
>>15587307
>I can move the camera way back and set the fov to something very small - this would end up with a perspective almost identical to ortho. But that fucks with ambient occlusion, which adds a lot to the game's looks. But yeah, that SHOULD be fixable
This is why you write your own engine, kids.
4518d1 No.15587514
>>15587417
Shit man guess he's got to write his own engine now
a0109d No.15587791
>There is a demand for updating my dusty mod for Factorio 0.16.51
>try to port my mod over to this newer version
>hurr durr missing icon size
>now I need to edit over 6 million files just to add one more entry
>and other cryptic errors such as missing Key "underground_remove_belts_sprite" not found
I am aware that with newer game version some shit will break but god damn this icon_size parameter is one of the most necessary change ever, fuck you.
174594 No.15588661
I feel like a massive brainlet, I can't even raycast properly and whenever I try to look for inspiration somewhere else, it's a moron who copied and pasted lodev's raycasting code.
add85c No.15589342
>>15587393
The design that comes first to my mind is: Put it into a class with references to the combatants, and have each phase give specific options to the combatants that could advance it to the next phase. e.g. Player attacks -> combat moves to a state that gives combatant 2 options for response to an attack. Or player casts a spell -> go to a state that just checks combatant 2's spell resist, deals damage, and moves to a state where combatant 2 can act. AI code could have a combatant interface similar to the player but that automatically checks conditions before delivering its "choice" of event to the combat state machine. I hope that's even remotely helpful.
8fd1ba No.15589358
>>15587791
Oh yeah, how is Factorio modding? t. playing it right now instead of my game
174594 No.15589528
>>15589513
The cape physics are oddly realistic compared to the rest, is it going to be a part of the gameplay ?
1955e3 No.15589596
>>15589528
No it's just that I can't animate and would rather let Godot simulate a softbody.
b9cb6c No.15589692
>>15589513
Is that the house from courage the cowardly dog?
8fd1ba No.15590023
On a scale of 1 to "please don't sue me Factorio" how does this look?
12e3be No.15590076
>>15590023
It looks like a flash game, but one of those well made flash games.
8fd1ba No.15590083
>>15590076
I wasn't going for that specifically, but I'll take it
a0109d No.15590169
>>15589358
Modding it from 0.13-0.15 was relative painless because they didn't made too many breaking changes but from 0.15 and to 0.16? It is going to take much longer because they broke a lot of shit that needs to be fixed now to make it work, I haven't managed to port it yet due to that. But apart from that it is relative "easy" as you just need a text editor to make mods for it so no specialized tools which is a both a blessing and a curse. The worst part is the action/final_action parameter because it allows you to chain shit it in the worst possible way so you end up ending with spaced brackets such as this:
}
}
}
}
}
}
}
1955e3 No.15590225
Item pickups. Will be used for simply puzzle elements, but the player will also be able to throw objects to make noise and distract enemies.
1955e3 No.15590284
Simple throwing
>>15590225
>be used for simply puzzle elements
time for bed
b04b5a No.15590369
>>15587417
You're right, what was I thinking? And I was making progress as well…
I'll se y'all in 2040, when I'll finally get a black triangle to render. I'm sure it will all go smoothly after that.
9ada52 No.15590411
>>15587417
What the fuck stupid shit is this? Do you even know what ambient occlusion is?
248a9a No.15590463
>>15586462
I'm genuinely curious, what engine is this? What engine doesn't allow orthographic projection? Do you not have access to the perspective transform? Can you not just tell it to *dont do that*. Can you not just open the shader and tear out the perspective transform part?
b04b5a No.15590476
>>15590463
>What engine doesn't allow orthographic projection
but when Beelzebox started, it was using orthographic perspective exclusively. I added it after this suggestion >>>/agdg/29920, the video poseted shows ortho being switched on as well.
If you
b04b5a No.15590477
b04b5a No.15590484
>>15590477
>>15590476
fuck, VNC went awol and caused me to post prematurely.
>>15590463
if you can help me with tearing the shader up for it to work both in ortho and perspective views, I'd be obliged.
https://github.com/i-saint/Unity5Effects/tree/master/Assets/Ist/ScreenSpaceBoolean
As I always said, Unity shaders are goddamn black magic to me. (this also answers your question what engine this is)
ac4921 No.15590509
>>15590484
If I remember correctly, to go from orthographic to perpective you divide x y and z by w. If you look at the shaders, you'll generally see a divide by w at some point, and just adding an if there to do it when perspective mode and not do it when orthographic should fix it.
I see three instances of dividing by a "something.w" in StencilMask.shader, and two in Depth.shader. Very first thing I'd try is just changing those things, and if that doesn't work try just the two that are in both files or just the unique one is the stencilmask.
t. guy who has no idea about shaders either
aa5bff No.15590853
>>15590369
Sorry, I should have more carefully chosen my words:
This is why you don't use Unity (or any other closed source engine), kids.
d4f918 No.15591162
>>15590509
In shaders, the w divide is implicit, and not something you explivitly do through shader code.
This is because the w divide is used as part of the interpolation process to create the perspective effect.
Anyway, to create an orthogonal perspective you yourself divide z by w, leave x and y untouched (scaling does apply for the scene to fit your viewport!) and w is set to 1 so that the implicit w divide does nothing.
This can all be achieved by using a different projection matrix, which is why literally every engine supports it.
8665ac No.15591661
>>15590853
Mhm, salty post
553677 No.15591726
>>15587307
>>15586462
Dude, just draw the character in perspective and the environment in ortho (or the other way around, I am not paying attention on which is causing the problem).
Have two cameras, one that draws the character (and the lights) and the other camera draws only the environment, put one on top of the other, and voila. Issue solved in 2 days at most.
Making the character go behind stuff might be an issue, I am unsure on how the depth buffer would interact, if at all
b752eb No.15591801
>>15583937
This is hilarious. I would angle this as a comedy-action strategy-rpg
553677 No.15592513
>making a farming game
>only discover recently that carrots have flowers
64 triangles on this one
ffce55 No.15592528
>>15583937
>that ak melee
Top kek.
4c38a8 No.15593083
553677 No.15593095
>>15593083
And a single carrot can give you a fuckton of seeds
7cedd7 No.15593099
>>15592513
>carrots are flowering plants
well fuck. Potato plants have flowers and fruits too, but they're hideously poisonous.
553677 No.15593130
>>15593099
You have no idea how much i learned since i decided to start making this game
ea16ab No.15593144
>>15592513
When they flower like that they're usually no good to eat since they're probably woody.
553677 No.15593155
>>15593144
We have quality and many other attributes on food items, so harvesting them late will change their properties
8fd1ba No.15593716
These don't visually clash too much do they?
1955e3 No.15593748
>>15593716
Is that an actual level? It might look better in a more logical setting.
8fd1ba No.15593759
>>15593748
It's a bunch of textures drawn to the screen in the worst way possible
b88782 No.15593846
>>15593716
Those black empty squares need to be filled to avoid looking jarring, but the two textures play nice with one another. If you wanted a visual clash assessment maybe you should've had no black empty squares in there.
8fd1ba No.15593856
>>15593846
No that's good feedback, since the empty tiles are probably going to look like that. Well, I could put something that isn't distracting in there instead. We'll see how it goes
b88782 No.15593955
>>15593856
Are they walls or floor?
8fd1ba No.15593960
>>15593955
Ah, floor tiles. But the game is kind of a weird perspective. It's a clone of SubTerra / Chip's Challenge / Boulderdash, and some objects fall, some have a top-down perspective
b88782 No.15594036
>>15593960
You can always make a concrete tile. Why would they stay black? At least use some gray closer to the other tiles.
8fd1ba No.15594075
>>15594036
Well here's a mockup of what a random level might look like, as it currently stands
b88782 No.15594094
>>15594075
Oh, I see. Thought it was top down look all the way. Still, I'd change the black background. Even a dark grey would fit better.
1955e3 No.15594191
Posting a WIP version of the game I'm making for GodotWildJam2. Not much to do, just explore. Make sure you go in the doorway next to the big TV in the cottage. Try to break it as much as possible, you shouldn't be able to get out of bounds. Also, let me know how the downloads go. Mega is usually stable but I know sometimes people get bandwidth errors. Third mirror is on the way.
Windows 64bit:
mega.nz/#!v7wUWaaS!qLa5LM30Pt4UIeWy01_ul9Mk2UhvC5zRLHhgU3ob5Kw
files.catbox.moe/44757p.7z
Windows 32bit:
https://mega.nz/#!H2pmAKZB!6D1XWsAZ6MXKpVlW3aQoz-fIoRSqZHctW8GsHf5UMBA
files.catbox.moe/il6pr7.7z
Linux (untested):
mega.nz/#!jjgiwCDR!6R920Oe0VpnkJXw3Wvi3Rvcg6dNqT7TzlFHOwTzhBl0
files.catbox.moe/6llen3.7z
Controls
Keyboard
WASD move
E pickup/drop (hold to throw)
Space jump
Left shift run
Xbox controller
Left stick move
B pickup/drop (hold to throw)
A jump
RB run
dualshock
Left stick move
Circle pickup/drop (hold to throw)
Cross jump
R1 run
1955e3 No.15594207
>>15594191
Windows 32bit:
void.cat/feb59d4fc1a41beb9b43b1089bf6fc02cfa1c90f
Windows 64bit:
void.cat/#a8a5fbcd241b0d7f96c4cec244598d996ed9d3b9
Linux:
void.cat/#cbc2ecf2eb2a156cec076dcee674df0683bf0ec0
704fdf No.15594895
Okay so my project is currently composed of a mail .cpp file and then a bunch of .hpps that I'm including into the main file.
How do I progress towards having object files for each to speed up compilation?
Try to work on converting each .hpp to .cpp, then work on getting the individual files to produce object files, and then work on linkage?
7eeef3 No.15594935
>>15594895
files: main.cpp obj.cpp
cc -c main.cpp // compile main.cpp into main.o
cc -c obj.cpp // compile obj.cpp into obj.o
ld -o game.exe main.o obj.o // link main.o and obj.o into game.exe
This is the basic workflow. Replace cc and ld with the compiler and linker of your choice (e.g. g++ for both), and configure your makefile to not compile .o's of .cpp's that don't need recompilation. Google for examples on how you do that. And google how to make a makefile that's not shit.
Of course this is all irrelevant if you use an IDE, because IDE's do all of this for you.
704fdf No.15594993
>>15594935
I know how to write a makefile, I'm mostly just wondering about how to progress to the stage in which my makefile will work.
add85c No.15595030
>>15594895
I would do it one at a time. Separate the implementation out of one header, make sure it compiles, make sure it links, move to the next
0240f8 No.15595071
This was posted here once, does anyone one if this is being worked on still?
also sauce on the music in the video please
078e70 No.15595126
>>15595071
It's on hold iirc. Same dev is working on Die Tokenmask 2 (spelling?). Check the wiki. Music is a remix of Cheryl Lynn's "Got to be Real", don't know it exactly.
078e70 No.15595142
bedf8c No.15595611
Godot developer laying out the bait.
eb6bc4 No.15595618
>>15595611
When you're so afraid of backlash on the internet you end every sentence in a question mark.
1955e3 No.15595633
>>15595618
>interrogative sentence ends with question mark
>must be afraid
eb6bc4 No.15595649
>>15595633
If I talk like this then I never have to commit to anything I say? Because if I'm wrong then I can just say I was being rhetorical? Ummm, racist and sexist, much?
8a27dd No.15595668
>>15595611
>Massive homophobia
<When they just don't allow for LGBTWTFBBQ to change marriage laws
Pride parades are allowed however, unless President of major city disagrees.
>far right government
<When it's centre-right at best.
What a fag.
98b01d No.15595694
What's the most efficient way to load and render images in OpenGL C++? Read through numerous examples and guides but all seem damn complicated and I'm not certain if all they do is necessary. I'm trying to avoid libraries as best I can, though it seems unavoidable in this case, at least for reading the image format.
Speaking of that, are certain image file formats faster than others? Couldn't find a real answer to that anywhere. I tend to stick to PNGs in general.
I hate asking retard questions but it doesn't seem like there's any real comprehensive resource for questions like this.
>>15595668
These are the same people that think Trump and Kavanaugh are far-right nazis. It's a fucking circus.
1955e3 No.15595702
>>15595649
"What's not to like about Poland." is wrong, you ESL nigger.
8fd1ba No.15595724
>>15595633
>>15595618
>>15595649
Actually the question mark in this context denotes a certain tone. It's not that he's afraid of backlash, it's that putting a snarky comment with a question mark like that is a linguistic signal to be outraged; if he had a period or nothing at all, people would just roll their eyes and continue on
18b2ee No.15595726
add85c No.15595727
>>15595694
You load the texture data from an image, maybe assemble a texture atlas or something, and batch a bunch of quads (two triangles) that will be drawn with those textures. The efficiency is really about reducing draw calls as I understand it, which is why you batch the vertices together
8fd1ba No.15595741
>>15595727
>>15595694
Reducing draw calls is nice, but the big killer is texture swapping. Of course, it's easy to do both.
For example, even though it's about 120 objects being drawn in >>15593716, there's only 1 texture and 1 draw call, so it doesn't even register as 1% cpu usage on task manager
8665ac No.15595744
>>15595668
Anyone who's even slightly right to the extreme left is "far right". Including the classic liberals and moderates.
As proven by the hidden tribes study recently, that's greater than 90% of the voter base in the USA that's labeled "far right".
The extreme SJW left is a very loud minority, but a minority none the less.
As proven recently by the study, they're at about 8% of the voting population. Also, a funny note, they have extreme conformity in their base identifying characteristics: white, upper class… all other voter bases are more "diverse", even the "far right", which is the opposite of what they claim so vehemently; it's fucking laughable.
>>15595618
It's a loaded question, any response is a "good response", because it's just bait.
add85c No.15595768
>>15595741
Ah, good to know. So assembling the texture atlas is really the more important step
5c1e57 No.15595790
>>15595702
You do realize he was talking about the second guy and his "uh, THINGS I don't LIKE?" reply, right? Cause if not, you might want to watch your own webm a few times.
0240f8 No.15595830
>>15595142
Thank you kindly
8a27dd No.15595877
>>15595694
>These are the same people that think Trump and Kavanaugh are far-right nazis. It's a fucking circus.
>>15595744
>Anyone who's even slightly right to the extreme left is "far right". Including the classic liberals and moderates.
I forget about this, sometimes.
12e3be No.15595889
>>15595694
>Speaking of that, are certain image file formats faster than others? Couldn't find a real answer to that anywhere. I tend to stick to PNGs in general.
Theoretically, yes. All modern file formats on some level use compression and some algorithms that decode the compression may be faster than others. However the size of an image has a greater effect on it's "speed" than anything else.
What you want is the smallest image with the least loss in quality. Sometimes PNG's can provide that, sometimes they cannot.
PNG's are good for
>images with few total colors, such as a white background with black text, a logo, or a cartoony drawing. Jpeg artifacts are apparent in cases like this.
>a situation where the png image is smaller than the equivalent jpeg image. The included images demonstrate this, and it relates to the above. The Jpeg is at 90% quality, the png is at compression level 9.
>images where it's imperative you use lossless compression, such as with normal maps or pixel art
Jpegs are good for
>photographs's or photo-realistic images, such as the color texture of a material
If you want your game to look nice and not take up 55GB you'll need to carefully choose your file compression and compression levels. Jpeg is a very good compression algorithm in a very specific realm but it quickly falls apart outside of that realm.
8fd1ba No.15595917
>>15595889
png can also support color indexing, which makes it very suitable for pixel art, too
bedf8c No.15595928
>>15595694
>>15595889
Whatever you do, don't fucking store textures as jpegs because converting them to a lossy texture format will create more compression artefacts.
98b01d No.15595958
>>15595889
>>15595917
>>15595928
So PNG is a good bet, then, or should I go with .tga like >>15595726 suggested? Any quality or speed concerns with .tga?
18b2ee No.15595984
>>15595958
TGA is simple and uncompressed. It works fine for small/medium projects, but I would look into compressed formats if you find yourself with gigabytes of texture data.
12e3be No.15596075
>>15595928
Huh. I thought it was OK to store the albedo part of a texture as a jpeg provided it's at a high quality, high resolution texture, but I looked it up and it turns out in an engine like UE4 they already compress the files for you, so lossy to lossy just gives more artifacts. In addition to this it's data isn't organized in a way that makes it suitable for textures in the first place.
>>15595958
Apparently tga is the industry standard, at least for transferring files from software to the engine.
http://www.informit.com/articles/article.aspx?p=2819033&seqNum=5
Then when actually storing the images DDS is used by Unreal, which uses the proprietary S3 Texture Compression, allowing for hardware decompression.
https://en.wikipedia.org/wiki/DirectDraw_Surface
Here's a list on Wikipedia of these formats specifically designed for compressing textures. If you are worried about speed, find the texture formats that allow for hardware decompression.
https://en.wikipedia.org/wiki/Texture_compression#Availability
933ee6 No.15596076
>>15595611
Does that queer not realize that this is exactly why Poland is so great?
adead2 No.15596141
In Unity, I declared a variable 'public int myVar = 10'.
Now I changed my mind and I want 'public int myVar = 20', but the old value is stuck in the inspector and doesn't update. How do I force Unity to update it to the new hard coded value in all game objects that have this script as if it wasn't a public variable? If I remember correctly adding [HideInInspector] used make Unity update the value with recompilation, but it doesn't seem to work now.
12e3be No.15596142
>>15595958
I found this, software from AMD that can create DDS files. It is free to use for commercial purposes.
https://github.com/GPUOpen-Tools/Compressonator
Then this open-gl tutorial give sample code for loading DDS files.
http://www.opengl-tutorial.org/beginners-tutorials/tutorial-5-a-textured-cube/
I haven't used this tutorial so there may be stuff on tutorial 1-4 that you need as well to get the code there to work, I don't know.
So the workflow for compressed and fast 3d textures should be as follows.
>Create texture in photo editing or texture creation software
>Export as tga or png
>Run through the Compressonator and get back a DDS file
>Put the DDS file in the game directory
>Load the DDS file in code and use it as you please
adead2 No.15596166
>>15596141
Nevermind. That [HideInInpector] method still works, I was just retarded.
33c9db No.15596176
>>15592513
I've only seen wild carrots bloom, and their flowers smell exactly like carrots when stroked. It's really a cognitive challenge when you stroke a flower and it smells like the root.
553677 No.15596183
>>15596141
the default value is whatever the value is in the field declaration.
If it is currently 20, then any new instance of this script will have yourint as 20.
However, any old instance of the script will retain whatever value it used to have. It is done like that to avoid data loss (where something you didn't want to change is changed, or deleted).
To revert the value to the defaults, remove and add the script again. Maybe right click on the gear icon will have a reset option, I never tested this with custom scripts
To change the value on all instances of the script, you can delete the variable and then add it again but it might only work on currently open scenes
8665ac No.15596209
>>15596183
I almost replied with the same exact response.
Tho, they want a hardcoded value that's updated across all gameobjects that have the component.
Which means they just need to hide it in the inspector, or set it to private, or do a public property instead.
ef153d No.15596236
>Got offered a job running a gamedev project by company I left (partially) because I wanted to learn gamedev
It's weird how things turn out sometimes.
553677 No.15596238
>>15596209
or make it static, so that it does not ever get serialized at all and it's always up to date on every single object at all times
adead2 No.15596241
>>15596183
>>15596209
Just to add, [HideInInspector] seems to be the only method that always works because setting variable to private if it is used by other scripts which I forgot to add was the reason I had trouble with this will result in error during compilation, and it'll revert back to previous value when you make it public again.
8fd1ba No.15596257
>>15596238
Please don't clutter up the global namespace with statics like this.
553677 No.15596337
>>15596257
is that a problem on non-javascript languages?
8fd1ba No.15596358
>>15596337
Love it or hate it, good code in an OOP paradigm is organized and tucked away in a sensible location
553677 No.15596436
>>15596358
>good code
that's just like, your opinion man. Unless you come up with a objective argument, like in the case of javascript, you won't convince me
In case anyone is curious about why it's bad in javascript:
This simplified explanation don't go into details because I don't know them, there are plenty of gross oversimplifications
Every single fucking javascript object is a hashmap
The root object, also known as global namespace is, therefore, also a hashmap
The more entries in a hashmap, the slower is to access it's content, even if just a little bit slower
By polluting the global namespace, you make slower the acquisition of every single other global variable
REMINDER: The general idea of why it's bad in javascript can be understood through this explanation, but the finer details are probably wrong
OBS: A hashmap is a key-value pair collection, where you send the key and get the value. It have a look up table that uses the hash of the key. Each entry of the table is a list of all key-value pairs whose key have the same hash modulo, then it goes linearly through the list to find the first occurence of the key. Oversimplified: It's a array of arrays which cleverly finds the first index and then iterates to find the second one
9810ef No.15596464
>>15594191
>Silky-smooth 7.5 fps
>Snow flakes clip through everything
>Thought I found a hole
>It just goes to engine room
>Platforming when input delay is at its max
>Camera cannot pan, must move against walls in order to turn it.
>There is a glow that seeps out from the cottage. There is nothing there though.
>Snow flakes give out when you almost reach the end of the fence.
>Foot step sound from snow same even if indoors.
Well, at least the snow flakes aren't in the Engine Room.
12e3be No.15597030
I'm making a character controller and working on the movement. I would like feedback if any of this stuff should be changed or if any of it is too confusing.
>Look around with the mouse
>Walk with WASD
>Lean left/right with q/e
>Hold ctrl to creep, that is, move at a fraction of the normal speed at a given stance
>While holding ctrl, scroll to move to a higher or lower stance.
>>Three heights for three stances, giving 9 total stances. High, mid, and low for stand, crouch, and prone
>>The lower the stance the lower the speed, except for the "tip toe" stance which is a high stance with a low speed.
>>While prone the player cannot jump.
>>The lowest stance is low enough that it can be used to look under the crack of a door, but the player cannot move while in this stance.
>Jump with space
>Hold jump while being adjacent to and facing a ledge to grab on and mantle the ledge. The player will climb up the ledge, and at the top of the ledge mantle onto it.
>Hold ctrl to just grab onto the ledge but not automatically climb and mantle. If space is held and the character is climbing, then the character stops.
>>Use the up and down arrow keys to climb up and down the ledge. This automatically sets the stance to mid stand. This cannot be used to mantle.
>>Use the left and right arrow keys to move left and right along the ledge
>>Hold space and adjust the stance down to mantle the ledge at the current height above the ledge. Once the bottom of the character is near the level of the ledge the player will mantle it. This is like bringing your feet up to climb onto a ledge. If you don't hold space you will just adjust your stance but you won't mantle the ledge.
>>While grabbing the ledge the player can look away without falling off, something they can't do while mantling. However in this state they cannot climb up and down, and trying to move left/right makes them fall off the ledge.
>>>While looking away from the ledge being grabbed, press space to jump up and away from it.
8fd1ba No.15597092
>>15597030
>First paragraph
Why can't I rebind them?
b88782 No.15597198
>>15597152
Conveyor belts, right? Try a yellow/black pattern instead of red for the bands. Red looks like danger.
8fd1ba No.15597230
>>15597198
There's going to actually be a few different colors, that's just the one I happened to record. The ones that look like Factorio are floor tiles that move objects inside that very tile, while the ones in >>15597152 are conveyor blocks, which block movement, but move falling objects in the tile above it (such as rocks).
12e3be No.15597267
>>15597092
You will be able to. When I talk about things that need to be changed I'm referring to the mechanics not the keys bound to those mechanics. For example, is it really a good idea to stop the player's climbing when the "grab ledge" key, in this case ctrl, is pressed? It seems like a good idea but I want more feedback.
Sorry for not being more clear. I can see how the way I put it makes it seem like I was talking about the key-binds. But if you have better default bindings then I'd like to hear them. Also I forgot to mention that alt lets you free look while moving in a direction.
I've already implemented all of this, so I'll flesh out the test level and create a build of the game. That will result in better feedback anyway.
1955e3 No.15597552
>>15597030
I think I'm confused about the difference between mantling and climbing.
Using the same key for creep and ledge-grabbing would be weird. Why not space to ledge-grab, then space again while facing the ledge to mantle.
>arrows keys to climb the ledge
Do you mean WASD?
12e3be No.15597686
>>15597552
Yes I mean WASD. I don't expect Rachmaninoff will be playing my game any time soon.
>I think I'm confused about the difference between mantling and climbing.
There are three states. Grabbing, climbing, and mantling.
Grabbing is where you are holding onto the ledge, but not necessarily moving up or down. (ctrl)
Climbing is when you are grabbing a ledge and moving up or down. (ctrl + w/s or space w/o ctrl)
Mantling is when you finish climbing up a ledge and move forward to be on top of it. (space only)
>Why not space to ledge-grab, then space again while facing the ledge to mantle.
I've put that on my to-do list I'll try to get a method that works that way once I finish the test build. As it is now feels fine to me but that doesn't mean it will be easy for a newcomer to pick up.
The reason for ctrl doing what it does is so that you can hold ctrl while walking backwards off a ledge and the character will automatically transition from creeping backwards to climbing down the ledge without needing code to check for that specific case.
0d8a96 No.15597898
>>15592513
it's beautiful
>>15586129
what happened to his body?
12e3be No.15598392
Here's a Windows and Linux build of what I talked about here >>15597686 and here >>15597030
Tab free the mouse. Escape to quit.
Linux
https://files.catbox.moe/ie7cqs.zip
Windows 64-bit
https://files.catbox.moe/l9jg82.zip
I've included an edge case, the hole in the wall, to show where this system currently breaks down. I am working to figure out a fix. Other than that everything should work. Keybinding has not been added to the game just yet, this is testing default bindings. Give me feedback on anything you think I should bring my attention to.
1955e3 No.15598485
>>15598392
(Used W64 version)
Alt-tabbing and tabbing back in fucks the directional movement, the axes are locked to wherever the view last was when I tabbed out (vid related). I really like the peek under doors mechanic, that feels very natural. It's unclear when I'm crouched, standing, and prone, a HUD icon would help that. The highest stance slows me down, I guess the player's supposed to be on their tip toes. Grabbing a ledge, manually climbing to the top, then holding space and adjusting the stance is kind of awkward, maybe lose the stance adjustment and do it manually. Also, it's possible to go to the "peek under door" stance while climbing, which seems wrong. I was able to climb and mantle the tall wall with the hole in it but not the tall wall to the left of it (I guess this is the edge case problem). Overall a very cool movement system.
1955e3 No.15598486
>>15598485
>maybe lose the stance adjustment and do it manually.
I meant adjust the stance automatically when space is pressed.
8fd1ba No.15598501
>Duplicator
>Receives a falling object from above (such as rock or keys or gems)
>Consumes object, but spits out a copy on the sides if there's space
>This means you can receive 0, 1, or 2 copies of the item
Sorry these are just random tiles just placed everywhere, I don't really have much set up yet.
12e3be No.15598591
>>15598485
>>15598486
I thought about adding a HUD element as I was playing around with the system. It really is a problem not knowing what stance you are in without scrolling to the edges and counting as you scroll back. You are correct about the tip toes stance, which really, really needs a HUD element clearly communicating what's going on.
>Alt-tabbing
Good catch, I was able to replicate it and figure out the problem. The "Alt" key is bound to free look, that is, it locks the character's movement direction but lets you look around. When you alt tab you hold the alt key, then leave the game. When you alt tab back in it never catches you releasing the alt key, so it thinks you are still holding it. Pressing the alt key again gets you your directional movement back, but that's a workaround not a fix. At the moment I don't really know how I can fix this, but I'll see if I can find a solution.
>Adjust the stance automatically when space is pressed.
I don't know why I didn't think of that. It's a way better solution.
>it's possible to go to the "peek under door" stance while climbing.
I thought I fixed that bug. I'll need take another look I guess.
>I was able to climb and mantle the tall wall with the hole in it but not the tall wall to the left of it (I guess this is the edge case problem).
That is a problem I am aware of, caused by the wall being built of stacked blocks that fool the raycast, but it's not the edge case I was referring to. The hole is just big enough to climb into, but if you try to climb into it then some some bug will jitter the character around before settling inside, and sometimes you'll be shot through like a cannon.
>I really like the peek under doors mechanic, that feels very natural. Overall a very cool movement system.
Thanks man. I appreciate your input. It really helps.
8fd1ba No.15598630
>>15598501
Feels like I'm making Actual Progress™. Finally satisfied with my approach to my drawing code and how performant it is to more onto other things. Also drawing it to a render texture and just kinda putting it on screen.
1955e3 No.15598643
>>15598591
>Pressing the alt key again gets you your directional movement back, but that's a workaround not a fix. At the moment I don't really know how I can fix this, but I'll see if I can find a solution.
Instead of checking if alt is pressed or released, check if alt is held down with is_action_pressed("freelook"). Be sure to use Godot's action/input binding system, it's God-tier among engines. The new action strength system is great too; it reduces a lot of repeated code: https://godotengine.org/article/handling-axis-godot
>I thought I fixed that bug. I'll need take another look I guess.
Maybe disable going down to peek unless the player is on the floor.
1955e3 No.15598649
>>15598643
>check if alt is held down with is_action_pressed("freelook)
Hmmm just tried this with alt and it sticks as well. Could be a bug, Godot frankly has lots of problems with alt-tabbing fucking things up in the editor.
1955e3 No.15598679
>>15598649
Did some digging, it is a bug:
https://github.com/godotengine/godot/issues/18785
Workaround (place in a singleton):
func _notification(what):
if what == MainLoop.NOTIFICATION_WM_FOCUS_OUT: # window loses focus
for setting in ProjectSettings.get_property_list(): # loop through all actions and release them
if setting.name.substr(0,6) == "input/":
Input.action_release(setting.name.substr(6, setting.name.length()-6))
12e3be No.15598719
>>15598643
>The new action strength system is great too; it reduces a lot of repeated code
Oh sweet. I'll be using that. Reminds me of UE4's axis system, something they mention at the bottom.
>Maybe disable going down to peek unless the player is on the floor.
Yeah, that's the fix I ended up going with. Previously the scope was way too narrow to prevent it.
>>15598649
>>15598679
>it is a bug
Wait a second, I know C++. Godot is written in C++. It already has functions for recognizing when the window gains focus, and supposedly functions for checking input. I'd have to figure out how to run through the inputs and update which are still held and which aren't, but that can't be too hard.
I thought I used godot to escape engine dev, but it follows me like a curse.
241531 No.15598962
>>15595611
I really hate social media cause of shit like this. It's absurd these people feel the need to cram their bullshit into anything unprompted.
Reminder never to answer loaded questions or comments like those. Just ignore them and move on, engaging isn't worth it.
904cd0 No.15599014
>>15595958
Good luck writing a PNG loader, BMP is an easier option if you are on windows.
958247 No.15599286
Okay, I need some expert help here, mainly to figure out what's the best aproach to this. I'm sure someone has already made a better solution that this shit.
I have a system where any action the player does will put a message describing it on an action log. Pretty standard stuff really, that part's done.
The issue is that the text for each action should come from a source that has variants for the same action depending on a few parameters but also variants of variants so even with the same parameters you can get a different sentence for some variety.
To achieve this, I'm currently making several Lists that house the text for each specific situation, but I've noticed I'm already having a fuckton of Lists for just a few actions and this really isn't very good for scaling later on. So I'd like to know if there's a better way to handle text resources like this, to be invoked later.
Note 1: I'm not currently using wildcards to replace relevant parts of the text with names, pronouns, etc but I will in the future, this is still just a prototype. That part I don't see as being a chalenge at all.
Note 2: Later on, there will be part of the script that loads all the text from XML files or something similar so anyone can contribute text or customize their own.
c312c3 No.15599337
>>15599286
I think your two notes cover it, using a template and moving it out into a data file are the main improvements I could think of that would help scaling
8665ac No.15599745
got this neat fractal shot while debugging
c6c590 No.15599841
>>15596236
Congrats on not being a bridge burner anon. how the fuck do you pull that off? I’ve got to do it in a couple days.
>>15599286
Maybe make some implicit postfix trees for message construction?
958247 No.15599951
>>15599841
>postfix trees
First time I hear about that and it looks sexy as fuck.
However, I'm not sure it'd fit the bill. If the idea is simply to have an objective and quick description of an action, that's more than enough. But if the whole point is to add some prose to them (for the sick fucks that get off to that shit) then this would have a subpar result.
Then again, it might just be a matter of syntax and how big the operands are expected to be.
8e46ac No.15600139
bossman wants me to make a graph for the fps counter
i made a mesh with rectangles, but i'm thinking you could somehow cut down on the triangles, but i keep getting the same triangle count
am i wrong or is there really a way?
7eeef3 No.15600154
>>15600139
why are you optimizing an fps graph
also you could do it the logical way and instead of drawing bars, just draw a fucking line. optimally using a line strip
8e46ac No.15600160
>>15600154
he specifically wanted a mesh, i already suggested just using a texture
>why are you optimizing an fps graph
why not?
7eeef3 No.15600166
>>15600160
>why not?
because its a waste of fucking time that can be better spent making your game run better or, god forbid, actually make your game work in the first place
8e46ac No.15600177
>>15600166
i've been working for bossman for only 3 days, this is the simplest task he can really give me at the moment on a project that's already past the midpoint
5c1e57 No.15600250
>>15600177
>on a project that's already past the midpoint
Get on my level, my bossman's soon going to have me work on a game that's already finished.
Anyhow, assuming every bar is a single update I'd go with your first solution, where you can just add two triangles on the next update. Because having to recalculate the new combination of convex and concave shapes each time is both going to be hard and going to be slow as fuck for what you want. Even if you'd find a solution that uses less triangles, there's no way in hell that computing it every time is faster than just drawing the extra triangles.
b04b5a No.15600388
Funnt result of physics
>>15596236
must be a cool company
>>15597898
>what happened to his body?
it healed. But that's more of a proof of concept than anything
The zombies are gonna have fleshy bits ripped out tho.
8fd1ba No.15600400
>>15600139
If it has to be a bar graph, then first pis works best, because theres no guarantee of concave shapes with a fan. And besides, some things don't support fans.
Also like anon said you can just use a series of lines
1955e3 No.15600428
>>15598719
>I'd have to figure out how to run through the inputs and update which are still held and which aren't
That example code releases all actions when the window focus is lost, which should provide desirable results.
553677 No.15600792
All textured and with correct size proportions now
47951d No.15600986
Making a game where you capture mutated humans with chicken tendies to fight each other. Here is a color test of overworld. It's still in pre-production / pre-alpha.
8665ac No.15601089
>>15600792
iso view looks a-mazingly like anime when it's in thumbnail form, perspective looks good.
Honestly both look better in thumbnail form.
You want to experiment with some post-processing to capture that level of detail, or more anime-esque look.
If that's what you're going for then maybe some post-processing could assist to give that more "only the important details" anime look.
Maybe even a slight blur for your textures could help too… you can apply that in substance painter in a few moments.
Because honestly the detailed textures sort of clash with the mid-low poly anime-esque artstyle.
As, once I see those textures (enlarged picture, i.e. !thumbnail form), my brain expects similar levels of detail from the characters, and it gives me a bit of an uncanny valley feeling when it's the anime mid-low poly characters. As in, it doesn't quite jive, and feels inconsistent to me.
Just a suggestion tho
>>15600986
cute artstyle, it has character
8fd1ba No.15601091
>>15600986
Okay but replace them with Kobolds and you basically have the lewd audience and guaranteed shekels
8665ac No.15601117
>>15601091
>goblin slayer
>must kill all goblins
>kobold slayer
>must lay all kobolds
I swear I'm not a scalie, but those hips are cuhrazy
553677 No.15601135
>>15601089
We are going to play with post processing later, right now we are getting some basic stuff ready to make demos
8665ac No.15601192
>>15601135
Not exactly sure what u have envisioned in terms of artstyle so u do u.
Tho, I'd probably try certain things like blurring textures and so on first before post-processing
553677 No.15601210
>>15601192
I don't really know what artstyle we are going for, whatever look better will do
If you look closely the characters are using cel-shading and the clothes aren't, so we didn't even decide what are we going to do with it. First we gonna add some features, my bro started making the farming system and i am modeling veggies right now
b8da61 No.15601215
>>15601117
>I'm not a scalie, but
8665ac No.15601298
>>15601215
thick thighs save lives billy
>>15601210
understandable, artstyle is hard to pin down. it's one of those "I'll know it when I see it" type of deals
8d7fea No.15601305
>>15601091
>Okay but replace them with oppai lolis
ftfy
8fd1ba No.15601376
>>15601305
Before you do that, best wait and see how that CREEPER bill turns out
1955e3 No.15601379
>>15601376
I thought that was for sex dolls, not depictions. Also, wouldn't oppai loli be excluded anyway? No IRL child looks like an oppai loli.
8fd1ba No.15601385
>>15601379
Yeah but it depends on the precedent it will set. Even though it's technically completely unrelated, Jews will be Jews and find a way to continue to censor and debase things you enjoy
01c2b2 No.15601486
>>15596142
DDS is a really good option. The loader is pretty easy to write as well.
He could also use one of those runtime compression libraries that provide a virtual filesystem with a file. All the game files get compressed and it's more or less transparent. Then just use BMP or something easy.
01c2b2 No.15601498
>>15599286
Clearly you need to implement or use a natural language parsing library.
Just thinking to myself, localizing a game like this would be sadistic.
01c2b2 No.15601509
>>15600139
do it all in a pixel shader. sphere-traced lines. No polygons! (OK ok, 2 to fill the FPS draw area) MAXIMUM polycount optimization!
4bafb3 No.15602653
Do people talk about modding in here? Anybody make mods for Dungeons of Dredmor?
I'm looking to make a Divine Cybermancy skill tree, along with passive buffs if possible. Suggestions are welcome
12e3be No.15602684
>>15602653
Yeah you can talk about modding. There's a section in the wiki that lists some agdg modding projects.
http://8agdg.wikidot.com/general:vmprojects
I'm interested in the Divine Cybermancy mod, though I don't really have any suggestions.
4bafb3 No.15602706
>>15602684
Unfortunately I haven't played Dredmor in a looong ass time, and to make a mod kind of requires an understanding of game mechanics
This is what I've got so far, mind you, whether or not these mechanics would be possible I can't say!
Divine Cybermancy mod for Dungeons of Dredmore
Permanent bonuses
You gain brouzouf
3% bonus to all money received
6% chance to proc status message 'You gain Brouzouf'
Cyber-Necromancy
+3% resistance to all effects/debuffs
Skills
My legs are OK…
nimbleness buff, removes
Deja Vu
Recast all abilities from the previous 3 cycles of guilt(turns)
Don't make waves
Possible stealth related ability?
Jian't Pain
Summons a Jian to fight enemies, will turn hostile when all enemies are dead - awards no XP
Substitution Gate
Lifesteal or however it works in Cybermancy
Triangular Gate
Final skill, causes a fuckton of damage and debuffs, has a chance to fuck the player up in a few 'interesting' ways
12e3be No.15602773
>>15602706
Seems interesting. Do you have plans to add weapons from Divine Cybermancy?
I should say that I've played EYE quite a bit, but I have no experience in Dungeons of Dredmore.
4bafb3 No.15602800
>>15602773
Eventually I would like to add in some of the weapons.
Most weapons need to be crafted during a run, so that does pose a challenge if I'm to emulate the research system of EYE. I'd love to have the smg that shoots faster than the minigun, bearkiller, and the nuclear powered hammer.
4bafb3 No.15602808
>>15602800
I imagine my first hurdle once I get some basic mechanics coded is actually having it be BALANCED and actually fun to use in your character's loadout. Something very hard to do for a novice Dredmor player!
4bafb3 No.15602811
>>15602808
coded being used in the loosest form possible as all modding in this game is done with xml files
8fbb00 No.15602963
Finished water handling. I slowed down the drain rate for this webm.
e31ac2 No.15602997
>>15602963
noooic
what happens when you drop it?
4bafb3 No.15603001
Lesson number 1: backup wiki's when you have the opportunity
The wiki for Dredmor died years ago. Fortunately archive.org has some wikidumps.
The only offline wikidump reader that seems to actually be in a working, maintained order is https://www.yunqa.de/delphi/products/wikitaxi/index
You'll likely need to use the new importer. Also, images don't get imported to the created db and I don't feel like self-hosting a wiki myself right now (shouldn't have neglected learning docker)
8fbb00 No.15603008
>>15602997
The cup has a rigid body so it would just roll
e31ac2 No.15603019
>>15603008
but the water, does it all spill out instantly and end up with a puddle?
8fbb00 No.15603028
>>15603019
No puddle. I wasn't planning on making a puddle. Maybe later?
b04b5a No.15603031
Need opinion on the look of the zomboids
conna call 'em Flayed Corpses
a5b5b3 No.15603036
>>15603031
I'd say they'd be much more spooky if they didnt have the normal eyes.
8fbb00 No.15603039
>>15603031
Maybe they are too shiny. Try tinting the specularity some subtle violet or red. Or maybe give them a subtle red rim light.
4bafb3 No.15603046
>>15603031
any way to have part of their flesh/guts hanging off/away from the body? that would sell it i think
79385c No.15603053
b04b5a No.15603062
>>15603036
I am planning on adding uncanny eyeballs to them, I'll see how they will turn out when the skull is carved
>>15603039
Here's a screenshot with proper lighting. Maybe they are a bit too shiny.
The second screenshot has some violet tint and emission added. GGotta edit the shader for a rim light.
>>15603046
There will be intestines hanging out. When you hit them, just like shelle bones, organs will fly out. I'm hoping to get the eyeballs to hang from the skull too.
8fbb00 No.15603087
>>15603062
It just looks really shiny. For some reason now the first pic looks better. Especially the dude on the back with the light.
b04b5a No.15603097
>>15603087
how about less smoothness, more normal map strength?
553677 No.15603571
>>15603097
Shiny is better, go back
770480 No.15603677
>>15603097
Jesus yeah keep it shiny. How about removing the eyebrows?
169f52 No.15603706
I'm working on a model for the MC of my project.
This is the first time I've ever done human modeling
Critique?
Something's wrong with the hand but I can't tell exactly what.
a0109d No.15603745
>>15603706
>Bobo70Oz
>70Oz
Are you a fine connoisseur of 40oz malt liquor too?
73f332 No.15603986
>>15603706
>This is the first time I've ever done human modeling
1) honestly i think you are lying anon
2) i know it looks better and all but it is easier to see topology in hard not soft edges
but to flatly answer your question i think the thumb is a bit off from a "proper" rest position. along with the rest of the fingers. but don't worry the fingers are easier to rig that way but the thumb "base?" needs to be brought down/ in and bend the palm not to be flat. or the whole hand needs to be flat.
553677 No.15604411
>>15603706
Proportion wise it is good for a cartoony style
Show the topology, here a reference if you need one
169f52 No.15604537
>>15603986
>1) honestly i think you are lying anon
You better believe it man, this is bizarre territory for me.
>>15604411
pic related is shaded flat
553677 No.15604553
>>15604537
Change to edit mode so we can see the wires, by the look of that shoulder joint it doesn't looks like it will deform well
Also you should try doing it in T-pose
adead2 No.15604688
>>15603097
Maybe monsters should be devoid of cartoon-human features? Pic is what I mean.
Skellys and meat look pretty good, the problem is those eyes and eyebrows, they just make them look too friendly.
4bafb3 No.15604748
>>15603745
>steeleworldwide poster
553677 No.15605522
Potato and Carrot with flowers and without, there are 2 extra stages, buds and seeds
a5d62a No.15606158
>>15603031
>>15603062
>>15603097
What happened to the fleshy face?
6ce63c No.15606317
>>15604411
What does it mean by "loops"?
All I see here is quads. I see no loops.
bedf8c No.15606339
>>15606317
He means the lööps, brööther
6ce63c No.15606447
>>15606339
epic instagram meme dude
08a0f6 No.15606485
Shmup anon here.
Did generalize my memory layout with a function, also noticed I made a mistake previously when manually declaring memory usage but it didn't crash since I was using more space than needed.
Made enemy aim toward player, I can add some randomness or prepare pattern if needed.
Fixed player input, I had a problem when rapidly switching direction but now it's fine.
Somewhere I realized I'm also evaluating stuff every frame call, and understood why japs also did this. Simpler to code, and afterward you can arrange numbers that depends on it when balacing gameplay. I will need to arrange that later.
Then I wanted to make lasers and thought hard about instanciation, but since I want a "slow laser" that you see expanding fast and possibly curve I realized that I was retarded and needed something else.
Searching around I found some purely fragment shader laser, but I would need to have some quad covering the screen, or go with a separate framebuffer. But I didn't like this idea much since it forego the geometry and would "bleed" over it in some case, ignoring any possible occlusion.
So it's related with generating vertices, and apparently a polyboard would suit nicely for what I want.
Current roadblock: polyboard for laser.
bedf8c No.15606494
553677 No.15606715
>>15606317
Are you colorblind anon? The loops are colored
958247 No.15607605
I had an idea for a game, but I only got the gameplay down, and need a framing device for it.
I'm shit at modelling, so a couple months ago I started working on a cyberspace game, because I like cyberspace and even with Cyberpunk 2077 coming, no one seems to be doing cyberspace games.
BUT: at the time I was also thinking about making a game that begins on the same premise Sim Medieval had (single village where all heroes come from) and take them on adventures, return to shag the queen, all that classic ye-old kingdom tropes. But again, I'm shit at modelling and using pre-made assets from some store makes the game look like crap so I put down that idea until I can pay a decent modeler.
Then, in a feverish dream, I had the idea of combining both. Cyber-Village where programs are the villagers, the houses/castles are processes and the monsters are viruses.
The player could be an hacker inserting himself to do something, it could be a scientist that started the whole thing as a simulation to study it, it could be a self-aware program, it could be… well, it could be a lot of things and that's where I lost my train of thought.
All I can imagine is some Tron-looking Guards come to request your audience to the King, and when you enter his chamber, he looks just like Master Program (but with a Crown).
He gives you a Task (quest) that shows up in your HUD under a Task Manager (complete with % of resources assigned to it and memory taken) to search for an evil cultist that infiltrated the town (trojan) and is murdering the villagers (crashing them). I even got a good chunk of assets to start with since they're basically made of primitive shapes.
It's easy to frame magic as part of the System, with "The System" being analog to "The Art" in D&D. Magic portals = Hyperlinks, fireballs = logic bombs.
The list goes on and I even went searching to see if anyone though about something like this but found absolutely nothing.
But I can't for the life of me come up with a decent framing story for this. Who is the player, what is he doing there? What is the goal of the game? Standard RPG with a story? Comical or serious? Should I involve Real-Life people that drop in or not?
I guess I just discovered that my writing talent is on par with my modelling talent too.
5c1e57 No.15607652
>>15607605
<I only got the gameplay down, can't think of story
>entire post is storyshit/events/setting, with zero actual gameplay aside from "there's a magic system"
958247 No.15607748
>>15607652
I posted storyshit because that's what I'm struggling with.
The gameplay was much easier to come up with since it's standard RPG stuff with a twist (since it's cyberspace). You spawn a character and have three diferent fields you can improve: EQUIPMENT / STATS / ABILITIES.
Standard RPG stuff. The twist is that they only look like it but work in diferent ways:
>EQUIPMENT is given to you at set points (or found in set locations), there's no shops
>Equipment is not numerous, but drastically changes how your ABILITIES work.
>After unlocking a piece of Equipment, you can change into it at will (might limit this to Retrofit stations or some other shit)
>Equipment also symbolizes how powerfull your program is to the world, this changes "social" events and increases threats sent to you if you equip a lot of powerfull stuff
>STATS are the usual STR/DEX stuff, but they start at 1.
>you find hidden data/information in the world, essentially pieces of lore
>you can slot this pieces in a limited memory pool to increase your stats
>some pieces are smaller, 1 Mem = 1 DEX, others are bigger and more complex, 7 MEM = 3 STR, 2 DEX, -1 INT
>hidden pieces open new hidden stats that are not normally available without them
>SKILLS are every action you can take to interact with other programs/processes
>Nearly all EQUIPMENT gives you acess to one or more SKILLS
>Some skills are learnt and you have acess to them regarless of equipment
>there's multiple versions of the same skill, that perform diferent things
>SKILLS are "fired" into a sort of ALU stack: while there they take time to resolve before the stack moves on to the next skill
>on battles or challenging puzzles, you'll need to fire 6 or 7 diferent skills in the correct order, mistakes will cost you
>your STATS determine how fast the ALU stack resolves each skill and their sucess rate
>typical fight: you stack ATACK->DEFEND->ATACK->COUNTER->SPELL and the stack will resolve it
>typical puzzle would be something like: SCAN->SLICE->SCAN->MOVETO->PULL->DISCONNECT
>normally, every enemy has a certain "loop" of actions it performs, same goes for puzzles
>recognizing and memorizing this loop is key to win, the other half is looking at your SKILLS and STATS and coming up with a counter to their loop
>harder fights/enemies will start surprising the player by killing processes in their own stack to throw your plans off-timing
>the player can also respond by killing SKILL commands in his own ALU to re-shape the stack and respond to it
958247 No.15607749
>>15607748
The most important aspect is that everything, be it an NPC, his house, the Statue in the Courtyard or a tree is built with these 3 systems. Everything has one or more pieces of EQUIPMENT, several STATS and one or more SKILLS.
There's no HP or MP, to destroy something, you must perform 3 steps:
>bypass the ARMOR
This is a value that increases with EQUIPMENT, representing how hard it is to gain acess to the inner code of that creature. Basically, how strong it's encryption is. SCANNING is important in this step. Without scanning, your hitting random equipment parts, but scanning lets you pick which part of EQUIPMENT you want to use. Focusing on a weaker part will lower his overall ARMOR increasing damage to ther rest of the parts. Damaged EQUIPMENT might disable the use of skills, lower stats or otherwise hinder. This happens to the enemy and to you.
>Override the PROCESS
The PROCESS is kinda like the Armor but derives itself from the stats. The highest stat will also change the type of PROCESS (High DEX = Runner, High STR = Brute) and diferent types will require that you have a certain configuration of stats higher than his to win this process. For instance, a high DEX enemy (Runner) would require you to pummel him with lots of fast hitting, small attacks, while the heavier attacks that normally cause AoE's would leave him unscathed. Every hit temporarily damages the defenders status (this happens to you too).
>Deploy a PAYLOAD
At this point, you already won, so you fire a specific skill to finish the enemy. On a normal RPG, the fighter would deliver a "Crushing Blow", the Ranger would "Skin for meat and skins" and a Necromancer would "Rise as a Skeleton Minion".
These payload skills are also things you need to find/learn and without them, you can't actually crash a program: you can beat on it until all his SKILLS are disabled and his STATS are 0, but he'll never die completely and might regenerate later.
Kinda like on RPG's you have the "Dying State" so people don't die as easily and clerics have a chance to help them.
It's a simple loop of:
>find new EQUIPMENT/STAT/SKILL
<slot it in and figure out a new combat LOOP with it
>find programs to beat up
>find puzzles to solve
>talk to friendly programs
>repeat
The problem is that I have no story or goal for this. It's a simple loop that should keep the game fun for some 6-8 hours (not planning on making it much longer) but without an actual goal, or a reason for things to exist, it's just a weird-looking crappy hack&slash with puzzle-like gameplay.
958247 No.15607763
>>15607749
One last thing:
>the Ranger would "Skin for meat and skins"
This is basically the way the player would adquire 80% of his EQUIPMENT/STATS / SKILLS.
Beating up a program and derezzing it actually destroys it, but before doing so, you can analyze it and retrive wathever EQUIPMENT/STATS/SKILLS were not destroyed/disabled in the fighting.
That's where the dificult part of the game comes in: you can obtain really good gear by finding an enemy to is currently using it. The downside is that you'll have to fight him without disabling or destroying that gear, which tyipcally means he'll be a lot more dangerous if you could go all out.
The only plot hook I had for an end goal would be having a BIG BAD EVIL that was running a very specific EQUIPMENT/STATS/SKILLS on his memory and you had to fight him without damaging the parts you want to save. After the fight, you rebuild yourself using the gear your took from him. You'd need some equipment to survive some apocalyptical event, some STATS to find out some story-related "truth" or use a specific set of SKILLS to stop something from happen, so basically opening the idea of having multiple endings based on what you salvage from him. Bad ending if you can't salvage anything.
3397ef No.15608624
I need some feedback on these character sprites for my game. The first two I'm mostly okay with but think could probably be better, the third seems really weird to me because of the differing shield angles/placements. From the front it looks like the shield is being held in front the whole time, while from the side the shield is only in front for one step then at the side the rest of them.
8665ac No.15608669
>>15608624
>variable pixel size
>low-res pixel shit to the max lvl
trigger me timbers
3397ef No.15608676
>>15608669
I took those caps with the snipping tool, sorry. After I get a rough draft of the 'full'sheet I'll post it.
8665ac No.15608708
>>15608676
i dislike most pixel art unless it has character anyways.
the current stuff u have reeks of generic-ness from early gen titles so it gains none of that usual leniency.
f.e. RoR is super low-res but has character.
Low-res can be done, but it has to have character. Or at least that's how it is for me.
3397ef No.15608731
>>15608708
That's absolutely meaningless in terms of how I can improve the sprites though. I'm trying to go for a gameboy style so it's kinda supposed to seem early gen. Bear in mind this is the first time I've ever tried doing something like this.
98b01d No.15608761
>>15608624
>Mixels
Die and go to hell.
Also, the level of detail varies so much that it's ridiculous. First guy, the hair has three shades of hair, but one shade on the shirt and pants. Second girl, two for hair, one for full outfit. Third girl, three for hair, two for outfit, One for both colors of the shield.
The big problem, though, is that it has no real style. It just looks like bootleg muh retro Final Fantasy NPCs. Try doing something original, with the color or the styling of it.
>>15608676
That's not what he meant, mate. Think of it like this, a 2x2 grid in your images is one pixel, since most pixels adhere to that. However, here are the spots you fucked up that I noticed at a glance.
>>15608731
>Gameboy
Learn about console limitations first. I'm not too familiar with the GBC but the NES allowed for 4 palettes of 4 colors each (3 custom + transparency) and that palette had to apply to everything on the sprite layer in the room. Backgrounds as well had 4 palettes of 4 colors (3 colors + 1 background color shared amongst all palettes). Background tiles were 8x8, but each 16x16 region had to share the same palette. For sprites, you could not have more than 8 (meaning more than 8 8x8 tiles) on the same scanline without flickering and lag. Mega Man is a good game to study. For most enemies and bosses, they used background layer tiles, moving the background in one direction or another, and anything really animated was done with sprites. And to get more colors out of Mega Man himself, the face is separate from the rest of his body and uses a separate sprite. It's why when the game lags, his face disappears sometimes. I don't know enough about GBC, so do your research.
Unless you're talking about GBA, in which case, fucking hang yourself and look at some actual GBA games.
8665ac No.15608764
>>15608731
well I'm just giving u advice that that style is overdone and generic.
Art is the first impression for people who happen to come across your game.
If the game looks like generic garbage it's generally judged as such.
>I'm trying to go for a gameboy style so it's kinda supposed to seem early gen.
Ah, good luck then, there are some purist jrpg people who like that style but the vast majority will stick with the above.
It also reeks of rpgmaker tier garbage
>That's absolutely meaningless in terms of how I can improve the sprites though.
How is, "that style is overdone" not direct advice?
Obviously it's not how you can improve those sprites, because I'm saying that style is overdone; they belong in the garbage or in the 90s or early 2000s.
>Bear in mind this is the first time I've ever tried doing something like this.
I'd look at the pixel logic art guide on the agdg wiki's graphics page.
Look for different styles, and even if you want to get inspired from the early gen pixel art; at least put some effort into injecting your own creativity into ffs.
3397ef No.15608975
>>15608761
I don't know what 'Mixels' means. Comments on detail are something I can actually use, thanks for that.
>GBA
Why the fuck would I mean the gba
>>15608764
Maybe I'm just autistic, but I'm getting caught up on how the problem seems to be "muh retro" when it's supposed to look like an early GB/GBC game. Not necessarily color-wise, though.
This is the only other thing I have you might consider 'original', but even I think it's shit. I'd had an idea of color-coding each area based on the major element used there kinda like how Pokemon Yellow has color-coded towns.
5c1e57 No.15609003
>>15608975
These are mixels, now fucking stop it.
>why the fuck would I mean GBA
Because if you meant the original gameboy or the GBC, you'd be using 4 colors total for the entire screen and would have to scale down your sprites considerably.
3397ef No.15609022
>>15609003
I still barely know anything about spritework but I'll pay more attention to that. I don't really have to scale anything down though, these were just really zoomed in so I could work on it. Zooming with a touchpad is bullshit, wouldn't recommend it.
8fd1ba No.15609036
>>15609022
Look at the sprites from Pokemon Crystal. Each one has black, white, and two other colors. Every single one. They somehow made it work wonderfully.
Also fuck you for making me go to Eevee's shit website to try and find them as an example
4c2f83 No.15609041
>>15608731
>mixels
>those many colors when GBC sprites have like 3 colors each
>posting with the grid on when asking how good the sprites look
>whatever the fuck you're doing with the hair
Design the sprites at 1:1 then blow them up later you retard.
I don't know the exact palettes and limitations for GB/GBC so look those up and actual games like Mario Tennis/Dragon Warrior/Pokemon Crystal for reference.
3397ef No.15609063
>>15609036
Do you mean bulbapedia or something, because I haven't gone there since the staff started their endless Trump bitching. Crystal seems like a good reference.
I hadn't really planned on limiting the number of colors I used but I guess I'll try it.
5c1e57 No.15609086
>>15609022
Gameboy has a maximum sprite width of 8 pixels, yours are 12 to 13 "pixels" wide when you don't fuck up with the mixels.
You claim you want gameboy style sprites, but you don't use the size limitations nor the color limitations, probably would put more than 40 sprites on the screen at one time, and likely won't use a 160x144 resolution. AKA you don't use a single GB limitation. Why don't you add some soft-shadows to it to make the pixelshit package complete, because you're doing exactly the stuff every faggot hipster indie puts out that causes everyone to hate pixel "art".
98b01d No.15609091
>>15608975
>I don't know what 'Mixels' means.
Mixels are mixed sizes of pixels, like what I circled everywhere. Rixels are when you take scaled up pixels like what you're using and rotate them in engine, causing blurred, JPEG-esque artifacts, or if you disable that, it still creates mixels.. Remember those two terms, and remember that if your game uses either, you deserve the fucking noose. Pic related. You're not using Rixels yet, so just keep it in mind.
>Why the fuck would I mean the gba
Because you're not adhering to GBC limitations by any means.
Read this.
https://forums.nesdev.com/viewtopic.php?f=20&t=17319
>There are eight four-color background palettes and eight three-color sprite palettes. GBC is the only tiled PPU I know of where all four colors are distinct across the different background palettes. On all others (NES, TG16, Genesis, Super NES, GBA), all tiles share the same color 0, but not on GBC.
>It has 12288 bytes of CHR RAM, enough for 768 tiles. That's 256 for sprites, 256 shared between sprites and background, and 256 for background.
>Up to 40 sprites can fit in OAM. They are all 8x8 or all 8x16, as on NES. But 10 can fit across the screen, covering up to 50% of a 160-pixel-wide scanline, compared to 8 on the NES, covering 25% of a 256-pixel-wide scanline.
The GBC is actually quite advanced, even more than the NES, but you need to respect the 8x8 grid and build a fucking shrine to it. Unless you wanna go full Shovel Knight and say "ehh, fuck this authenticity bullshit, who cares about that?
There's more detailed shit in there, but since you're having a hard time understanding everything so far, you should first learn the simple stuff.
In short, 8x8 or 8x16 grids (but you have to make all of them one or the other, can't pick and choose), 4 color background tiles, 3 color sprite tiles, 8 palettes of each can be used in one area. Unlike the NES, color palettes apply to 8x8 grids, so you're lucky there. 40 sprite tiles per area, so be sparing with what you're animating.
As for palette, it's the same as the SNES palette, 15 big RGB. Second pic related. Those are the colors you can choose from but other limitations apply.
3397ef No.15609170
>>15609091
I don't really plan on rotating anything so rixels shouldn't ever show up. Though I imagine that and mixels could in theory be put to good use on something that's supposed to look like it doesn't belong in the game.
Would something like LoZ Oracle games be good for reference, because just from a glance those and Pokemon Crystal don't seem like they use 8x8, it's probably 8x16.
8fd1ba No.15609234
>>15609091
I don't see a problem with the one on the right. Yes, it's not necessarily true to the original, but it looks better. As long as the base spritework is correct, I figure any sort of camera tricks is just fine.
98b01d No.15609294
>>15609170
>Though I imagine that and mixels could in theory be put to good use on something that's supposed to look like it doesn't belong in the game
Either an Undertalefag or a fan of those Sonic.exe-tier stories. Either way, no. Let your game stand on its own legs and make a really impressive looking sprite that adheres to the limitations. That's far more impressive than just OH LOL HERE'S THIS MODERN LOOKING ART IN MY NES GAME.
Check this out and learn.
https://www.youtube.com/watch?v=NMrlVcgR240
>Would something like LoZ Oracle games be good for reference, because just from a glance those and Pokemon Crystal don't seem like they use 8x8, it's probably 8x16.
Realize that you're intended to put 8x8 tiles together to form one character. Think about Mario in SMB3. Pic related, I broke him down into 8x8 tiles. You need to calculate that shit carefully.
>>15609234
The point is that the base sprite work is NOT okay when you do that. It's scaled up several times in engine, and that's fine, you're not going to use a 144p window on a 1080p monitor, but once rotated, there are smaller pixels in between the larger ones. It would only be possible on a 1080p monitor and modern engine. That, and rotation effects like that are not even close to being possible on proper old hardware.
If that's what you're going for, okay, but then do not dare even suggest that you're going for a retro style if you're going to break every rule like that. What's next, you're gonna shove shaders and bloom in an NES game and say that's authentic? Gonna add modern lighting to PS1-style models like hipsters do in their nu-poly styles? I bet you think Shovel Knight looks okay too. Fuck off.
3397ef No.15609343
>>15609294
I was thinking something like the Ultra Beasts from Pokemon SM but sure, whatever. The 8x8 boxes are a method to keep something looking 8-bit even though it's made of more than 64 pixels, right?
98b01d No.15609382
>>15609343
Not exactly a method, it's just how these consoles render their graphics. You can make each tile on a GBC have a different palette of three colors, that's why these grids are important. You can only have three colors per tile.
b551d7 No.15609390
I can't get myself to finish because the fun parts are done and only the tedious parts are left, implementing hex grids is bad, with facing it's worse, having movement controls be text based is the worst
>move to x y
>face dir
>repeat X times
>do action
But then again without it it won't be much of a tactical game, Is there a better way to handle distance and facing?
553677 No.15609687
Anyone here have experience merging skinned meshes in Unity?
I am having problems with the blend shapes, I get this useless error Bone weights do not match bones. which don't tell me how shit is wrong.
The mesh works perfectly until I start messing with the blend shapes
b88782 No.15609716
>>15609091
>Rixels
>Mixels
Cool, I didn't know there were terminologies for these things. Rixels piss me off so much. Low res graphics require a low res canvas to justify themselves. Rixels destroy the illusion and look downright dishonest.
64f7e7 No.15609766
>>15609170
>Though I imagine that and mixels could in theory be put to good use on something that's supposed to look like it doesn't belong in the game.
Don't have things that look like they wouldn't belong in your game. The reason people so quick and easily point out mixels is that they do so clearly stick out, and you realize that they shouldn't be in the game. Everything on the screen will look so much better in general if the style is consistent as opposed to having variably-sized pixels here and there. You might think you can apply them somewhere to some effect, but don't. It will look like a flash cartoon from 2003 where they just downloaded a bunch of sprites from various games.
>>15609063
>I hadn't really planned on limiting the number of colors I used but I guess I'll try it.
A good side note on limiting your colors is that these sorts of graphics are supposed to set off the imagination. If you look at the first picture here >>15609041 and imagine that person standing in front of you, your imagination would fill in the details. That helps a lot when you're viewing such low resolution pixel art. You surely don't have to limit yourself to just two colors on a character, but if you cut down on your color usage, you'll find yourself making your details in more creative ways that look better in this format.
If you're ever having trouble eliminating some of your mixels, post a picture here and ask for ways to eliminate them in your sprite. People here are more than happy to help guide someone who is trying to become better.
>>15609716
I never knew it either, I always called mixels "faggot sticks", because I usually see them when people are trying to cram lines where they wouldn't fit, or make a detail stick out where it shouldn't.
98b01d No.15609778
>>15609766
>faggot sticks
>bundle of sticks is a faggot
>a faggot is a faggot
>faggot faggots
c98121 No.15609828
>>15609382
And that's why all the overworld sprites for Crystal use black and two colors, so they can all (or at least the human characters) fit in one 8x8 box? Or do even those use more than one box? Like I said, they seem to be 8x16 at a glance. And if for whatever reason an 8x8 sprite had to be resized into 16x16 would it be enough to just 'double' the size of the pixels?
>>15609766
I think that kind of effect would work better if it was a higher res object appearing in a generally lower res game, like Sigma's final form and that one sword miniboss from X2.
Iirc Pokemon RBG used 4 colors for battle sprites, though black and white might be counted in those 4 colors.
If it isn't obvious yet I'm trying to make an rpg and the shit I posted earlier were supposed to be map sprites
7fd49c No.15609915
>>15609893
#include <iostream>
int main() {
for(;;) {
for(int a = -10; a == 10; ++a) {
if(a < 0) {
int c = -a;
}
for(int b = 0; c < b; ++b) {
cout << ' ';
}
std::cout << "test\n";
}
}
}
That code is messed up. What are you trying to do? Also, c is not going to be visible outside of the scope of that if. That wouldn't even compile. Do you expect the for to be inside of that if?
b88782 No.15609925
>>15609893
First of all, the while(69) will always be true, so it'll never stop.
Second, your first for() is fucked. a starts being -10, and the loop will continue for as long as a is 10? You might want to change that middle part from a==10 to a<10. The it'll go until a equals 10.
I'm not reading beyond that, so you're on your own.
7fd49c No.15609959
>>15609943
C++ isn't like other languages. When a variable is used, it needs to be declared ahead of time. There's no data structure storing the variables. Think of the code structure as a tree. The declaration needs to be equal to or above where it is used.
ae8f93 No.15610025
>>15609959
You recall that you actually declare it afterwards in C. I forgot how though.
4c2f83 No.15610066
>>15609828
Characters are made of multiple sprites. Essentially every GBC RPG uses one palette per sprite character, so do that if you want to stay in style.
You could double the size but it wouldn't look very good. Just don't use smaller pixels or you'll break your imaginary fixed resolution.
SNES did have a high res mode but it was mostly used for menus and text, and I think those had to be displayed in a separate low-color layer. X2 and X3 use a special chip to display wire frames but that still had to fit in SNES resolution.
If 7th Stand User is anything to go by, you don't have to autistically adhere to hardware limitations to have a certain look. Just make gud gaem and don't be retarded.
98b01d No.15610069
>>15609828
Actually, the Pokemon sprites you're referring to (I assume it's these ones in particular) are 16x16. Whether they were drawn in 8x8 or 8x16, I can't say. They could have more colors, but they don't. That's not a hardware issue, though. If I had to guess, maybe they were using 8x16 tiles, which meant they didn't have many opportunities to make sprites look better by using more than one palette. It could also be a storage limitation.
Check out pictures 2 and 3 now. I'm not a fan of the game but it's good for this showcase. Sprites only allow for three colors per palette. This sprite I picked is actually not too efficient, has two tiles that are just the outline, but it might be different in animation, I didn't check.
Now, look at the makeup of the sprite. In total, it has six colors: black, skin, light blue, yellow, purple and red. I tried to break it down into how I assume the devs did it, into an efficient manner. This limits it to three palettes (remember, sprites can have 8 3-color palettes on GBC per room, so that would leave us with five):
Black-Red-Skin,
Black-LightBlue-Skin,
Black-Yellow-Purple.
What I realized by examining the sprite is that the hair is drawn separately from the rest of the sprite, much like Mega Man's face being separate from his body.
Pic 3. You can see that Frame 2, the body is the same as Frame 1, but the hair isn't. And for Frame 3, the hair is the same as Frame 2, but the body isn't. It's why the animations are so smooth, saves memory. I was suspicious because I couldn't break the tiles down nicely.
Pic 4 is how I would imagine the tiles being broken down. Doesn't seem very efficient, so I might have been wrong. Either way though, the Pokemon devs were either just lazy or pressed for memory, probably the latter.
6ce63c No.15610070
Should I keep doing Godot or should I look into modding the Cube 2 Engine?
6ce63c No.15610074
>>15606715
Those aren't loops though. Those are curves. I see no loops.
4c2f83 No.15610099
>>15610066
Oh and I guess GB (I don't know in which games) and NES (in Mario 3/Contra/others) have a 8x16 sprite mode but that doesn't help you because you don't have a hard sprite count to follow. They don't look any different from two sprites stacked on one another.
b88782 No.15610103
>>15610074
That figure is broken into front and back view. If you follow the yellow collarbone "curve", you'll see it loops around the front and back sides, creating a loop.
a31f91 No.15610132
>>15609893
>>15609943
Errors and dubious stuff (I'm assuming you're trying to write C++ code here) :
>that while is using a constant with a positive (>0) value as condition, so it'll always be evaluated as true. Thus the while loop will run forever.
>Also what's that 69 for?
<Solution: you can either put a variable that changes within the scope of that while and then evaluate a logical statement (such as "a == VALUE"), or indeed put a positive constant but evaluate another condition inside the while's scope using the break keyword to exit the while loop (which is messier most of the time)
>that outer for loop will end inmediately, the counting variable (a) starts at -10 but then the loop condition is " continue if a == 10"
<Solution: it should be "continue if a <= 10" or "a < 10", depending on whether you want the loop to use a=10 or a=9 before ending, respectively.
>you're declaring the same variable "c" everytime a is less than zero.
<Just declare "int c" once, outside the for loops (but the compiler will probably optimize that)
<I'm assuming you want that inner for loop to use the "a" value everytime, whether it's positive or negative. In that case, just assign a to c before that if statement.
<or you could just use the if-else operator: " a < 0 ? c = -a : c = a; "
>there's no need to multiply a basic data type by -1, when the minus itself works as a prefix unary operand. That is, c = -a is the same as c = a*-1, but much more readable.
>that inner for loop will never run: the loop condition is "continue while c is less than b" (c < b) , but because of that if statement c will always be positive, thus said condition will always start as false and no loops will be done. Also the counting variable in this case is b
<it should be "continue while b is less than c" (which is either c > b or b < c; note the symbols' direction)
>you could use <<endl instead of a newline character, but that's up to you
-----
Basically, your code's messy and your shit's all retarded and I seriously recommend you re/read one of the free official C or C++ books just so you can grasp not "how to do X specific thing" like tutorials teach you, but rather the fundamental concepts behind the language, such as Scopes, Loops, Variables' lifetimes, etc. Especially scopes, a lot of beginners always expect their variables to be the same everywhere, everytime, when in fact shit changes without any warning whatsoever depending on which scope is being used.
c98121 No.15610134
>>15610069
If the 'small' pokemon sprites like those are 16x16, the human overworld sprites probably are too, which makes much more sense. So in that case, they could have had the other half of the sprite use different colors. I'm not entirely clear on what 8 palettes per room means, though I assume it's basically how many can be shown on screen at once? So if Shantae's sprite uses 5 palettes, the devs would have had 3 palettes 'free' for everything else at a given time right? If that's how it works, would the Link's Awakening devs have been technically limited to 7, since Link himself uses 1 palette? The gen 2 Pokemon games were hard pressed for both time and space. The only reason Kanto exists in GSC is because Iwata himself personally re-coded the entire game from scratch to make it work.
4c2f83 No.15610143
If you're interested in seeing how an image is displayed and how the graphics are loaded, look at the video ram.
a31f91 No.15610157
>>15610122
It's still wrong man
>the for loop won't use the a = 10 value, the used range will be [-10, 9]
>it will run forever, it's all enclosed within a while loop whose condition is a positive number. It'll be always true.
6ce63c No.15610216
>literally shitting up the thread with gay shit
wew
8665ac No.15610241
>>15610074
that model is unwrapped and projected to be a 2d so it appears to not be cohesive edge loops.
Attached images are the images everyone else is constructing in their mind's eye to understand; they are, in fact, loops (when all is said and done).
>>15610161
>>15610171
>>15610192
>>15610210
>.webm
>doesn't know the rules
kek
6ce63c No.15610253
>>15610241
Cool. Got anymore in depth tutorials on character creation? Like something written?
8665ac No.15610264
>>15610253
yeah, definitely
look here: http://8agdg.wikidot.com/resources#toc17
There's vids on topology, and the polycount wiki for character modeling references and guides.
8665ac No.15610343
>>15610280
If you really enjoy coding at the low-level, want to do something current engines cannot do, and wouldn't mind a potentially huge investment of time; then go for making an engine.
It really depends on the level of complexity of what you're going for, for the potential time investment you're looking at.
Making an engine that's "modern" is a giant investment of time (especially for a 1MA).
Look at UE4's source code, godots source code, and so on; then decide if you want to recreate the features you need yourself.
If the complexity of your wanted features is so simple that making an engine isn't too huge of a timesink (multiply that projected time investment by 3 or 4 first tho) then it may be a good exercise, and especially so if you need practice.
So it might be a good idea to at least try it (to gain some perspective on the complexity of making an engine), and that's to say even if you fail; or decide to use an engine later down the line (worthwhile practice and experience).
If you want to, "just like make game", then use a pre-made engine. Most competent engines can be extended to do almost anything you'd want to do even if they don't explicit say it supports x,y, or z.
>is Godot 3 ok for 2d and 3d games
I don't use godot tho the word is that it's the best choice around for 2D atm, but 3D is unfinished (lacking things like (occlusion?) culling and so on).
8fd1ba No.15610365
>>15609294
The only way your limitations would be authentic is if you made it run on actual hardware
>What's next, you're gonna shove shaders and bloom in an NES game and say that's authentic? Gonna add modern lighting to PS1-style models like hipsters do in their nu-poly styles? I bet you think Shovel Knight looks okay too. Fuck off.
And your post was okay up until you started to sperg out
c98121 No.15610594
Bout of autism on my part, but does anyone else think it'd be feasible to have raid battles in a single-player game? The way I'd thought of it working is that the player is given several 'groups' of four generic units and that after a certain number of turns the group would get automatically switched out for the next one. If the player lets that number of turns pass on the final group (The player characters) it's game over.
6ce63c No.15610601
>ditto
Should I keep doing Godot or should I look into modding the Cube 2 Engine?
750a51 No.15610627
c98121 No.15610724
>>15610702
I need to look into this more.
0c6f98 No.15612737
>be me just now
>try to use printf("Index vals: %d\n", xstart) to print values so I can see what the index is when it segfaults
>THAT'S the line that segfaults
note that xstart is an unsigned int. It is not a pointer.
What the hell?
0c6f98 No.15612764
>>15612737
Nevermind, I've got it.
I think the stack was overflowing and that's all it was giving me.
7eeef3 No.15612771
>>15612737
Whilst it probably isn't relevant to your issue, it's important to remember that values that you throw into var args are never typecasted, even implicitly. So if you're throwing a long long int or short int in there, you have to make special care in using the appropriate size in the printf formatting. Also "%d" is for signed integers and not unsigned integers. You want %u for that.
8fd1ba No.15612820
>>15612737
Pointers ARE integers, to the memory address. This is why you physically couldn't have more than 4gb ram on a 32-bit OS
7eeef3 No.15612844
>>15612820
sizeof(void*) != sizeof(int) on most systems
8fd1ba No.15612875
>>15612844
Sorry, what I mean is that while they aren't directly convertible, they're basically analogous and it's essentially how it works under the hood
7fd49c No.15612884
>>15612820
There's exceptions to that like PAE or swappable hardware.
>>15612844
Compilers can't agree on how big an int is. C was trying to be platform agnostic by not specifying it.
4f3754 No.15614586
>get everything compiled on godot for gdnative, everything runs smoothly but nothing happens on screen
fug
0d8a96 No.15614639
>>15600388
No, I mean, why is he ripped now? What happened to the old body?
0d8a96 No.15614645
>>15601305
Flat is justice
0d8a96 No.15614660
>>15612603
>turnips have flowers too
okay now you're just making stuff up
52b031 No.15614746
Hypothetical question. I have two land provinces and there's a river between them. Let's say that I'm using object-oriented language like C# or C++. Each province is an object with AdjacentProvinces list of type Province. How do I get the information there's a river between those two provinces? Is it even possible or should I rather create class LandConnection in which I'd specify two connected provinces, their distance from each other and if there's a river between them and then create LandConnections list in each province?
12e3be No.15614835
>>15614746
This is a graph theory question. What you are talking about is an undirected weighted graph, where there is a number that gives some information about the connection but you can traverse between nodes in either direction.
Because you are developing a game you should be thinking data oriented design, not object oriented design. Do you need fast access? Do you need low memory complexity? What you use depends on what you need the data to do.
There are multiple different ways to store a graph of this type.
There's an object pointer model, where each object contains a pointer to the connected object and a parallel array containing the weights. This has O(n) memory and access complexity, except for the pointers and parallel array which can have worse memory complexity than that.
There's an Adjacency Matix that has O(n^2) memory complexity but O(1) access complexity. Because the graph is undirected the Adjacency Matix would be mirrored along the diagonal, meaning you could do some tricks to cut down on memory usage.
Here's a scientific paper on storing graphs in memory, which also mentions an edge list and adjacency list. It assumes an unweighted graph at the start when talking about storage, but for a weighted graph you would store a third value for each connection in the edge list that describes the nature of the connection. For the adjacency list each number would be two numbers, one for the province and one for the weight.
http://www.sommer.jp/aa10/aa8.pdf
6ce63c No.15614850
>>15614660
Most tubers do. Potatoes have flowers and even fruit.
792aec No.15614865
>>15614660
Almost all plants produce flowers/seeds, the ones that don't produce spores, but there are other ways for them to propagate as well, but when they do this you get a genetically identical plant, this is generally considered the ideal way to propagate plants for commercial use because it makes the produce homogeneous. the best example for this is Bananas, the most common variety of banana all come from the same tree genetically, probably because it doesn't produce seeds.
553677 No.15615237
>>15614660
Everything you knew is a lie
75c4f3 No.15615347
>work with C# at my day job
>work with .NET Framework 4.7.2 + netstandard/netcore 2.1
>come home and work with Unity
>tfw it's like going back to the stone age
>mfw full bodied properties, full bodied properties everywhere
Given that we have netcore being multiplatform now, I hope they can get their shit in gear.
75c4f3 No.15615559
>>15615347
>tfw no byref properties for value types in Unity because it's not on C# 7
>tfw remembering that Vector3 is a struct
73bf90 No.15615573
I've posted previously in the Freecities thread on HGG but decided to keep it short as it wasn't Freecities related completely, so i thought i'd post here as well to show off my progress and general idea, the idea of the game is basically Fallout + Freecities using the fallout world lore as a base and Freecities for the city building & slavery part, i thought it would be a simple enough cut and paste project on ue4 to learn off of and maybe we'd get a good fun game as well.
in my first picture i've gone for a board game type feel to the city building aspect while in the 2nd picture i decided to take a small screenshot of all the enums i put together in my struct for my human blueprint the list goes further down a bit, my third picture is my hastily thrown together idea of what the city might look like in totality, and finally the last picture is the epilogue or last page of the character creation/settings introduction right before you see the actual game, there's probably a lot of spelling errors and such but for 3 days of work off and on i feel like i've done a fair bit
6fd0a5 No.15616313
>>15615573
the slaves better have 3d models too.
6fd0a5 No.15616633
>>15616628
because you don't know how computers work.
neither do i.
d4518b No.15616635
>>15616628
You need to download more IQs
98b01d No.15616638
>>15616628
Get Game Maker.
From experience, depending on what you use, it's usually an overbearingly difficult start since you have no idea where to even begin and you need to learn fifty things right off the bat and no tutorial really tells you WHY you need to know them. It gets easier as you go along, as it just becomes basic logic puzzles.
98b01d No.15616659
>>15616644
it was a joke, anon. the rest of my post was supposed to be encouraging.
0b38e7 No.15616731
>>15616628
Knowing and using math makes it easier.
3b6740 No.15616812
>>15600139
Anon, I hate to break it to you, but one rectangle is made of at least two triangles, that's just how it is. If you want to have rectangular bars made of triangles, a repeating value could be represented by one rectangle. E.g. if the lowest value repeats, you could put one big rectangle at the bottom (made of two triangles) and only those bars reaching above it would need their own rectangles.
3b6740 No.15616910
>>15616812
>one rectangle is made of at least two triangles, that's just how it is
Disregard that. I suck cocks :/
7eeef3 No.15616929
>>15616910
Except that literally only works if one bar is EXACTLY 2x the size of a bar next to it. Which is essentially equivalent to "repeating value" nonsense.
It does save you fucking performance for shit to write code that checks for it. All you do is move some vertex shader computation to the CPU in an extremely inefficient manner.
3b6740 No.15617023
>>15616929
What part of I suck cocks didn't you understand?
792aec No.15617051
>>15617023
Can you give us a visual example of you doing this?
91a16f No.15617238
>>15612942
Any more videos like this? it's really great
8fbb00 No.15617348
>>15615559
Vector3 isnt purely a struct. Unity preallocates them on the stack. There is no performance calling the same new Vector every frame. Hence why you can have Vector3.zero.
8fd1ba No.15617494
>>15614850
Excuse me but what's the deal with strawberries? They have runners, and yet they still need pollination?
8fd1ba No.15617504
>>15615347
>>15615559
>Using W7 Home Premium with neutered telemetry
>IDE of choice is Visual Studio, specifically VS C# 2010 Express edition
>Only includes a compiler up to C#4.0
>Even if I install a newer compiler, it is hardcoded and fails to recognize it (specifically the program will complain when it loads a project - starting a new one and adding all the files and compiling it works fine here, just loading it and saving it doesn't)
>Can't upgrade to a newer VS edition because they all fucking need W8 or W10
792aec No.15617568
>>15617152
Well I can't fap to this.
>>15617528
I hate that .gif so much.
792aec No.15617587
>>15617494
Pollination is how you get fruits/seeds, the runners are asexual reproduction, they produce genetically identical plants, much in the same way growing a new plant from the bulb/tuber creates a genetically identical plant. Plants need to exchange pollen to reproduce sexually, some plants have male and female flowers and can pollinate themselves, some plants only produce male or female flowers and need to exchange pollen to produce fruit, only the female flowers will produce fruit.
6e97ee No.15617913
What would be a good number of skills for an rpg character to have at endgame? I was thinking an average of 16, not counting 'generic' skills that can also come from equipment. Party would have 7 characters.
98b01d No.15617990
<shitty turbo-potato with parts from 99 that runs XP
>couldn't even load XP off of a fucking flash drive properly
>needed to use Easy2Boot and load XP into RAM to get it to work
>such a piece of shit that I can't get passed the legal request when you open up Google
>runs my engine with no problems
<office PC from 2009
>runs Windows 7
>is crap but useable
>can't initialize OpenGL window
Even better, the old piece of shit can only use OpenGL 1.5 and runs fine. Had to get graphics drivers to get it to work on the newer one. What a strange world.
>>15617913
I know a few skills would overlap, but still, it seems that, unless you're exceedingly creative and have many different types of attack damage and status ailments and the like, characters could become fairly samey pretty easily with that many. I'd say 12 would me more manageable, and again, you need quite a bit of creativity and complexity to prevent staleness.
73bf90 No.15617994
>>15616313
No its like Freecities but with 3d buildings, plus i dont have the capacity for that
d21423 No.15618011
>>15617504
So stop using the sack of shit that is VS. It seems like an easy choice
6e97ee No.15618060
>>15617990
As far as number of damage types, there's Fire/Water/Lightning/Earth/Air/Ice/Light/Dark for magic and Slash/Pierce/Strike for physical, with those being broken down further for each weapon type. I.E., Bow and Spear both count as Pierce damage, so equipment may block Bow or Spear damage individually, but can also block Pierce altogether. Then there's hybrid damage, which is just magic+weapon. like a flaming sword
I'm counting basic magic skills as both generic and unique, so while most characters can use most magic with equipment, the character who specializes in it will have stronger, natural versions. The number of unique skills a character has includes passive skills like 'Wind Boost', too. This is what I've thought of so far for the thief character:
Poison Strike: Inflict low Slash damage on an enemy. May inflict Poison.
Pocket Sand: May inflict Blind on an enemy.
Swift Strike: Inflict medium Slash damage on an enemy and raise own Speed until end of next turn.
Blade Flash: Inflict high Slash/Air damage on an enemy based on own Speed.
Wind/Gale/Tornado Boost: Raise the power of Air magic. [passive]
Rush Stance: Enhance own Speed and Critical rate for 3 turns.
And then of course he has Air magic and Steal skills. I think this is around 10.
98b01d No.15618173
>>15618060
I just have one question: a fan of Persona?
Seriously though, with that kind of breakdown, you could probably fill up 16 moves a character, but whether it would be unique enough, I'm not sure.
The way I'd consider doing it is a semi-job system, where instead of jobs, you get branching paths for each character at certain points of leveling up. Perhaps let's say, one character could chose from either focusing more on powerful physical attacks, powerful magic attacks or learn hybrid attacks which could generally be weaker or cost more AP to perform, but would have the benefits you described.
16 moves with that kind of variety still seems potentially overpowered unless you limit very heavily what kinds of moves each party member can learn and that's not too interesting, then. The system I described adds a bit of strategy and some more consequences to long-term decisions and planning. That way, you would need to strategize your way out of a bad party setup instead of just grinding. 12 moves would give plenty of space for variety and strategizing while still limiting it somewhat.
Even then, 12 moves, you'd need some damn well planned out bosses with tough strategies, ones that change up over the course of the battle. Most bosses in RPGs, especially status-ailment-heavy ones like Persona have one strategy you needed to repeat to win. Avoid that.
Generally, though, your idea's not bad, I'd just need to see a more planned out implementation to comment more. Just throwing around some thoughts.
64f7e7 No.15618178
>>15617913
It depends. Are you only asking about active skills? Do characters have utility outside of their active skills? Characters with flexibility in their skill sets or natural value in their base characteristics (like high survivability and great attacks) don't necessarily need to match other characters with the length of their skill lists. That sort of thing can vary depending on whether it's a combat grid, real time, or a final fantasy-like combat system. If this is a turn-based game without a combat grid like a final fantasy game, I imagine 16 skills would be more than enough, with most of the skills having unique effects, and perhaps a couple of repeats that are just more powerful versions of earlier abilities. Just try to focus on having unique effects and abilities.
I have been getting positive feedback about characters in the game I am working on. I am using a battle grid, and characters tend to have very small amounts of specialized abilities. I have a fighter-archetype character that has a 'heavy' or 'skill' attack, which varies on the weapon, and a fatigue reduction skill, and that's it. Through testing, I've found that giving him any more abilities clouds his purpose - you want his actions to be focused on getting into good positions, blocking enemies that need blocked, and killing enemies in close combat. Wasting his actions on other effects seems to contradict his purpose. But the point I'm getting at, is that having 16 options will not matter if every battle is focused on the same 2 options. Look at your encounters, look at your skill lists, and figure out if you're benching certain skills in 99% of encounters. It's okay to have some skills that are more focused for fighting bosses and strong encounters, but sometimes you might be able to get more utility out of them with small changes, or by having other party members work well together. For example,
>>15618060
>Poison Strike: Inflict low Slash damage on an enemy. May inflict Poison.
This is a skill that would probably be overlooked in the majority of encounters, but thrown up at the start of combat on a strong enemy. If the character had some sort of power that would, say, use wind to inflict weak cuts on all creatures while simultaneously sharing poison effects they may have, you'd occasionally have more use for it. Or if you had another character that performed a similar action (spreading poisons), that'd increase the value of poison strike, and let players make more valuable decisions on which skills they will use.
6e97ee No.15618270
>>15618173
>a fan of Persona?
Yes, though I like the higher difficulty of SMT more. Haven't played SMT 4 but I heard it's easy. I'd considered a system like what you've described, but also thought about saving it for a different game. One thing I could do is make it so that characters learn a large amount of skills, the player can only use a certain number of them at once, and can swap them out between battles. I definitely agree about boss strategies, I'd noticed the thing about one repeatable strategy and do need to avoid it. One way I thought of to do that is to segment the boss' HP into multiple bars and after each bar is broken, they purge their debuffs and change tactics. Imagine if halfway through the fight Matador turned into Mot.
>>15618178
Active skills, I guess. I suppose the warrior-type character wouldn't need as many skills as the thief would since the point of that chracter type is to make big numbers pop up. Enemy variety can go a long way towards mitigating the issue of dominant abilities. One idea I'd had was to make most party members technically optional and let the player hire mercenaries to fill in gaps they think they might have.
98b01d No.15618325
>>15618270
>Yes, though I like the higher difficulty of SMT more. Haven't played SMT 4 but I heard it's easy.
Just a hunch, since your physical attack types are pulled straight from Persona.
>I'd considered a system like what you've described, but also thought about saving it for a different game.
Fair enough. It is quite an evolution from what you originally described.
>One thing I could do is make it so that characters learn a large amount of skills, the player can only use a certain number of them at once, and can swap them out between battles.
That would need to have some in-universe explanation to make the concept easier to understand. It's quite similar to the concept of Personas, just applied to the whole party, and without the ability to swap mid-battle. The only way I could think to explain it in-universe without something like Persona or Pokemon or the like would be to just severely limit SP and make SP management crucial, but that may be tough.
>One way I thought of to do that is to segment the boss' HP into multiple bars and after each bar is broken, they purge their debuffs and change tactics. Imagine if halfway through the fight Matador turned into Mot.
You could do that in early fights to explain the concept, but then later on stop highlighting it, to keep the player on their toes.
>One idea I'd had was to make most party members technically optional and let the player hire mercenaries to fill in gaps they think they might have.
While it's not a bad idea in theory, with only having seven characters in total in the game, that's iffy. I'd say have more characters in total, then. As for the mercenaries, an important lesson is that they're doing it as long as it benefits them, as in, they get paid. If their life's in danger, they should just run, saying "fuck it, this isn't worth the money." Between potentially having less unique skills, and needing such hard maintenance, it would heavily encourage seeking out permanent party members.
You could also consider each optional party member having their own sidequest line, and perhaps some members could simply not agree with each other, for various reasons (a heavily-religious mage would consider a thief's actions dishonorable and decide that she cannot associate with that), and you would have to choose between one and the other. Just spitballing ideas again.
6e97ee No.15618381
>>15618325
Yeah, mercenaries wouldn't be able to be taken out of the area they're based in, would only stick around for a certain number of fights and leave if things get too dicey for them. The 'optional' members having a chain of sidequests attached to them was something I'd already decided on, since by being optional they wouldn't exactly be integral to the main story. though a hypothetical sequel would assume the player recruited them A way I thought of to encourage recruiting them permanently was to have them join the player temporarily beforehand for the area's major dungeon so they could get a feel for how the characters would work. Some mercenaries would have similar skillsets but generally weaker stats.
On another note, what would you think about difficulty also affecting a game's length, basically making it so the player would have to play higher difficulties to see everything?
98b01d No.15618417
>>15618381
>A way I thought of to encourage recruiting them permanently was to have them join the player temporarily beforehand for the area's major dungeon so they could get a feel for how the characters would work.
An interesting thought. I'd say you could mix it up a bit, introduce the player to the idea of recruiting optional party members using the method you described, yet keep a few hidden outright, as a bit of a reward for exploring.
>On another note, what would you think about difficulty also affecting a game's length, basically making it so the player would have to play higher difficulties to see everything?
A classic NES type of deal? I'm not too fond of the idea. I'd say a better way to handle it would be to have a general plot involving the main cast, and a more overarching plot that you see glimpses of as you adventure, a plot that affects the whole world. The party beats the final boss, boom, roll credits, happy ending, but then you're given the option of a post game story involving that overarching plot.
One of my favorite RPG moments was in Final Fantasy IV's Endgame, where you start encountering bosses and super hard enemies as regular encounters and they play the Boss music. Needless to say the actual bosses are hard as shit. It was such a cool fucking touch.
6ce63c No.15618418
Is Cube 2 a good engine to base my game off of?
a0109d No.15619004
>>15618418
>>15610070
>>15610601
How about you try those 2 engines if you are willing to write this question 3 times?
d50883 No.15619587
>>15615573
You'd benefit by making your enums less specific. Do you really need separate enums for breast, nipple, areola, pussy, labia, clit, ass, anus, hip, dick sizes that a single size enum wouldn't suffice? The same for colors. And clothing and collar would be more useful as structs or objects so you can have a variety of information like material, mesh, value, color, condition, cleanliness or whatever else you need, and be able to track changes to them over time.
75c4f3 No.15619895
>>15617504
Dude that is patently untrue. I've developed for years in a professional and casual capacity with VS2012, 13 and now 17 on Windows 7. You can't target Win Store but that's about it.
3b2cbb No.15620226
>>15617504
Try the only VS i don't hate : "Visual Studio Code", has decent support for C# too.
5c1e57 No.15620368
>>15620226
>Micro$haft product
>open source
>multiplatform
>supposedly without all the bloat VS comes with
What letters do those barneysteyn bears have here?
Remember to search for telemetry.enableTelemetry and maybe @tag:usesOnlineServices in File > Preferences > Settings to hopefully not report back to Bill Gates all the time.
ef2c47 No.15620502
>>15620368
Microsoft is trying desperately to rebrand as open source friendly, hence the acquisition of github and the open sourcing of several of their projects
0ccddf No.15620620
Finally something to show
The map data raycaster is finally done. I can add any number or arbitrary data and associate it with groups of triangles. It will be used for Soil Information, fish spawn, animal spawn and whatever else we need to associate with the ground or zones. Currently any grass is plantable, but it's not actually associated with the texture (We probably will though, since it needs to be easy for the player to know where he can plant and where he cannot).
The video shows the Soil detection, if you hit your hoe in any position, it will either produce a red or green marker, a green marker means you can plant here, a red one means you can't.
It works by checking the groups of datas in each triangle, (each triangle can be part of one to three faces, it will therefore have all the information of all faces it is part of) if all groups have a Soil Information data set, then it will enable planting, if any of the data sets don't have Soil Information, then it does not enable planting, since it will either be invalid terrain, or a triangle that borders invalid terrain.
The picture is a example of where each data set starts and ends. The system creates "faces" which connects through borders, therefore, some triangles, have two or more face data. Where B have no soil information and A have, therefore the whole A+B zone is invalid, even though there is grass there
2a7a33 No.15621705
does anyone here have experience with godot's networking system? And if so, how shit was it?
8fd1ba No.15622242
>>15619895
It must require a KB pack I've ripped out then, because a lot of "modern" stuff that requires Windows doesn't work for some reason
8fd1ba No.15622416
Oh hey, I have a question on how to design components. Like I want to split my game loop into an "update component" and "draw component" so that they can have different tick rates (eg it could update 120 times a sec, while only drawing 60 fps).
When I'm passing elapsed time into each one's update/timing loop, where am I getting the delta time from? Like, should I still have one actual main loop, that passes the time from the last tick, or should each component track the time since its last update?
98b01d No.15622433
>>15622416
>limiting framerate
No.
>When I'm passing elapsed time into each one's update/timing loop, where am I getting the delta time from? Like, should I still have one actual main loop, that passes the time from the last tick, or should each component track the time since its last update?
Keep one timer that handles everything. I don't see any benefit to using independent timers.
8fd1ba No.15622897
>>15622433
I can absolutely limit framerate, because otherwise the game will shoot CPU cycles up to 25% usage (one full core). There's no reason to do that
8fd1ba No.15622926
73bf90 No.15623473
>>15619587
they use different naming conventions for example a nipple type can be inverted but an ass or a hip cant, later on i can change clothing into a struct if i need to but they are just descriptors for now