/v/ - Video Games

Artistfags lend me your knowledge.

I have a bunch of tree images and want the root to be in the same spot in every image.

What is the best way to align them pixel perfect?

>>14552184

Not an artist, but have you tried putting them on the correct position?

>>14552184

Do you mean you want to change the images so their roots are aligned or need some sort of metadata for aligning them all when rendering them?

>>14552225

I would prefer to manually align them but the only way i can think of is to go through all the images and move the into place manually. Surely there is a better and more reliable way?

>>14552184

>>14552247

i think you'd waste more time trying to think of an algorithm to unfuck your trees rather than just doing it manually

>>14552247

That's probably what you're going to have to do. Depending on your graphics program it might have a batch export/save to take most of the tedium out of it.

>>14552222

STATUS: ENLIGHTENED

>>14552297

B looks weird probably because it's too exaggerated. Reduce the curve a lot, it wouldn't hike THAT high, especially if it's specifically designed for women.

>>14552317

meant to

>>14552306

>>14552099

For pathfinding interactions, have the NPCs ask the doors if they have permission to pass through

For interaction, make the door order the NPCs to unlock them. The unlointck animation is a property of the door, not of the NPC.

Otherwise, if you add 10 interactable objects, the NPC code will already be unmanageable

>>14552361

>unlointck

unlock

I don't know how that happened

>>14552099

This actually made me think of something funny to try implementing.

So there's already a piston in minecraft, but why isn't there a spring-loaded trap? Your webm made me think of cartoons where a character opens a door only to get socked in the jaw by a boxing glove on a spring.

That's going on my to-do list.

>>14552361

that much i can do, i'm already handling off mesh traversion by script

the issue is that if the ai can't pass through the door, what does he do then?

unity's nav mesh has no "find a different path" option, so he'd be stuck trying to get through a locked door because it's the closest path, even though there's a different way

>>14552375

can't you set the door weight to 999999999999999999999999999999?

>>14552381

it would be global

would fuck everyone up, even though we want it to be per agent

>>14552384

Then I guess you will need two pathfind solutions.

One custom made, where it's just a simple graph path, which tells NPCs where they can and can't go

And Unity's pathfind, to tell NPCs how to follow the graph's path

Or you could fucking google how to solve it, it's impossible that you are literally the first person to have to deal with locked doors and unit's pathfind

>>14552391

>it's impossible that you are literally the first person to have to deal with locked doors

i disagree

everyone on unity forums just says "lmao, just like raycast" which isn't really a solution to the pathfinding issue

>>14552397

Unity seem to have a massive hard on for recommending raycasts for everything. Even their tutorial for a 2D grid based game was using raycasts to check neighbouring tiles.

>>14552397

https://docs.unity3d.com/ScriptReference/AI.NavMeshObstacle.html

check if this updates instantly when you order a pathfind.

You could check all doors before a pathfind operation, enable this for all doors that are locked, disable for those that are unlocked. Do this for all navmesh operations.

But I still believe you should have a macro and micro navigation solutions, you shouldn't be asking Unity's navmesh to walk a npc from one city to the next

>>14552406

>>14552391

>>14552381

as far as i can understand, the only real solution to my problem is to do >>14552381

for a single frame and or so and then turn it back and make sure that my navmeshagents don't automatically repath

it still sounds like a horribly messy solution though

>>14552423

alternatively, enabling and disabling offmesh links

http://docs.unity3d.com/Documentation/Components/class-OffMeshLink.html

>>14552431

enabling and disabling offmesh links is what i'm currently doing and it works fine for the player, but the issue with that is that i can't figure out how to get the AI to go to a door to enable the offmesh link (assuming that they have a key or something)

>>14552437

before doing the pathfind order, check all doors, disable locked ones, enable unlocked ones, do pathfind order.

Make npcs go through doors as a 'animation' rather than a pathfind navigation, if needed

>>14552447

let me just reiterate what the issue(s) are, to avoid confusion

if i somehow disable part of the navmesh, either via a navmeshobstacle or a disabled navmeshlink, the ai doesn't know that there's a door there, and even if he has the ability to unlock it, he won't go there

if i always keep the navmesh intact, i can stop the agents from passing through the door, but can't force them to go any other way so they'd be stuck

i think the easiest way of doing this is to:

1) have a link at the door as it currently is. ai goes to door.

2) check ai inventory for key or something that can unlock the door

if so, use it to unlock the door. repath

3) if not, disable the link, repath. and all agents must NOT repath, otherwise when i reenable the link for everyone else to use, the ai would come back to the door

might work, will try

I have a question for you anons here; How does one go about making sound files sound like the sound files used in the DOS Era? (Such as Duke Nukems voice clips where they are heavily compressed?)

I am asking for authenticity reasons, but can't seem to get the effect right

>>14552461

The steps I suggest are as follows

1) For all doors, check if the NPC can go through them, disable/enable as needed

2) Do the pathfind

The npcs will never reach a door they can't open unless they have their keys stolen mid-path

Also, the other npcs can't repath, that's a given

I still think a macro-level pathfind is needed, this is decided which rooms/streets/plazas the npcs can go through, not how they go through them. You could also add 'polite' links with it, to avoid NPCs going through shops to save time

>>14552558

so?

The same whore is under development, and every time it was about a different part of her development

>>14552574

What the fuck do you think these threads are for, you double nigger?

>>14552582

>annoyed at the frequency

Maybe if you spent your time developing instead complaining, you could too

>almost entirely pointless updates

That's very subjective

>>14552582

>slightly changing the colors of socks

>minute change

Maybe you should consider killing yourself, you disgusting normalfag

>>14552503

ultimately that doesn't work because i have areas that can be accessed from multiple doors

example being the current building - main entrance is a door, but you can also climb up on a ladder, run over the roof, onto the balcony and then climb down the stairs

even if the npc doesn't have access to the main entrance, he should still be able to go through the other entrances

also, i managed to get the navmeshlinks to not let an agent through, but if i reenable the links afterwards, the agent repaths back to them

apparently navmeshagents auto repath variable is only for no longer accessible paths, not for shorter/longer paths

i can't figure out how to get him to only do the pathfinding once

>>14552582

>hey guys what color do you think I should give these socks

<try green/purple or rainbow

>okay, here's what that looks like, but I don't like them

<try blue white

>ooh that looks good, here's a pic

>now I made a shirt, how do you think it should work with big tits

<[more input]

And then you come in REEEEing.

>>14552621

Rotation is played out since the 2014 fad. Try translating it instead.

>>14552621

>>14552625

have you tried killing yourself, faggot?

Is there a way to make a game like Honey Select not boring?

>>14552632

Using more varied puzzles might do it.

>>14552621

Use blender to make porn

>>14552541

>being a goon

>>14552640

Looking triple AAA there.

>>14552640

>>14552625

>>14552621

filter and move on

>>14552621

>>14552625

>>14552640

Have you ever made anything or do you just get angry when other people do?

>>14552666

no satan, don't feed, just filter and move on

plenty of arguments were already wasted on him

>>14552618

really don't want to have to make whole new systems for something that was nearly working, but as far as i can tell, the navmeshagent ALWAYS repaths, so i can't just give him a path once and expect him to follow it

you'd think that this is something that should be doable by default

i can calculate a working path ignoring the door link, tell the agent to follow it, he'll start, but as soon as he reaches a corner, he'll go back to the door that's been reenabled for the rest of the agents

fucking unity

>>14552184

>>14552247

do you mean like having a tool to help you with? why not just gimp layers?

>>14552671

Wise, but unfortunately the application I use doesn't support it. I guess I will do it manually.

>>14552674

You would need the macro-level pathfind for going from City A to City B anyways, unless NPCs never migrate

>>14552367

Sick, figured out how to get existing blocks with different metadata to be named accordingly, it's annoying to see logs, planks, saplings and leaves all just use a generic name instead of telling you what they are.

>>14552744

is there a source for these?

>>14552753

This, they look very helpful for research…

>>14552711

well the idea was for levels to be separated, so they probably won't migrate

anyways, here's an even dumber solution

https://answers.unity.com/questions/1103247/off-mesh-link-limiting-access-to-certain-agents.html

just have two areas - one for regular doors, and one for locked doors

depending on his inventory, an npc will have locked/unlocked doors assigned to his nav area mask

only problem with this is that you only need one key to unlock any door

>>14552873

yeah, you would need one area for each key

>>14552910

in the end, the amount of work is not dissimilar from making a macro-level pathfind system, it seems

>>14552554

Where'd all the facebook/disagreement niggers come from? We didn't use to communicate in stock emoticons.

>>14553065

>stock

It's not stock, though, it's an edit. Ironic usage and mutation of normalfag memes is common. You can mutate normalfag material until it's not normalfag material anymore, and vice versa.

>>14553092

It's worse than normalfag, they're aping fucking niggers. Besides that, I'd prefer we not ~ironically~ make them think this is a suitable place to settle down.

>>14553124

If we're including 4chan, then Risk of Rain in terms of popularity. Speebot in terms of technicality (the entire 3D engine made from scratch).

>>14552744

>have you ever seen a girl in real life anon?

Wait… do they actually exist?

>>14553124

Speebot as far as i know

http://store.steampowered.com/search/?snr=1_4_4__12&term=speedbot

He is working on a new game too, looks pretty good so far

>>14553130

>(the entire 3D engine made from scratch).

I didn't know that. That is pretty cool.

>>14553148

>animu stalker

IDEA STOLEN!

>>14553142

What game is he working on?

>>14553163

im trying to find some of the webms he posted, looked like a bullet hell game

>>14553148

>>14553163

>aminay STALKER

Double stolen

>>14553117

>he googled "facebook meme thinking" and copied the first image result to call me a newfag

>>14553148

So once we've been replaced, where do we go?

Am I the only one that thinks the wiki is a bit of a mess?

There are some projects still in the active category which I know have been given up or put on indefinite hiatus.

Problem is that some of these articles weren't made by the people who worked on these projects so they feel no desire to change them.

>>14553288

pretty much

i already post progress here, i don't feel like bothering to also put it in the wiki

>>14553288

I don't think anyone really reads or cares about the wiki

>>14553293

I'm also pretty sure your project is on the hiatus page.

>>14553285

>So once we've been replaced, where do we I go

trash is a fitting place

>>14553301

so, are you cuckchan or reddit?

>>14553307

why make a question which the answer you refuse to believe unless it conforms to your bias?

are we supposed to compare dick sizes, which cannot be confirmed or denied?

fuck off newfag

>>14553319

I'm learning towards reddit.

>>14553288

There should be a general listing that lists all projects no matter what the status, that way it's less harmful for anyone to pull projects that haven't been seen in a while off from the active list.

>>14553301

>>14553319

Filter, my friend

>>14553334

In case af2d49(32) is having trouble understanding, it's like a personal downvote button

>>14553321

I bet you lean towards men too.

>>14553354

We're quickly reaching the point where all jokes and content are completely ironic

>>14552933

check this shit out

https://forum.unity.com/threads/k-pathfinder.447929/

since it's both free and seems to cover what unity's already does, maybe it's easier to hook some code into the navigation code

>>14553431

Sounds like one of the graveyard projects

>>14553431

>There was another thread here about a concept for a 6DOF dogfighting game involving flying magical girls using wands, machine guns/rifles, and swords in an alternate history WW2

That's totally not "Saga of Tanya the Evil" the videogame

>>14553496

If it's under active development it's not on hiatus you goof

>>14553566

>he doesn't embed nsaware in his games so he can spy on fellow nodevs

>>14553496

>kike-tier word raping

>on my agdg

>>14553566

> he thinks there aren't feds on 8ch

> he thinks I'm not a fed

> he thinks I'm going to believe he's not a fed

> he thinks everyone here isn't just a fed trying to fit in by making vidya

>>14553596

Shit, we're onto us.

>>14553611

Can I jingle her bell?

>>14553628

>her

>>14553611

how many copies do you think you'll sell? 50,000 minimum, right?

>>14553611

Is there a character creator? please include some light kemo stuff like mabinogi

>>14553680

no, i need weebs to draw sleazy fanart of my waifus. fuck character creator.

>>14553684

Character creator hasn't kept people from drawing Dragon's Dogma fanart.

>>14553722

thanks anon, it's all for the greater good!!

also my good, obviously.

>>14553690

but i want to see fanart from MY waifus, not yours, your waifu a shit

>>14553880

please no, this gets old super fast, was the worse addition to pokemon ever made

>>14553906

>not wanting a post-battle hugs/headpats RPG system to level your waifu's abilities

>hell depending on how autistic the devfag wants to get there could be social link levels, different kinds of affection level different abilities etc

it's like you don't want to fug your waifu until she learns how to fug female enemies to death or something

>>14554456

I don't think any of this does anything at all, it's basically just companies agreeing to tell their bloggers and HR team to bitch about online games. It has less weight than the Accord de Paris

>>14554321

>>14554321

So according to this thread, a number of companies are "standing in solidarity" and making a consortium group to fight "online toxicity", eg being straight or white or posting on imageboards.

Specifically Unity Inc. and Epic Games are on this list, so if you use their engine, that means you support censorship. Incidentally, making your own engine for games means you're a bigot.

>>14554444

Wow I should have read the thread. /agdg/ was my first shot at seeing that. Glad another anon thought the same thing.

>>14554444

There's already a thread for this

>>14554732

It affects /agdg/'s usage of game engines.

>>14554739

How so? Is anyone going to change engine because of this? It's just PR

>>14554444

Four fours confirm you need to dump SchlomoWare UE4 is already owned by the ricekikes of Tencent and Unity is getting worse by the day

>>14554739

>>14554750

This is standard good goy behavior. If anyone here does use Unity, don't be dumb enough to buy any assets.

>>14553288

Eh, it is what it is.

It's good for archiving and for getting information on what people are working on even if a lot of information is inaccurate or outdated, but personally, I'm more interested in making my game than working on a wikipage for it.

Still, it's good to know that there are people interested enough in the stuff I'm making. That pushes me forward.

>>14554444

As if I needed any more reasons to not want to continue to use Unreal in future.

Don't get me wrong. It's an amazing piece of technology and I enjoy working with it, but that using it means that I give money to Tencent already rubbed me the wrong way. I also don't need many of its state of the art features, because I'm not working on anything even remotely AAA. So I'm open to change, but CryTek is being run by idiots and Godot gets its fair share of complaints wherever I look. I also won't switch the engine in the current project again. That would most definitively kill it.

I question how much will actually come out of this initiative though. For all I know of could be nothing but a PR stunt, nothing but virtue signalling. Then again, it might not.

>>14554894

is crytek even alive anymore? i thought amazon bought them

>>14554960

Amazon bought their technology and rebranded it to Lumberyard for their own use. They didn't buy CryTek itself.

I've been paying attention to them for years now. It's downright incredible how many bad business decisions they made. Right now they should be a huge competitor to Epic. Instead, they're pretty much irrelevant.

>>14554960

Presumably it was the stipen of money to save their asses from becoming bankrupt.

Lumberyard is the result of that deal in terms of what amazon gained

>>14554972

he wasn't personally talking to you. This is a good example of why race mixers can never be valuable, though, they're far too invested in their degeneracy.

>>14554996

Nice attempt to bait, step up your game anon-kun

>>14555005

It's true, a white man with a cute Japanese wife would never give up his "relationship" for something such as a morality. He's too invested in it. In the exact same way, a unity-fag would never give up his current project because he's too invested in it.

>>14555009

Switching engines makes more sense than abandoning wives.

>>14555015

If you have any sense of morality, and you're aware that creating happas is wrong, you won't do it. That's worse than just giving money to Jews.

>>14555009

Now you’re just breaking rule 8, also see >>14555005

>>14555072

You think that a plausible excuse is just as good as doing the right thing. Don't use Unity.

>>14555102

next you're gonna convince drug addicts to quit drugs by telling them that drugs are bad

>>14555137

Well, if they knew drugs were bad, and had any self respect, they wouldn't do them

>>14555192

It's not quite that simple. I would say this holds true for starting to do drugs. People who start doing drugs almost universally have no self respect. But those drugs fuck addicts head so much that self respect isn't what's going to get them out of it.

Take something like meth. Having listened to people give their experience on meth they feel like total and complete shit when they're off it, but when they're on meth they could be a skeleton covered in skin, scars, and bruises but they are totally convinced that they are the coolest person in the world. Meanwhile something like heroine people take to feel normal after they've built up a dependency. They take it because of a self preservation instinct, not only because of a lack of self respect.

I would say your sentiment holds more true for weed and LSD. But you can't really get addicted to LSD because your body builds up an immunity so fast, and the most addictive property of weed is only present if you're life is shit in the first place it. Weed makes everything feel alright. That isn't universal though.

Sage for off topic.

>>14555118

Sigmanon you should make your engine even more /pol/ themed, it will be funny to watch the (((MSM))) reports about it

>8chains hackers make game engine to develop school shooting simulators, now look at this exploding van

>>14554444

>>14554996

yea i'm not to far into my dev cycle and i could walk away from unity, but….. i don't really see how it is any worse than giving 5% to the chinks or hoping that Godot works for me. while losing access to tutorials and the heard of fucktards that is their help forums* i think the latter may be a plus thew** and after all that shit its hopeing that the godot devs are not going to sign on. or whatever other engine you want to use. and the make your own engine dev cycle is a dead end. but all of this is mute to the question of

"is it ok to use cucked software to make red pill the game?"

i mean what the fuck can they do? if i ever make a big enough impact i will have to licence the engine from them and if i get my foot in the door they cant stop me. what the fuck they going to do? refuse service? soft ban my unity help questions/responses? C&D my ass? worst case i huff and puff and make ether free press for me or damage them. i have the free time. if i duck under the 100,000$ a year they don't get shit. and if i make more than that i can get help and use anything else for red pill 2, the happening. (now free for owners of red pill 1, unity devs suck donkey dicks)

>>14555800

Really there are very few things I can do at this point besides use "racist numbers" as the default port, beyond making the logo of my project a swastika or something, which I don't want to do since turning the entire project into a /pol/ meme would kind of eclipse everything else about it. Since, it's not so much about the engine, but what kind of content you make on the engine. If I made a school shooting game, it would be pretty easy to get some entertaining press coverage about it, but i'm not at that kind of point where I can make any kind of real content.

It would be rather funny if, in the future my engine is used to make a popular multiplayer game, and all of the articles online about "1488 is a racist number" are buried under a million " hey guys, check out my sigmacraft server it's at 192.168.1.8:1488" posts online.

>>14555920

yeah, good point

>>14555825

>i mean what the fuck can they do?

<Unity will have the right in its sole discretion, and without prior notice to you, to suspend or disable your Unity Account or terminate the Agreement and/or your right or ability to access or use any of the Services if: (a) you breach this Agreement; (b) your use of the Services poses a security risk to, or otherwise adversely impacts, the Services or any third party; (c) your use of the Services subjects Unity, our affiliates or any third party to liability; (d) your use of the Services may be fraudulent; (e) you have ceased to operate in the ordinary course, made an assignment for the benefit of creditors or similar disposition of your assets, or become the subject of any bankruptcy, reorganization, liquidation, dissolution or similar proceeding.

<if you are a Legal Entity, Unity or its authorized representatives may upon reasonable prior notice access and inspect your facilities and computer systems to review and verify your compliance with the Agreement. Any such inspection shall be conducted during regular business hours at your facilities or electronically via remote access. In the event you have impermissibly used Unity Personal (or other products) or have not paid the applicable fees for all Services you have deployed or used, you agree to immediately pay for such Services, as well as the reasonable inspection costs, upon Unity’s demand.

>>14556099

i guess ill have to be quiet until i have a stable build then

>Engine/framework EULA/ToS doesn't matter to anons here

>Because nobody will ever finish a fucking game and if by chance someone does, it's because they had the autism to do so

>If they have the autism, guaranteed they make their own engine

>>14556099

I'm not a lawyer, but, I doubt any of this would be enforceable in an American court, and since Unity updates so frequently you could easily have them attempt to use this against you for a project that wasn't created with agreeing to this ToS being a thing that you ever had to do.

Silently changing their ToS on their site without forewarning or headlines would also likely get it thrown out.

No person or company can ever be expected to forfeit constitutional rights to comply with a ToS or EULA that isn't an explicitly formed signed contract, and even then it's a gray area.

>>14554404

looking good anon, breast shape is much better, the shirt was also nice

>>14556099

You cant reasonably argue that making a game that they don't like falls into any of those categories beyond the vague "otherwise adversely impacts" clause which would allow them to revoke the unity licenses to any kind of game they don't want to be associated with, which has no precedent of being invoked in such a way from my knowledge.

Just build your own engine guys, it is that simple.

>>14556310

>>14556226

I will admit that the Unity Police stuff is unlikely to the point that it's basically never going to happen, fourth amendment and all that. Which is why at the bottom of the EULA they have "If any provision of the Agreement is held to be unenforceable or invalid that provision will be enforced to the maximum extent possible and the other provisions will remain in full force and effect."

But the question is what they could do and I took that as a worst case scenario. The stuff like cutting off access to services is more likely especially when using a free version of the engine.

For something with precedent there is what Epic did with Hatred, they asked for their logo to be removed from the final product. However that's not possible, or at least very heavily discouraged, with the free version of Unity. So I'm not sure how Unity would deal with that. If you were to create "Hatred 2: School Shooter Boogaloo" in Unity I'm not sure how they would react to the controversy that would surround it. But something like "The Jewish Problem: The Game" would likely receive a stronger reaction.

>>14556099

I’ve got a bit of experience with lawyer talk, I’ll give a brief overview in layman.

First, there’s been precedences set via the release of previous unity games, where they didn’t violate “the agreement”, and thus there’s a very large degree of wiggle room here in terms of what they deem worth enforcing their EULA (f.e. the literally who beat up sim was made in unity, no case was levied against them).

The first point is there to protect them, in terms of liability, and gives them the ability to disassociate their legal entity from anything that could otherwise harm them (reputation wise, etc). It’s standard, and is pragmatic from a legal point of view; to not understand this point is to be legally naive to the implications of a corporation’s reputation, and associated guilt even through non-direct means such as providing access to commercial software (which they give you, in legal terms, as a privilege; not a right). To legally disassociate, and revoke access; to say a person using their software for creating a “game” that is intentionally carrying a zero-day ransomware, or something like a bitcoin miner; which causes damage to property, is legally disassociating themselves.

The second point is basically, “don’t pirate shit”, and is protection of their assets; again, legally pragmatic.

Say a large game company pirates the pro version for a large set of employees.

Now, normally one would just report it if they had tangential proof, but, with this clause they can get real proof; and press charges for the full worth that they lost in revenue with proof. It’s protecting themselves, plain and simple.

Also

>The Agreement is governed by and construed in accordance with the laws of Denmark

If shit goes down in the US, then the case would be held up in US courts with US laws; in some cases EULAs are thrown out entirely due to the incompability of the legal systems.

>>14556466

The thing about EULAs and who they’re enforced against is generallly corporations (legal entities), and are entities that matter (large losses of income, big business presence that they’re associated with); I.e. they’ve got bigger fish to fry than some nobody anon on a mongolian basket weaving imageboard.

Though I guess it makes sense that ‘’anons’’ would be a bit more paranoid than most, kek

>>14556466

Also intellectual property laws, anything you make is yours, and only the software/services that was provided initially can be “revoked”.

So, if you’re really worried; best to make a game framework with heavy inversion of control so it’s easy to port to another engine.

>>14556595

>To legally disassociate, and revoke access; to say a person using their software for creating a “game” that is intentionally carrying a zero-day ransomware, or something like a bitcoin miner; which causes damage to property, is legally disassociating themselves.

If they disassociate themselves from you (freeing themselves from what you create), does that same thing allow them to revoke your access to their product? Or is it just a way for them to say "sorry, we're not involved here"?>>14556595

>>14554972

Your game is already shit if its using Unity or UE4.

>>14556744

So there’s a few factors here: the jurisdiction it’s being enforced in, the compability of the legal system with the EULA, and the degree at which the legal entity itself decides to enforce their EULA (or other rights like trademark copyright, etc); I.e. it depends.

So in this case it’s mostly untread waters… so we need to use examples.

Take UE and Hatred as an example, they just gave a gentle “warning” alluding to the possible legal enforcement that was within their legal rights, and did so by publicly giving notice (so from a corporations perspective they were “reputationally distancing” themselves, without actually revoking access to services or being legally threatening; while saving face by saying “we explicitly don’t exert any sort of creative control or censorship over projects.”). In the end all it took was saying, “the video is using the trademarked Unreal Engine 4 logo without permission from Epic, and we’ve asked for the removal of our logo from all marketing associated with this product,” and that’s all that was pragmatically necessary to disassociate themselves.

In essence, EULAs are for big fish + pragmatic legal/reputation protections, and gentle coaxing words to an “underdog” developer are much better for their reputation than say threatening legal action via being serving the devs court documents and essentially by-proxy censorship/disproval of a creative work made using their software.

In the real world, especially for corporations that depend on the trust of their consumers (as there’s competing alternatives); has to be very pragmatic with these sorts of decisions. So, consumer/developer trust is huge; ruining that with a stupid move like attacking an underdog is reputation suicide, and completely unnecessary.

>>14556424

I'm guessing you're going for a silent hill type atmosphere? That first synth instrument is irritating as hell but I like the rest

>>14557967

They will replace whatever other value y and z will have at that point in the animation.

Keeping them ensures the animation plays corretly regardless of the object state

If you are playing two animations at once, one of them might replace those y and z values and make your animation play wrong

Although, it's unlikely to happen

I've been seeing a lot of shit going around urging everyone to learn to code, and literal calls to "democratize coding". Has democracy already ruined gamdev? There's not much money in it, so far I think it's kept most people out.

>>14556745

Even though my math sucks and I probably wont understand this, I'm saving it anyway.

>>14558255

That would make cents. I've seen gamedev companies whose owners spread terrible ideas.

>be thankful we're letting you work 60 hours a week for peanuts

>do you know how many people would love to work in game development?

>they'd gladly replace you

>by the way, we're always looking to be more inclusive and democratic so more people get into game development :)

>>14558281

please enable AA

>>14558281

Looks a bit blank. You could try looking at the final boss of Kirby Star Allies, it pulled off space pretty fucking beautifully. and it's fresh on my mind.

>>14558246

A large percentage of the population is in fact, on average, too incompetent/low-IQ to ever properly understand even design patterns; none the less programming. Programming is a set of skills that, at minimum, requires the faculties to abstract thought from a low to high level, and inversely; as to fully grasp what is it you’re doing.

Not to mention it requires dedication, and consistency; which is what a lot of people lack for what is considered a meritocratic driven craft.

>teaching the general populous basic logic

>through something that can make even math, which is boring to most, fun

Not a bad idea tbh, but as mentioned most will not pursue it as a career or hobby; nor will they possess the drive/faculties to do so.

>>14558305

>Not a bad idea tbh, but as mentioned most will not pursue it as a career or hobby; nor will they possess the drive/faculties to do so.

I hope you're right, but I've the masses of horribly low quality programming books (particularly gamedev, webdev, or javadev) makes me think that billions of pajeets are out there, waiting, for the moment when they can create an entire game by gluing unity assets together.

>not developing your web game as a java applet

>>14558305

For what is considered an essential skill set in a meritocratically driven craft*

>>14558314

how young are you that you have never seen the countless amounts of java applet games when the internet came about?

>>14558331

anon, I obviously know they exist, I mentioned them by name

Yo, MoM anon here, its been awhile since I posted. Heres a video of that Dragon summon from the official trailer I posted awhile back. All the dragons in MoM serve as optional bosses, who, when defeated, can be used as summons.

I tweaked his design a bit a tad too, to make his apparel more dapper looking!

>>14558452

Looking fun. What's your plan for the next demo?

>>14558474

>Looking fun. What's your plan for the next demo?

my plan for my next demo is to Finish the game this year

>>14558477

>actually finishing a game

I thought this thread was an elaborate joke like KOTH online

>>14558486

Actually, I am not far off from finishing. I have finished 9 maps in the last 2 months, and if I continue with the rate I am going at, I SHOULD be able to be done relatively soon, since they is only 25-30ish maps left!

>>14558497

You're going to gross $750,000 minimum, right?

>>14558505

>Selling this game

>The game I spent roughly 3-4 years on

HAH~!

>>14558511

hey, I'm sorry you don't believe in your own project :^)

>>14558516

Its more so I just want my first project not to be sold, is all. Plus, the sounds of a selling a TC mod, no matter how much you changed, sounds iffy to me. I rather just go the cave story route with my first game, and then go up from there!

I am not sure why people want to insist me to sell my game though despite me saying it should be free for the last 3 years

>>14558519

this does include boss arenas, which are circular in nature. Not all of the maps are big and sprawling, and even the bigger maps are Ocarina of Time styled, so they are not as complex as your typical doom map. Unlike a map pack, the emphasis is more so on the whole package then the individual maps. The demo I posted awhile back should give you a pretty good idea of such things, where some of the maps are the shops/bank/gallery, etc.

>>14558316

> playing kirby whatever

> I like the contrasting bright colors

you need to over 18 to post here

>>14558281

it looks pretty fucking good, but more like a mystical environment kinda thing

whats your goal there?

>>14558314

Pajeets are a special exception as their entire culture is non-creative, and they’re entirely stagnant (due to various factors like religion (~89% of the population of India follows Hinduism, a religion that promotes a stagnant way of life if utilized in a culture driven manner), way of life, teaching style, and overpopulation/resource distribution.

So, they can use it, but nothing of value is produced from their own creativity; so they do not create, they follow instructions, and emulate.

Also, programming is massive in terms of impact on society, culture, and way of life all across the world; it makes sense that there’s so much garbage, and opportunists. However, the cream always rises to the top.

I’d go into it further, but gotta get back to stuff.

>>14556301

Thanks, gonna work on the sandals now

>>14558751

tell us more, anon

>>14559317

The effect at the beginning was pretty cool, actually.

>>14559380

Aurora Borealis

If I'm generating Perlin noise, how much data do the vertices need? For 2D noise, I think I need just the XY component, and for 3D, I believe it's the XYZ component. Or do I need the Z for both 2d and 3d noise?

>>14559326

yeah, the sunflower would go pretty well on her hair, tbqh

but I agree with atelier thing, doing so much manually can be a drag,

looking for references should help though, just keep it simple

I need an opinion, since I'm moving blocks of similarity into shared IDs, what would be a better way to handle things here?

>create a block constructor class based on functionality

>static block, fluid block, moving block, ticking block

>create a block constructor class based on type

>rock (stone/cobble/mossy/end/netherrack), ore (gold/iron/coal/diamond), ground (sand/gravel/dirt), log (whatever type of tree), planks (etc)

>>14559512

I wouldn't make it OO in the first place

>>14559607

Yes anon, he'll just rewrite minecraft to not be made in java just to mod it.

>>14559621

that would honestly be a better use of your time

I can only find a single decent looking book on .net's IL. I might have to settle for a generic book on virtual machines, or delve into pajeetland

>>14559683

Who's the author?

>>14559607

Well I'm trying to reduce the object oriented bullshit as is. I suppose I could shove fucking everything into one class but then I don't know how that would impact everything else. I mean, how would performance be impacted if every block had the same constructor and just different variables passed to it? I would assume performance to tank because there would be more data per block held in memory, but I'm dumb so what do I know?

This isn't to say I wasn't considering doing that.

>>14559631

Yeah and it's shit because practically no functionality was put into the engine, it's easier just to put an API into vanilla minecraft.

>>14559721

>I mean, how would performance be impacted if every block had the same constructor and just different variables passed to it? I would assume performance to tank because there would be more data per block held in memory, but I'm dumb so what do I know?

Constructors are just functions, there's no overhead in calling them in that regard. However, note that it allocates a new object, which is expensive. So try to minimize constructor calls. Ideally, you should just have to initialize the data structure that holds the blocks in a chunk, then perform SetBlock() style operators, and never call any constructors.

>>14559738

If you have questions I'd suggest hitting up the Rimworld modding community's pardeike. The guy is an expert in IL.

>>14559738

>.Net 2.0

Wew

While I come from a C# background, any books by Jon Skeet or Eric Lippert in particular are worth reading. They'll mostly be focused on the C# implementation of things and how they work under the hood, but these guys probably have 3 decades of experience between them, and Eric worked on the MS compiler team, so he knows his shit. Jon is just an autist.

So your mileage may vary depending on what they write, but generally IL should be cross compatible between .Net languages, right? Like that's the whole point of it

>>14559733

>try to minimize constructor calls

well, there's some 124 blocks (not including those that are just subtypes of others), and most of those 124 all have their own constructor class

but those constructors extend a base Block constructor class

so I have no idea how many calls are made, but surely it's excessive to have so many constructors

>>14559892

>Marching cubes

>128 nested ifs

You don't understand what the fuck it does, then.

Given a block, you look at adjacency and set a byte value which has each of the 8 bits act as a flag for a particular corner. There are only 15 unique cases (33 if extended) and many are mirrors and translations of each other.

So basically, you take a byte, and transform it through a lookup table to find which geometry to use. Almost no if's involved

>>14559758

>>14559764

I'm not too interested in modding, I just wanted to learn a high level assembly language for fun. I might be stuck with the implementation documents, but that sounds too boring even for me.

>>14559781

The constructors are part of the bad design, but I don't think they're a problem in themselves. I don't know how JVMs typically lays it out in memory, but I wouldn't worry about the actual block creation code. More that it's creating a shitload of objects.

>each block array just has a list of pointers

>each block's data could be placed anywhere in memory

>have to call a function for every block

>all this is wrapped in a shitty virtual machine

The fact that MC even runs at a playable framerate amazes me.

>>14559892

The best option depends on the nature of your terrain.

For most types of terrains heightmaps are great. If you're dealing with steep cliffs (think grand canyon for example), then you use separate 3d models on top of the heightmap. You hide the seam between the two with bushes, rocks, etc. (or you use a special shader that uses world position for texture mapping).

If you have overhangs and/or caves, you'll run into the problem that Unity's terrain system doesn't support cutting holes through the surface of the heightmaps (at least it didn't back when I switched to UE4). In this case you'll have to rely on standard meshes for everything. Marching cubes would be one really simple algorithm with which you could generate the mesh. The resulting mesh will be far from optimal though, even if you do use a MC implementation that uses an index buffer. MC is something you'd want to use to either give you the base terrain for further editing in Blender or whichever tool you use, or if your terrain changes dynamically.

>>14559927

what if i wanted to make a randomly generated open world meme game? poaching simulator 2019

>>14559975

I would probably go with a hybrid approach. Use a heightmap to generate your terrain itself (there are many tutorials for this online).

Caves can be made as lazily as Bethesda does them in Fallout: A couple of pre-made cave entrance models which let you press F to enter. Their insides can be generated similarly to the outside world, although you would need a different algorithm than using Perlin noise to generate a 2d heightmap. You also wouldn't use displacement mapping to generate the mesh as you did with the heightmap, but marching cubes (or something similar). I think that would be the most straightforward approach.

Sorry that I can't be more helpful. Procedural generation isn't my strong point and I'm tired as hell right now.

>>14559975

>>14560071

You could also use portals for caves. Almost as lazy, and a lot better looking for the player. Just make sure to compare the cave area to the surface so you don't end up with non-Euclidean geography. Unless you're into that sort of thing.

>>14560453

i really just want pretty cliffs and didn't want it to look like skyrim. currently looking at some marching cube code off of the web, but i don't get any of it. i might just use mincraft cubes for now considering i got that to work, kind of. laggs hella lot more than the example project that makes some silky smooth meshes.

>>14559621

what? modding through javassist is the easiest thing in the world, I've made a couple of cheats for Wurm Online with that, like painting black in the minimap the tiles that the horse can't climb and Xray vision for mobs/players

And never going to share with you faggots because I lost and/or forgot where I put/how to install

>>14559912

Not a problem in and of themselves, but since I want to at some point get rid of these hard coded blocks and load them through a config, boiling everything down to a smaller set of constructors should simplify that process

eventually I'll get to how it stores world data, but that's pretty far off for now

>>14560783

I've never heard of javassist

>>14561087

>Not a problem in and of themselves, but since I want to at some point get rid of these hard coded blocks and load them through a config

Since you're stuck with objects, you should express blocks as a set of behaviors rather than form.

>destructable (x hits)

>spawns item (x ID)

>gravity

>emits light

So a (Lua?) file would have something that just defines everything on startup.

{
name = "SmoothStone"
attributes = {
destructible {health = 5},
drops = {CobbleStone},
texture = {xPos = 0.2, yPos = 0.2} -- Fuck I don't know I assume it uses a texture atlas
}

>>14561087

>>14560783

>javassist

Sounds like a japanese porn site featuring office ladies.

>>14561122

I'm not sure how you would do this without gang raping your cache, but going off of what you're working with, I'd set up a component based system.

>individual minecraft blocks still exist as objects

>initialize all block behaviors in the config file as their corresponding behavior objects (along with their data)

>a block notifies its behaviors on update, tick, etc

The main issue is that this is a lot slower and creates a lot more garbage.

There's a way to more or less completely mitigate this, but it's a lot of extra pain. Basically, having a static array of blocks for every type's data, and a block would just keep track of its references to its data types and its position in that array. It'd be incredibly fast because you'd be able to update all blocks based on their type, and the block objects would have almost no data in them, aside from a pointer and an integer. You'd have to deal with allocating, cleanup, etc.

>player hits a block

>find the block's health component

>reduce its health by one

>only one layer of indirection for infrequent updates

and the very frequent updates, like game ticks

>update all sand at once

>all sand data is tightly packed in memory, because you have a single array

>behavior loop goes through them, no extra function calls, just operating on data

>>14560783

Learn to read, I didn't imply that it's hard to mod java or anything else along those lines.

>>14561764

>gookspeak

>no korean in the image

Anon, you are getting the Asians mixed up.

>>14561122

I'm hoping to get a working file handler written with the json library already included, which handles player statistics, but that layout is pretty similar to how I imagine it being setup.

>texture

the 'texture atlas' is a 16x16 grid of 16x16 pixel textures, where position is set by one number where 0 starts at the upper left corner and increments horizontally to 15, then the next row at 16 from the left, etc.

I have to change that at some point too because hitting that texture limit will be a pain in the ass.

>>14561208

>gang raping your cache

It already rapes your disk with it's stupid stat tracking. With every stat it tracks, whether it be how many blocks you mined, the distance you traveled, how many times you died, etc, it reads your stat file to check and see if it needs to update, then updates it it needs to or creates appends the new stat to be updated to the file. I can't find anything that suggests the stat is just held in an array and written out on a timed thread, nope it seems like the stat page is updated every time the player does a thing tracked by stats.

RIP disk I/O

>>14562213

Are you going to try to change the way it handles chunk generation and saving? The method used last time I played was "generate chunk from seed and immediately save to disk, and overwrite it whenever a change is made". It seems far kinder to the disk to just store the changes made and apply them to regenerated chunks on the fly.

>>14562466

>anime autists

guilty as charged

>>14562213

>the 'texture atlas' is a 16x16 grid of 16x16 pixel textures, where position is set by one number where 0 starts at the upper left corner and increments horizontally to 15, then the next row at 16 from the left, etc.

That's still a texture atlas.

>It seems far kinder to the disk to just store the changes made and apply them to regenerated chunks on the fly.

That's what word used to do before they realized it was a massive security vulnerability. He'd have to create an optimized structure to actually store the individual block changes, and I'm not really sure what would be suitable. At some point, the player would make so many changes you'd want to revert that chunk back to the old system.

>>14562466

gamedev is the most degenerate act known to mankind

>>14562466

I watch various /co/ and /a/ things including the first season or so of MLP and some titty animay stuff, but at least I don't tell others about it unless I know them well enough and they care about those things too.

Funnily enough though, the biggest "open" degenerate I've met was a designer, he even used a Marie Rose tiny oppai mousepad at work.

Time to start on my new project

Stalker - Call of Blyat

>>14562466

If he's got a deviantart he's obviously an artist larping as a programmer.

>>14561790

young fags these days don't understand

>>14561764

gook is for koreans

chink is for chinese

nips is for japanese

moonspeak for the written language

good thing I've even found this https://en.wikipedia.org/wiki/List_of_ethnic_slurs_by_ethnicity

and dat some good drawings

>>14562466

real programmers don't produce art and don't own any art related faggy social media accounts.

look for github accounts instead

>>14562656

they are all slanted-eyed vermin anyway

still, sorry for the mistake

>>14562792

well clearly japs are honorary aryans, with the biggest gripe laid against them being that they sided with evil old hitler.and their biggest flaw, "how hard they got fucked threw the marshal plan"

>>14562656

>https://en.wikipedia.org/wiki/List_of_ethnic_slurs_by_ethnicity

grossly incomplete, but still a good reference

>>14562841

I'm more fond of Koreans myself, at least the non-thot ones that embraced christianity (30% of the population is pretty good for far east Asia), something that won't happen in Japan

>>14562862

hey the Tibetans seem cool too but this is getting off topic, but the idea of the group of X that abandoned their culture seems like a shitty group to judge a people buy especially considering how well the Talmud sold in worst Korea

>>14562477

There is one act more degenerate than gamedev, it is– and may Allah forgive me for saying this word– game journalism

>>14562879

So what kind of game is it?

>>14563063

glorified Minecraft but not shit i guess

>>14563180

I always had the problem of not enjoying the process but the product. I just beat myself up when I'm not making progress fast enough but when I do accomplish something, I feel like it's the way things are supposed to be as opposed to some sort of achievement. Not a good trait for a programmer, especially not when dealing with shit like networking.

>>14562477

Most of the people interested in gamedev I met, that have talents in other fields were your average normalfag Joe. This is clearly programming related thing.

>>14562545

>How the fuck do you even know these people have a dA?

The first guy was a special case. He told me he likes to draw so I asked him to show me some of his art He opened his dA and started showing off his pony stuff. He just didn't give a fuck. To his credit, the art was pretty good.

For most cases though, when I meet someone new I try to look for any emails or nicknames they might have written on their belongings. Simply paying attention during regular interactions is enough. For example some people use the same email for all accounts so that's easy if they send anything to you. Some sign their drawings with a nickname or write it when scribbling, they might use that particular nickname in an online game or write it in a filename. That's how I found the dA of the third guy, he was showing me something on his computer and one of the filesnames on his desktop caught my eye because it sounded like an edgy MMO name. Later when I searched for "[filename] deviantart" I found his dA with furry and mlp art. Some people do post their dA on social media, but those don't contain anything interesting, usually.

>>14563101

If what you just purposefully dodged saying was "Rust but with lovecraft" then fuck off and take your blogposts with you redditor

>>14563432

the fuck are you on about?

>>14562879

Sorry Anon, but gamedev is sacrifice, even if you complete your dream game, by the time you release it you will have more playtime through testing than most players will have in years after launch, also there will be no surprise or exploration because you already know everything about the game

Your only hope is that it becomes so popular someone makes a decent clone of your game

>>14563611

I know where you're coming from, but my favorite games are ones that I can go back to time and time again. While some of the lore won't be impactful to me, and the closest I could get to that is if I see others play it one day and see their reactions, but the game itself, the core gameplay should be something I could come back to and have fun. And if it's not, that's how I know that my game is a failure.

>>14563445

I see rottenredditors in my sleep

>>14563624

I was talking about people I met in person, not online, Anon. I guess I could've mention that at the start for the sake of clarity.

>>14562879

Do you even have any responsibilities, like a job or chores? Sounds like you're making your game all day everyday.

>>14563749

I finished HS two years ago and haven't done anything since. I'm a hermit and will remain a hermit for about a year until I move out and have to be a normal human being from there on out. My day consists pretty much entirely of working on my game at the moment though.

>>14563854

Use the time wisely.

I’m luckily getting a similar deal in a few days (final day of work soon) for about a month, but it took some planning. I took a term off uni, worked for a few months, saved up, and now I’m going back to uni after that month of pure dev time; as I took a term off as to be able to have at least of month of pure game dev while having extra saving to deal with other IRL money bs (as I’m a fully independent adult, unfortunately).

>>14563945

>>14563854

Good luck to both of you

>>14563920

me too, I had awfully slow couple of weeks in terms of development,

I blame it on interesting games such as KCD, procrastinating watching vidya and others things, and chilling out with a demanding gf

>>14563854

Thought so based on your post. Honestly, how much time per day do you spend actually working on your game and how much of that is lurking or shitposting here? It sounds to me like you can't handle having this much freetime and you got completely bored of your typical time killing activities but keep doing them because you can't or are to lazy to find anything else to do. Cut off the power from your computer for about a week to make your brain think of some gratifying activities other than the lowest hanging fruit.

>>14563940

"pol bullshit" is just regular /v/ posting, you'd know that if you didn't spend so much crossposting between here and reddit

>>14563945

I spend a lot of my time practicing things related to gamedev, I think I'm using the time well.

>>14564105

A lot of the day I spend on gamedev, and if I'm not doing that, I'm generally spending the free time learning artistic things like animation, sound, art, music, etc. or watching game critiques like the kind of stuff Joseph Anderson makes. I have a bit of a fixation on how much time I've got overall until I die so I tend to spend most of my time productively if I'm able to. Videogames are productive because what better way is there to learn about videogames but play them? :^)

>>14564143

/pol/ bullshit is /pol/ bullshit

>>14564271

we're constantly being barraged by cancer, we need the /pol/ to detect and repel reddit. They see racism and out themselves pretty quickly because their conditioning instructs them screech until the evil race realists go away.

>>14562466

Where and how do you meet them? This could be sampling bias.

>>14562879

show game and vagene

>>14564285

you're replying to the redditor

>>14564439

What if both are redditors

>>14564285

>we need /pol/

Speak for yourself. /pol/ exhibits the same kind of zealotry that you see in sites like Tumblr, just pretends it's not. We've warded off newfags with gore chemo for years and we can continue to do so without contaminating the place with politics

>>14564369

maybe sometime soon

>>14564545

>Speak for yourself. /pol/ exhibits the same kind of zealotry that you see in sites like Tumblr, just pretends it's not.

No one here accused /pol/ of compromising.

>>14564439

>you're replying to the redditor

Yeah, I know. They've already outed himself by screeching about /pol/. They just can't help it.

>>14564545

>we've warded off newfags

Except you are the newfag

>>14563653

Well if you add great modding support and it gets popular you could get a lot of fun from mods when you go back to it

>>14564996

I've thought about this, actually. not in much detail, since I need to have some functionality first but, since it's a Mario 64/Spyro-esque platformer, complete with world map and everything, would custom stages just have to be relegated to a menu of some kind? Seems a bit detached from the rest of the game.

My biggest worry is someone screwing around and putting some malicious code in their mods. I know it's the user's responsibility at the end of the day, but it'd still be unfortunate.

>>14565075

>malicious code in their mods

If you trust mods you're doing something wrong

>>14564813

Quality post

RR can’t hide behind anonymity because he’s too predictable. His posts always end up in the same pattern; self-pity, and self-loathing.

I cannot believe SFML has no built-in sprite batching options

>>14564813

What a horrifying idea.

>>14565207

The devs' attitude of 'one batching method wouldn't be ideal for every situation so we won't bother at all' is pretty shit.

>>14565237

Guess I'll be using someone else's solution, I guess I'm receptive to the reasoning but at the same time I sort of expect animation and batching in a 2D API.

>>14565207

>>14565237

>>14565362

It may come as a surprise to you, but the purpose of the Simple and Fast Multimedia Library is to be simple. It provides the most basic building blocks necessary to make multimedia stuff. Simple is the opposite of complex, where complex means made up of multiple simple things interwoven to create a new thing. In this case an example of something complex could be a program such as your game, or a system such as sprite batching or animation.

>>14563940

>why is it, that no one gets upset when thread is derailed by /pol/ bullshit

I like to think that the reason for this is that the average /agdg/ poster is smart enough to realize that "getting upset" is nothing but taking the bait, which will result in the thread getting derailed.

>>14563920

I'm sure you'll get back on track, Anon.

>>14565412

That's nice and all but I was criticizing the actual reason they gave for not including batching. Which wasn't that it was too complex to be appropriate.

>>14565412

Yeah, that's why I stated I'm receptive to their argument. I understand it has broader uses than sprite based videogames. If I implied it's defficient I meant to imply it's defficient for my use case, though really only marginally so.

>>14565412

Although SFML easily fits into your broad definition of complex

Anyone know much about Singletons? I want to use them to store things like the player's inventory, but I don't really know what are good practices with them and such.

>>14565690

The only good practices I can think of when it comes to singletons are these:

>only use them when necessary (that's a given, use the right tool for the right job)

>enforce them properly (only access through the getter function/property, private constructor, private instance)

>choose a naming convention and stick to it (I like to call the access function/property "Instance", so I get the object with class.Instance)

There isn't much to say about them, because they're really simple.

>>14565795

I see. Well I'll start implementing my inventory with it and see if it's a decent way to do it.

Just implemented a new block that's just named 'ore', gave it 4 subtypes, so the base block is gold, but subtype 1-4 goes Iron, Coal, Lapis, Diamond

I didn't add redstone because that has a lot of special constructors to make it tick and glow and sparkle shit. Since the client version I'm modifying is so ancient it lacks the ability to swap from creative to survival mode, so it's a bit cumbersome to test if these blocks will drop the right items.

>>14566106

>Just implemented a new block that's just named 'ore'

You're doing it wrong

>>14566217

I went over that earlier. It's a level of abstractness that's completely pointless. Those objects don't have "being an ore" in common, they have several behaviors in common; They have a certain amount of health, they drop an item when destroyed, and a texture. Some only drop items when mined. Some ores have an entirely different(!) behavior, like emitting light. You'd be able to represent any block in the game as just a list of behaviors, rather than a giant web of constructors.

>>14565690

What if you have multiple players?

What if you have multiple player-like objects/characters? What if you have to swap eg inventories or similar objects around?

Singletons should be reserved for system/manager-level things. Look at it this way: since OOP forces you to instantiate everything, why not make the object work in such a way it shouldn't care about how many copies of it are floating around? At the very least, it's not static, right?

>>14565362

>>14565207

What language are you using SFML with?

>>14566235

Write a tool to automatically convert all these definition files to java source files when you compile. Modders can work on a much higher level of abstractness and combine any base behaviors, there's no runtime speed loss, and it's a million times easier to manage. Defining the code for new behaviors becomes a lot easier and you can add them to existing classes without having to actually touch any individual block, just their high level definitions.

>>14566252

>>14566217

>>14566249

>What if you have multiple player-like objects/characters? What if you have to swap eg inventories or similar objects around?

Good point. I think I need to do some more planning for my inventory system then. I'm aiming to develop an RPG that relies heavily on inventory being the main mechanic for world interaction and combat.

>>14566252

>write a tool to automatically convert all these definition files to java source files when you compile

uh, why though? I mean the goal is to have minecraft parse some json files that contain block definitions and load those into memory on startup, having a tool to convert external files into java source during compile seems wrong

and also would be trying to run before I can crawl

>>14566251

Ceeplusplus, I've found a lot of options for sprite batching and animated sprites in the short time since I posted so I'm back to the game logic before I worry too much about rendering

>>14567055

i'm confused, didn't unity already support multithreading?

i could swear i used it ages ago when i was trying to figure out A* with my rts

how are the jobs different?

>>14567244

So since Unity uses .net we have always had access to the (now ancient) multithreading API from .net 2.0/3.5 and 4.0 with ported multithreading libraries.

So stuff like like threads, threadpools, and the like.

Actually, heh, I was one of the anons giving u walls of text of advice on why threadpool were better than your initial single threads approach, and how to optimize your AI with this.

However, that multithreading API is .net specific code which you have to wrangle yourself or find a library to wrangle for u; I.e. a manual, and generally tedious process requiring the construction of a multithreading framework to make it easy to maintain/use/keep bug-free/rapidly prototype while at the same time preventing race-conditions, deadlocks, and other issues like dealing with cache coherence on multiple cores (headache with shitty GOs/monobehaviours).

So, the difference here being it’s a fully multithreading framework provided by Unity, built in a very performant oriented manner, while exposing low-level operations for data, and it also auto-magically solves all the issues I mentioned.

However, it also comes with the requirement of using the ECS API (which is far superior anyways), and makes perfect sense as you can’t have the current job system functionality without the foundational ECS-DoD API/architecture.

Basically, it’s a new paradigm for unity in terms of performance, and the ECS/Job system go hand in hand; I.e. u can’t use one without the other as they’re mutually dependent in terms of functionality that arises due to the functionality one exposes which can then be utilized by the other.

>>14567362

Last part of mutual dependent became a word salad… I’ll clarify.

So u can use the ECS api alone, that’s fine, u get a performance boost over unity’s EC-OOP API.

However, this provides the needed functionality for the job system to do what it does, and can’t be formalized without the ECS API.

Furthermore, the total potential benefit of the ECS API can’t be fully harness without utilizing the resulting multithreaded systems offered by the job system; so there’s mutual dependencies in terms of fully harnessing the new feature’s functionality here.

>>14567362

>>14567055

>>14567381

Ok mr unity shill and now explain how the average hipster using unity is gonna benefit from this?

unity games will still be known for piss poor performance because no one can program for shit

>>14567362

>>14567381

yeah, i think i'll just wait for them to release a video tutorial on it

i'm not even sure i have any complicated algorithms that need multithreading right now

>>14567450

>Freudian slip

I don't think you know what that means.

If you shill for this shit at least have some arguments other than "omg its so great guise"

>>14567578

oh its the european refugee sim dev

>>14566249

>What if you have multiple players?

>What if you have multiple player-like objects/characters? What if you have to swap eg inventories or similar objects around?

These are good questions to ask, but not at the coding stage. Stop, answer these questions, and document them in a simple text file. Name the text file "design doc - rough draft". These decisions will directly affect the type and the complexity of the implementation you will need. Do not write a single line of code that is unnecessary or "nice to have". Take it from me. You will never finish otherwise.

Ogre3D has massive amounts of singletons in it's codebase. Check it out. Singletons are essentially global variables with a bit more sanity and initialization control.

> Look at it this way: since OOP forces you to instantiate everything, why not…

No. Do not look at it that way. OOP should not be _forcing_ you to do anything. You may be trapped in the OOP methodology. I suggest 1 week of assembly coding to break you of this dangerous mental trap. Write a recursive, parametric generator for trees or a easing animation core in assembler.

>>14567868

Nah. I've thought about counting bits and reusing variables for multiple purposes, and it just fucking kills the readability. Not worth it for my projects.

>>14566353

What's a good approach you've seen? I'm in the same boat for C#. Worst case, I'll just poach XNAs source and make it better

>>14567868

I concur, premature genericization can be just as much a trap as premature optimization.

thoughts?

>>14568284

The last few gamenights we've had, we played some D&D 5e-based boardgame where you explore a dungeon by flipping over tiles. Every tile fits together, and you orient them with an arrow pointing towards the previous one, towards the entrance (tree-like) and monsters are placed, and their cards have specific AI, like:

>If ~ adjacent to a hero, use [attack 1] twice

>If ~ is within 1 tile of a hero, it uses [attack 2] on each hero in the tile

>Otherwise, ~ moves 2 tiles towards the hero with the least HP

I thought it was kinda neat