/v/ - Video Games

Hahaha I've been bashing my head against Vulkan for two months now no I'm not sigmadev because he's not this retarded.

>>16045822

I too, aspire to not be retarded

>>16045844

All jobs have bullshit but it's important to make sure you're in a liveable state first. This is amateur gamedev, you can't live off a hobby, otherwise you wouldn't be in this thread

>>16045974

looks like a space ship

>>16045844

It Depends. Where I'm working now is rife with literally criminally incompetent management, but I can still go home at the end of the day and not constantly fret about what will happen the next day.

>>16045974

Not bad. Did you use any hard modelling plugins?

>>16046238

They're not Federal Buildings so it's fine ;)

>>16046352

But does it render red when you take a picture of you holding her hand?

>>16045963

Looks like shit. Those tentacles look like they have 100x the polycount of the room they're in. Why do they have sharp claws and a spike on the end? Do they drill into your fucking head? Why are they reaching into some bucket? Where does the pilot sit? It's ass. Do some research first next time. Get yourself references of other sci-fi head-mounted "neural jacks" or whatever, and also look up real medical equipment as well. For fucks sake, at least try to make a plausible, functional space that a human being could possibly inhabit. I'm not even going to bother with the lighting. Just add some real lights please, not just some emissive textures on the wall and a bumped-up blue ambient.

I can't tell by looking at anything what it's made of or how. Is the floor ceramic tile? Are the walls metal? Are those blue tubes ethernet cables or blue play-doh squeezed through a round hole? Your whole scene, as well as your weapon, look like dog shit, because you are a lazy cunt. You have put in the absolute minimum of effort to make a small scene with all polys and no detail, that vaguely resembles scenes you've seen in other generic sci-fi video games. That's it.

>>16046352

Woah, that's some next level TECHNOLOGY

>>16046352

Fan-tastic!

>>16046651

yeah but knocking out communications of any kind usually precedes a greater attack, if your american the CIA will fuck you over anyway under some false claims, you'll be okay as long as you are a big guy

>>16046951

So, Aquaria but in 3D?

I'm down for that shit.

>>16046951

That sounds like a good combination of elements. Free 3D movement underwater sounds like it could be fun if you get the physics just right, and everyone loves mermaids.

What kind of puzzles, though? Will she be flipping switches or pushing blocks? Or, god help us, raising and lowering water levels?

>>16047100

Underwater sudoku

>>16046951

>mermaid protagonist

As long as its top half fish bottom half woman im in full support.

>>16046352

But why isn't her whole body red?

>>16047467

> she looks like a Klingon.

What Klingon looks like that?

>>16045963

>>16045974

Looks really neat. Keep on doing what you're doing.

>>16047361

You can download an add-on called Tiled map. Alternatively you could use ResourceSaver to create/generate a new tileset for you by code.

>>16047515

I get exactly what you are saying but I disagree personally. The model is anime style because it is an anime character so there it is.

What I disliked was the girl you posted. You have bad taste so I don't see any reason why he should listen to you.

>>16047554

>Monogatari

>obscure

>>16047554

>anon posts his shit taste and demands others to cater to him

>acts pissy when people don't like his taste

Like pottery

>>16047601

Guess what, I also didn't know who she was until I looked into who that was. You should stop acting like a nigger in any case.

>>16047601

Hey filtered+ dumbass, you do realize 4chan was founded on anime and that 8ch was an offshoot founded on that plus politics? Go back to Reddit.

>>16047616

ew, whats that thing on her forehead and nose bridge?

>>16047654

How the fuck do you not know who Shinobu is? Is this your first day on an imageboard? Is that why you keep calling Monogatari old?

>>16047654

>Implying you wouldn't.

>>16047676

And it's still present all over the internet, including 8chan. How long ago was it that the wonderful nutjob married Shinobu on stream? A year ago? And that was advertised all over /v/ at the time, too.

By the way, Mr. wonderful nutjob, I've always thought that your Shinobu model looks a bit off and that her animations don't really suit her.

>>16047703

No matter your excuses, you're a fucking newfag among newfags if you don't know who the fuck Shinobu is. Do you also not know who Asuka or Rei are? Their original anime aired even longer ago, after all.

>>16047697

> I've always thought that your Shinobu model looks a bit off and that her animations don't really suit her.

Yeah, this. There's something slightly off about the proportions that seriously bugs me. I'm not enough of an artist to know exactly what, though.

>>16047693

>How do people even go about getting voicework for their games?

I have a feeling a lot of the no-name ones are grabbed from opera/theater groups.

>>16047703

>I never really cared who your autistic anime girlfriends are!

Now you've done it anon!

>>16047703

>Alien gas cloud who can heal all your ills and possess the body of any woman you find attractive

yeah, okay, good taste

>>16047520

Oooh perfect. I'm hoping to eventually get to procedural map generation so both of those are really useful to know about. Thanks.

>>16047729

I'm pretty sure any faggot on /a/, despite their gigantic egos, would recognise who the fuck Mario is.

>accuse others of sperging

>while repeatedly attempting to make excuses for being a huge newfag that doesn't even recognize characters that get posted daily

I'm done with your retardation.

To be fair, I haven't the slightest clue who Shinobu is. A few months ago I even thought she was Shinobudev's OC. I do admire his persistence though, and genuine animation skills.

>>16047100

> What kind of puzzles, though?

Probably traditional puzzles you see from old games like silent hill, the sort of stuff that requires the player to have a brain. Also presumably agility puzzles since movement would be a huge part of the game.

>>16047616

> 3DPD space orc

> attractive

>

>>16047817

>only attractive if you make them look like green elves

Really makes me think

>>16047827

<Only attractive if you make them attractive.

>>16047983

>The fuck happened.

Some bitter anon was shitting up the place.

>>16047983

Twas a joke, hence the flag. Also, well done on the camera system. Mighty impressive.

>>16047983

Was that VR wedding video the intro cutscene to the game?

>>16048587

It's a shame.

I like to use IRC when it's available which are usually comfy, and the use of IRC itself being a filter, but Discord is very useful as a tool as most games now have a Discord server as standard, and communicating with devs is as easy as sending a PM.

Rather than sending a formal email to an address that may only get a canned response a week later, you can informally chat maybe only a few minutes after a simple introduction.

I don't like Discord, but for gamedev it's very useful, and for a lot of teams (as they coordinate using it), required.

>>16048651

Whats the big deal? Just run the Israeli spyware in a sandbox with a fake email. Done.

I like IRC but dislike how people can see what channels youre in

>>16048651

>>16048587

I've never seen Discord used in a corporate scenario, it's always Slack or some custom thing.

>>16049287

Hensk, did anything ever come of the 'trains and bunkers' concept for the Orwellian/Diesel faction? Or will the engine, I think it was RA2, even play nice with units on rails? I remember the concept of the bases being built jointed together, like the Eurasian Dynasty in Earth 2160, with armored trains or tracked crawlers as semi-mobile forward bases. Is that still on your mind at all? Or would it come too close to the Huxley factions' Hegemony style mobile base walkers?

>>16050323

I should also point out I haven't gotten much work done on the project at all over the last month since I've been busy with other stuff, so don't expect me to be too much of a regular on here for the time being.

>>16050323

>>16050333

Neat. What influenced you to decide to make trenches crushable instead of crossable, if you don't mind my asking? Keep them from being too exploitable balance wise for late game, so you can't just load up a bunch of dirt cheap blockers with rocket infantry and a couple MGs for point defense to create an army eating tarpit? To keep the Command & Conquer steamroller mechanic spirit alive even if the poor bloody infantry are in a hole? Just a matter of whim? I suppose I should ask for clarification as to what qualifies as 'large unit' conceptually. Medium tanks? Standard heavies? Big dumb mammoths like the super-Char?

>I should also point out I haven't gotten much work done on the project at all over the last month since I've been busy with other stuff, so don't expect me to be too much of a regular on here for the time being.

I'm just happy to see you're still kicking now and again, scrolling through the front page. I don't actively keep tabs, so it's nice to be reminded to poke my nose into your galleries now and again. When you left off on the older 3D model renders and didn't come back to it for a while, I thought you'd dropped the whole idea. Voxel dorfs was progress to me before, so's trenches now. I'm sure you'll get at it all in good time.

>>16050405

>What influenced you to decide to make trenches crushable instead of crossable, if you don't mind my asking?

There might be a way to make them crossable by tanks specifically, but it might require some programming, since it's not something I could just do in YAML. Making them crushable is just a balance thing. Trenches will be immediately available for construction as one of the first things you can build, but to avoid them being a death trap for higher-tier enemies they can just be crushed by tanks.

> I suppose I should ask for clarification as to what qualifies as 'large unit' conceptually. Medium tanks? Standard heavies? Big dumb mammoths like the super-Char?

Pretty much what qualifies as an infantry crusher in normal C&C. So buggies, trucks, cars, etc can't crush trenches, but tanks can.

>I'm just happy to see you're still kicking now and again, scrolling through the front page. I don't actively keep tabs, so it's nice to be reminded to poke my nose into your galleries now and again. When you left off on the older 3D model renders and didn't come back to it for a while, I thought you'd dropped the whole idea. Voxel dorfs was progress to me before, so's trenches now. I'm sure you'll get at it all in good time.

I originally dropped it mostly due to losing my NEET status and having to wageslave. I got back to it recently because after inheriting some money due to some unforeseen events, I managed to get into a lifestyle of just working from home doing online commissions, and I was able to quite my exhaustive retail job (admittedly I'm scraping by in terms of what I'm actually earning), and also while I've been aware of ORA for years, it only occurred to me semi-recently it's almost the perfect engine in terms of what I want.

I think there is also a bit of a mental aspect to it too, since I felt like I wasted my life not really doing much with any of my artistic ability after going to college for several years for it, so constantly drawing/doing 3D and gamedev stuff now seems like a way to vindicate that and realize what I originally wanted to do with my life.

>>16050454

>constantly drawing/doing 3D and gamedev stuff now seems like a way to vindicate that and realize what I originally wanted to do with my life.

I'll risk sounding presumptuous in saying it sounds to me like you've got a lust for crafts, with no greater complexity to it. Making tangible, useful, and/or beautiful things. In which case, stay the course in the best median of monetization & net loss fulfillment and you're probably golden. The best way to kill the shiftless feeling I'm reading out of 'realize what I originally wanted' will be to settle in, hone your already quite reasonable skills and intensify your output. More tanks, dorfs and fat anime tits. Can't go wrong. Maybe dabble in a skilled trade of some sort on the side that can key back into the arts and give you non-retail prospects at the same time, if the commissions ever dry up?

>>16046000

nope

>>16047520

thanks anon

>"Where is the previous thread you faggots let it die" edition

Last thread was >>15993425

Managed to archive just in time.

>tfw demo day in a few days and we're not ready yet

>>16052060

It's not like the guys and I don't have anything, but not as much as we should have.

>>16052666

looks ok but don't make the default colour of the square that ugly yellow, make it something neutral like grey or maybe blue or green those are eye pleasing

>>16054280

Its not something thats being fired in a tight loop thousands of times every frame, nor are you allocating new objects with it. Just maintain a collection of delegates that point to handlers for each behavior.

>>16054431

Is there any particular reason you're using MS3D and not, say, Blender?

>>16054509

It is what is used to export the models to the half life .mdl format, along with the recommendation from other anons/tutorials which recommend using MS3D over Blender. I tried using a Blender add-on which supposedly exports and imports mdl files, but that seemed to only work with Quake, not Half Life. MS3D has the ability to export to a format that works with Half-Life, with less fucking around so that is why I chose it.

>>16054280

there's nothing wrong with a switch statement, I doubt turning it into a pile of OOP with a dozen abstractions and virtual pointers will make it easier for you to understand unless you breath pajeet

>>16054280

Don't fall for the OOP meme, it can become a nightmare. Always use you own judgement.

>>16054531

MS3D can import and export a huge amount of formats. You might want to experiment with exporting to some other format and then using MS3D as a converter.

>>16054876

Could do so, I am investigating further on what other developers have done. Also recently found out that the devs of Half Life: CAGED allow you to download and mess around with the levels in the mod, so that should help a lot with the development and I will be going over their developer commentary to see if I can learn anything new.

>>16054936

It's manageable but only because I spend 75% of my time making sure the code doesn't become an unmanageable mess.

>>16054431

Dude, I cut my teeth on Milkshape3D. I was forced to learn Blender for university and realized I didn't know shit about modelling after many years of using MS3D, and I had to relearn it. Worse, as I fancied myself a good modeler, I had made a bunch of models in it for my game. I had to completely redo them in Blender afterwards. There was nothing to save.

MS3D doesn't have what you need to make models effectively. It lacks the tools and complexity. You spend time doing stuff better software solved and automated years ago. Its workflows are not analogous to real world modelling processes. Get away from it as quickly as possible. You will save hours, trust me.

Use Blender, pirate 3DSMax, whatever, just don't use Milkshape. It has nothing to teach you.

>>16052135

we gonna make it, bro

>>16054936

I have less than 2000 lines of functional code in my engine. A lot of it is fluff

I'm trying to make a generic inventory/container system that works via grids. The basic logic of it is simple enough, but I keep feeling overwhelmed by how to design it properly and all the corner cases.

Things like splitting and merging, how to account for adding an item at a specific (xy) spot, versus automatic insertion, even what data type to use (array vs list), etc

>>16056174

look into 2d bin packing.

sounds like you just need a tree structure, possibly some quad-tree routines.

>>16056199

Bin packing isn't the problem (and I've implemented that before). With an ARPG inventory system, you can give each item an XYWH/bounds and check for overlaps, or store the occupied tiles. Each tile has an implicit index as well.

I want to try making a simple multiplayer game right now so I have to implement it carefully

>>16056211

sounds simple, you could easily use a tilemap were each tile references the item's id and surrounding bounding box, then insertion becomes like tilemap collision.

or just compare the inserted item against a list of all item bounding boxes.

automatic insertion would be brute forced by attempting insertion at every point in the grid

>>16046352

That's the way it's usually done. For instance, I am quite sure that this is how they draw the orange selection outline in the scene view

>>16054280

>As in, it would surely be easier to read as a single CURRENT_JOB_BASE_CLASS.update(). But is this really worth the overhead of abstraction, memory fragmentation, virtual tables, and object orientedness? There would never be more than 1 instance of each job type and memory

This is exactly the type of situation you want to make it easier for you to code.

The only places you want optimal code is where it's actually a resource bottleneck or a hugeass look.

For instance, my map system is a bagazilhon abstract classes with a lot of reflection to automagically find these classes implementations, my map generation code is pretty well optimized and my grass placing code is ultra optimized.

If it's not critical, or could become critical, don't bother optimizing and just do what's easier to read and code

>>16056318

you literally go line by line, tile by tile, checking if it is a valid position

it will take nanoseconds to finish.

unless you are making a mmo where what people will be doing the most is managing inventories, you will never have to worry about this being too expensive to compute.

Also, this expensive computation can be done only on the client, and the server just verifies if the client got a correct answer

>>16054936

My engine is 3000 lines of compact code. It's going to blow up once I start adding levels. I keep mothballed code in a separate file which is 500 lines.

>>16056318

>Also, this expensive computation can be done only on the client, and the server just verifies if the client got a correct answer

And that's what I'm trying to figure out. I've never done networked stuff before, and I'm trying to basically figure out the server validation, which needs to be as lightweight as possible, and figuring out what needs to be offloaded to the client

>>16056449

client sends 2x4 item at position 22

validate 22, validate 23

validate 28, validate 29

validate 34, validate 35

fail 40, out of bounds

that's all that there is to it for the server

the loop is super small and validations could be against a bool array that stores which tiles are used and which are not

>>16056452

Kudos for your progress. Do remember gamedev is engineering and therefore is more about results than autism.

>>16056452

I'm using SFML instead of writing my own renderer which saves a lot of work. I hear you on the visual product looking the same. It's harder to see progress. It can be discouraging even for the developer.

>>16056510

I have some results, but not anything that I already posted. Eventually it will be very interesting, but since it's an engine, its taking such a long time…

>>16056519

SFML acts a lot like OpenGL 1.1, at least from what I have seen. I've had to use it before at one point, it all seemed familiar to me because it was all functionality that I had to implement at some point or the other, it's interesting.

My engine has looked the exact same for months now, it's driving me crazy.

>>16056668

>>16056519

Note that if you're using the official C# SFML binding, Vector3 math isn't correct

>>16056509

Like this this a rough sketch/implementation of what you're describing and it already feels very wonky to work with

https://pastebin.com/raw/4sk6Ah1C

>>16056805

You got all the tools needed to complete the objective with a performant-enough code. What's to complain about it?

Is it because you need to properly ensure that the occupied array never gets wrong info?

Just make sure that all operations on it are atomic, meaning, there is no such thing as "replacing" items, only "removing" and then "adding", so even if you drop an item in the corner of another, the replacement is easily implemented.

You will be unblocking a tile you are about to block again if there is overlaps when replacing items, but unless you are aiming at 100000 fps, this is not what will make or break the performance.

>>16057027

>meaning, there is no such thing as "replacing" items, only "removing" and then "adding", so even if you drop an item in the corner of another, the replacement is easily implemented

See but now I realized that if I implement the "drop on item to swap them" behavior, I have to have a way to check to see if lands on exactly zero, one, or two or more items, which means having a bool[] is too naive, since it only tracks that IS an item, not which one, which means that every container in the game is going to have a massive number of int32's associated with it to track item pointers

>>16057204

Return the slot index if the placement is valid, then use that to lookup the item occupying that slot. If the placement isn’t valid, like there being multiple items, return -1 instead and treat that as a failure. With that said I think you’re needlessly worrying when you probably don’t know how much the system needs to even scale yet.

>finally get time to dev

>spend it redoing my character basemeshes instead of making any progress

I guess they look nicer, but the autistic urge to constantly revise things is going to be the end of me.

>>16057637

That's a wide subject. Do you mean minecraft-like procgen where eveything is randomy made from noise maps or diablo-like with random assembly of hand crafted rooms?

>>16057637

I don't know shit about this game, but look at the Biomes section

https://curious-expedition.fandom.com/wiki/Modding:Home

>>16057637

This was brought to my attention a while ago and I've been holding onto it. There may be something here that can help you.

http://pcg.wikidot.com/

>>16057681

>Oh yeah Perlin Noise is a thing

>That's been superceded by Simplex noise

>It's covered by a software patent

>Owned by Nokia

https://patents.google.com/patent/US6867776

Retarded

>>16057665

Both, eventually. I need natural-feeling biomes and room-based dungeons inside those biomes.

>>16057675

>>16057681

Sweet, thanks.

>>16057705

Is that patent expired? Didn't Minecraft use Perlin noise?

>>16057889

Perlin was never patented, but the improved Simplex noise was

>>16057889

The patent status is marked "active", so no it's not expired. Patents tend to expire 20 years from the date of issue.

Perlin noise has been around since the 80's. The patent describes Simplex noise which is similar but with some improvements.

>>16057246

I used hammer for all of my levels. I couldn't comment on any other editor because I never used anything else. But I didn't feel that hammer was holding me back. I might have felt that way if the other editors out there are really better. It's worth a try, at least.

>>16057927

>>16057705

Who'd license a marginally better noise algorithm?

>>16057941

>Be from an A studio

>Could pay a hefty fee for licensing the patent from the owner

>Or have a subpar / impossible implementation that just wont work or be innovative

>Welp can't make procedural textures in our game, it has to all be pre-generated on our workstations instead

>Massive bloat

This kind of thing is more common that you'd think

>>16047361

Someone made a shader based tileset that only does visuals but is much faster: >>16057705

OpenSimplex

Speaking of random gen, does anyone know what the Dwarf Fortress guy used? Did he ever explicitly say?

https://thebookofshaders.com/12/

This is so cool

>>16058129

That really is awesome, thanks for sharing.

>>16054390

That's a really good idea thanks. I should have thought of using delegates.

>>16056402

It's not about premature optimization. The problem is trading off bloat for code readability. Bloat, in this case OOP, which can eventually lead to performance. Even if there are innocuous changes one wants to make, he should be neat about the overall design and never prioritize structure over readability.

>never prioritize structure over readability

Well I take this back since there are always exceptions.

>>16058129

>book of shaders

I worked with the guy who wrote that book. He works at Facebook. Not the nicest of guys in person unfortunately.

>>16058352

Oh yeah I remembered what I was going to write now

It's the same Item instance between the player and the world. Each Item basically has XY and Type, which contains more information (name, behavior, icon, width, height, etc). Eventually, it'd be nice to have the items placed anywhere within a tile, not just snapped to the middle

>>16045963

Sexy colour scheme/10

>>16057705

Heads up that its 3D simplex noise and higher dimensions that are patented.

>>16057204

replace the bool with a byte indicating the item slot in the list of items +1, 0 being empty and 1 being slot 0 and so forth, or a short if there will ever be a big-enough inventory. That would, however, need to be rebuilt every time a item changes, or you could use a sparse array for your items that keeps track of holes on it

Also, you are over optimizing a non-critical (performance-wise) system

>>16058246

>The problem is trading off bloat for code readability. Bloat, in this case OOP, which can eventually lead to performance

I can only ever see OOP being bloated code if you put literally all your classes in a single folder.

You would have to be very incompetent to make unreadable OOP code and readable structured code, as OOP is made for humans and structured for machines

>>16058354

use XY for tile position and xy for subtite position (preferably give better names for the variables).

The item is still `owned` by a tile, but that way, it can be placed anywhere as far as the player cares, especially if you can place more than one item per tile

>>16048587

>no one uses IRC because muh images

I am using it but I have no idea what to write there tbh.

>Riot

Is there a AGDG Riot/Matrix room?

>>16058814

There's one at https://matrix.to/#/#agdg:matrix.ordoevangelistarum.com , although I dunno if it's restricted to OE members or not.

>>16058721

That's what I was leaning towards and I can ignore the subtile bits when it's in a container that's not the world. I might be able to get away with using bytes for most things instead of ints if I chunk my maps appropriately too

>>16058696

Bools are compiled to bytes anyways, even bool[]. I have a bad habit of over optimization, yes

>>16059752

>Bools are compiled to bytes anyways, even bool[]

std::vector<bool> actually stores bools as bits. This has some implications for how you can access them, but it's an interesting thing to remember in case you for some reason require a fucktillion bools at some point.

>>16058716

>I can only ever see OOP being bloated code if you put literally all your classes in a single folder.

Fucking what? Do you know what code bloat is?

>OOP is made for humans and structured for machines

OOP is made BY humans and is not made for machines. OOP is abstraction which is by principle counter intuitive for machines.

>>16058716

The problem with OOP is that data and functions are contained together in classes.

>>16060173

funny enough, in Java, using bitwise operations on int[] is faster than using boolean[]

>>16060345

>Fucking what? Do you know what code bloat is?

Turns out I didn't knew, I was assuming "complicated and hard to read" rather than "too many repeated letters"

>OOP is made for humans and structured for machines

I was splitting the two methodologies as OOP and Structured coding

I am not saying that OOP is structured, I am calling 'non-OOP' as Structured

>>16060374

What's the problem with grouping related things together?

I am not saying that OOP is performant, even, I am stating that OOP is only ever good for when performance is secondary to ease of writing the code, which is 60-90% of the code depending on the project (If you are a engine fag, then I guess it's 20-30%)

>>16060409

>>16060173

Sorry im working with C#. I can see why Java would behave that way since C#'s CLR is optimized for the common case, being Int32. That is, all bitwise ops return an int32 even if youre working on bytes, which means casting.

>>16060409

>What's the problem with grouping related things together?

I can't find that website some anon once linked here. It made the case beautifully. Hopefully someone knows it, it was filled with the authors issues with different dev things, some of them humorous.

Basically, if you may want to use functions with different types of data. In OOP you'd have to make a hierarchy of inheritance which can become horrifying. If you end up branching that hierarchy you could end with two "leaf" objects and want one to use a method that's only available in another branch, thereby requiring copying code and defeating the whole point of OOP.

>>16060528

>end with two "leaf" objects and want one to use a method that's only available in another branch, thereby requiring copying code

or implementing the method in the nearest common parent class?

I guess the website can give better examples, but this seem like a case of bad organizing rather than 'OOP is bad'.

The best part of OOP, Structured and ECS and whichever else paradigms are out there is that neither of them force all of the code to be made in neither, you can always use the best tool for the job rather than forcing yourself on using exclusively one of them in your project

>>16060409

>rather than "too many repeated letters"

It isn't too many repeated letters, it is too many unnecessary operations.

>I was splitting the two methodologies as OOP and Structured coding

Which is nonsensical. OOP is just a brain damaged subset of imperative programming. Structured programming is the only kind anyone does. It says "create languages with constructs like "if" and "while" instead of rolling them yourself. Every OOP language on the planet forces you to use structured programming. There are two paradigms, imperative and declarative. OOP and procedural are two ways to do imperative programming. Functional and logical are two ways to do declarative programming.

>What's the problem with grouping related things together?

It isn't grouping them together, it is coupling them together. This forces horrible shit like inheritance and prototypes. This is why you have ugly and fragile "design patterns" like this: https://en.wikipedia.org/wiki/Composition_over_inheritance to try to work around the language intentionally crippling you. In a functional language, composition is natural and simple and used everywhere, no "design pattern" needed.

>>16060555

True, but often people take OOP too seriously and many universities will inculcate students on it.

>>16060666

>Like what?

It depends on the code. Look at the disassembly of your binary and see for yourself.

>I am pretty sure the only operations being run are the ones I program myself or those I willingly use from the libraries.

If you think something that naive and silly then why are you arguing? Learn something before trying to argue about it.

>Are you talking about shit like safe guards in arrays so that

No, that has nothing to do with OOP.

>Good thing you are not forced to use it, regardless of the language you use

Yes you are, and that's not an argument.

>I could see a math graduate using that, but making a game on it would be living hell

Yes, I can see how games, being giant piles of math, would be very hard to write in a programming language that you think a math graduate would use.

>dude, just make the main class and don't let your dreams be dreams

That does not solve the problem:

<this car has a broken transmission, it is no good

>lol dude just don't use the transmission problem solved!

Imperative programming is defective by design. You can't just decide not to use some of the features and that will make it ok. You need a language with good features, not one without features.

>Make a 300,000,000 lines single file code in the fucking main method if you so fucking wish

Are you retarded? What does this idiocy have to do with anything?

>the only requisite is the goto statement being available in the language

The only requisite for what? Gotos do not help at all. "Oh if OOP is bad just do something worse!11"

>If you don't like it, don't use it

I don't use it. Because I don't use broken languages. That does not make a broken language like java or C# acceptable.

>You are free to use when convenient and not when not

WTF?

<waiter I ordered a steak, why did you bring me a big smelly turd?

>you can just eat around it

<eat around it to what? it is just poop on a plate

>if you don't like it don't eat it!111

<I want a steak not an empty plate that smells like shit

>omg stop autistically complaining about the meal because it can only do one thing good

Taking the turd off my plate gives me an empty plate, not a steak. I want a steak. OOP does not do one thing well, it is always eating shit.

>>16060666

Oh yeah, and why did you post a picture of google telling you the same thing I did?

>Use a matrix transform to convert mouse/screen coordinates into game tiles

It works but this feels awkward. For example, it depends on where I render the game window, and it's possible that it could go into negative coordinates, so I need more validation as well

>>16061679

I guess it depends on how much time you're going to spend making assets and levels. Doom can get results that look good way faster. But goldsource is a way deeper engine in that area. If you're not going to put in the amount of time that's needed to get to that point, doom is going to be a better idea, I don't mean that in any kind of negative way, it's just the nature of these tools.

>Demo day is 3 days away

Well, how many projects do we have this time?

It almost feels like it's not worth a sticky

>>16061808

Doom has the advance of automap which is a awful rare feature in FPS engines so if you like doing big maps or something with heavy amount of puzzle it would be the perfect fit because the player can always check his GPS position.

Whatever happened to that anon making that pastel colored tank game?

>>16062128

I don't know much about it technically speaking, just that that's what ORA as an engine uses for vehicles.

Actually technically ORA uses sprites, but it can ALSO use voxels, with the advantage being that they can rotate around like a 3d model. Or at least, that will be an advantage whenever the ORA devs get to actually implementing voxel rotation beyond the lateral axis. Currently vehicles can only rotate to change directions on the map, but don't actual tilt forward or backwards when going up slopes, which is kind of funny looking.

Hopefully the devs get to actually implementing full rotation soon.

>>16062059

>>16062060

The heavy is like the weird bastard of a Char B1 and the Neubaufahrzeug. I love it. Especially the giraffe neck turret ring and the sponson gunshield being as externalized as physically possible with a gap you could stick your head in. A million awful little impracticalities that are neat to chew on. I like that it has ball cheek MGs that are not in any way synchronized with either main gun, because the gunner(s) and loader(s) really need to be able to independently traverse and mow down stray infantry while ignoring their primary firepower. The stepped pseudo-Turan III turret is an unexpected touch. I suppose the commander really needs the extra two or three feet of room to sit above the hull crew, plus it's a good excuse for a redesigned turret upgrade with some KV-2 derp guns. Or maybe that's where they started, and the top front of the turret was hacked off when the guns were downsized for reasons, but someone lobbied to keep the commanders' position high up because 'muh view range' and reasons?

What's the plan with all these tankettes, though? Deciding on a final variant, modeling them out as an exercise, or are you thinking you might actually have that many options for giggles? You've got a Bren carrier, what I'd guess is an infantry gun, AT rifle/light cannon carriers, with or without MG or overloaded light tank stubgun, and a midget pseudo-Hetzer. Four-fiveish units out of one chassis. Now, these could be purchased upgrades on individual units or through research, thus not bloating the unit build panel, but even then the issue of unit recognition might rear its head.

Personally, I'm a sucker for AFV variants. I like the granular changes that can develop over time. In the context of Dorf RTS, though, I'd say it would be wise to put some mind into what you need gameplay wise as you go along. While it's obvious you can crank these rough models out very fast and polish them about as quickly, don't bite off more than you can chew by trying to overstuff interesting voxel models into a unit tree without a definite aim (or existence) yet, especially since there's two more factions that'll also be contributing to the balance metadesign load later. Make sure to do some case studying in other RTS on what units each faction will need and where to put them, get some basic big picture paradigms set, maybe figure out some objectively balanced cost-to-stat ratios which can then be skewed up or down, and then consider the biggerer picture of granular extra/gimmick units, the kind you'd have in an expansion pack or a free patch.

Alternatively, I could be talking out of my ass high on pure nodev theorycrafting, and you can just wing it. Maybe having a Swiss army knife of purpose built units can be an Orwell faction gimmick, the same way the Mad Max/Golding raiders have oodles of scrap upgrades. Still, don't force yourself to find a place for placeless things if you get carried away modeling. Games tend to have scores of concept assets that are neat but went absolutely nowhere, and sometimes it's because those neat things would've been clutter. As an example, take modded up Supreme Commander; could you honestly say you make full use of the potential triple digit unit variety, as opposed to building mostly vanilla (for familiarity) or mod units (because you have them and may as well, and they do something useful?)

The answer is probably not, because there's a lot of units with a lot of inherent and situational redundancy. You will therefore pick the workhorse or the novelty whizzbang units. Slightly gimmicky almost-workhorses, like a fragile but good DPS sniper tankette or overgunned overpriced almost-light tank that comes a hair earlier in the tech tree than a real light tank but is an objectively inferior resource sink if it isn't used to rush in just the right way, might get lost in the cracks if there isn't a good balance for them to key into or subvert in the first place. Plus, having a unit for practically everything in one faction but not others might give them too much counterbuild ability, and if everyone gets it, the C&C style faction distinctions might blur a bit, to where they are fittingly more Total Annihilation/Supreme Commander like.

>>16062552

There's also Atomontage engine, in development since forever. Looks pretty.

https://www.youtube.com/watch?v=1sfWYUgxGBE

https://www.youtube.com/watch?v=LE3bjUrhXWk

>>16062744

Actually here's a more recent reel of the engine.

https://www.youtube.com/watch?v=nr5JqYYye3w

>>16062749

>>16062744

Oh right, I forgot about that one! It really shows how much potential this stuff has, and how few projects use it- only hobby projects similar to the ones made here are exploring this technology in a serous way… everything else just uses very trivial applications of the technology without exploring new places. It's all very interesting stuff.

>>16062753

I don't know man. Atomontage looks gorgeous, but they haven't done anything beyond tech demos and PR, and I remember they were advertising like crazy trying to find gaming applications of their tech years ago. I think high res voxels just isn't practical. Too much memory.

>>16062765

The real problem is that all of the games companies that work on their own engines, etc, have built these content pipelines around traditional models of rendering graphics, mostly because all of their third party tools, etc, are also built around this, and just because its easier… I don't think it's something out of reach for our hardware… just a logistical problem with software. We have more and more memory right now, and all of the tech demos that we look at were made years ago, and still look rather good, its just an issue of the toolchain and engine not being made around it, so few companies actually innovate with the game engine so it doesn't happen a lot..

>>16062765

One of the most promising applications of voxels is voxel-based global illumination, where you voxelize a scene, and do realtime bounce lighting on the simplified voxel version, then extrapolate that information back to the high polygon scene.

>>16062780

True. Nvidia is working on a voxel based gi solution, which no doubt will be purposely incompatible with AMD GPUs.

>>16062671

Those, plus the T-28 and T-35.

I didn't even notice that. The front hull step angle and rear radiator from the T-35, plus the crummy, tiny vision port. The zit turrets are almost the T-28's. I suppose I need to step up my tankspotting.

>I am thinking of adding a third mini-turret behind the driver's hatch but in front of the big turret, though that may be too much clutter.

I think it might be, if only because it could end up being too indistinct a detail once the model is voxelized and turned into a 3D pseudo-sprite. I'm actually curious as to how all these extra guns will be handled; does open RA's current codebase or any using it allow for independent weapons or multi-turrets? I only remember there being one slot for ground and air attacks in early C&C with no actual turrets.

>the way I want this faction to work is that every vehicle has a single, dedicated factory, rather than the faction getting a war factory that builds everything. The idea is that there will be three main ground vehicles: a tankette, a half track, and a heavy tank.

I think I remember you telling us this a year or two back. I'd outright forgotten. I think I forgot to ask even then; what happened to the old Light, Medium, Heavy, and Advanced Medium in this framework? Thoroughly obsoleted as too rational, since you're going hard on interwar tank anti-design? Or some of them rolled over to a very expensive late game multi-unit advanced factory or two, for if someone fucks up enough to let the Imperials get to late game but not bad enough to get killed by units from the main six war factories? Sort of like R.U.S.E and the Prototype Base.

I suppose the tier frame there would be Base Factories-Upgraded Capacity-Standardization Program. Start with lots of options, and still have those late in the game, while also being able to streamline with a 'normal' C&C war factory with a new lineup entirely in the late middle or endgame? Reinvent the wheel with WW2 tech, then even the odds with Huxley by getting some early cold war tech; monogun battle tanks, first or second gen jets, elektroboot bootleg ballistic rocket subs. Just as long as Huxley don't jump ahead in their own very expensive end tier with X-ray lasers and atom bombs.

>I haven't given much thought to their navy

Well, I have to wonder if you've given anyone's navy much thought at all at this early stage. I imagine that the Huxley-Cold War boys would be easy to pin as having good submarines, hunter-killers and boomers, as well as destroyers and missile cruisers. Advanced, long-range weaponry, good AA cover, middling durability. Some are cheap, others are expensive, all are reasonably cost effective and tend to be best utilized out of enemy LOS. Golding-Mad Max would have weird shit like dirt cheap fireships, Waterworld tankers with ungodly transport capacity and maybe gimmick production, and torpedo boats with far too much anti-capital firepower that'll capsize from a stiff breeze of escort fire. Maybe a double-submarine that can drill through land and surface on treads in addition to oceangoing, like an Atoragon made of spot welded scrap metal held together solely by fervent prayer. Wild cards that tend to die trivially, especially if misused, but earn back their cost or more if leveraged properly.

Therefore, Orwell-Imperials by process of elimination logically need to have the most expensive on average, high HP, mid-ranged loadsadakka diesel ships on the order of Deff Skwadron's 'Da Grimlug,' with everything being a little oversized and overarmed for their class- destroyers are recon cruisers, line cruisers are pocket battleships, battleships are Yamato scale pagoda messes- with increasingly spotty AA cover as size increases if you don't buy some refits like dual-purpose batteries, and increasingly passive & optional sub defenses, say from depth charge attacks, to layable mines, to buyable torpedo bulges. It even fits thematically, since 1984 did explicitly mention 'floating fortresses' as one of the militarily impractical to downright useless outlets for human effort and industrial capacity.

For variants, I suppose escorts could be turned into floatplane tenders for cheap recon & air attack or torpedo destroyers for glass cannonry, line cruisers distincted between high gun count AA farm American cruisers, meant more for dealing with the other two factions, and heavy panzerschiff with frontloaded damage cannons/torps for mirror matches, or escort cruisers for single engine fighters, with capitals either being dakkaships or ridiculous fleet cruisers that can load up on twin-engines. And subs are subs. It's just that they're either Seehund midget submarines (midget-midget submarines?) proper oceanic subs, or Surcouf style cruiser submarines, in keeping with the light-medium-heavy unit class pattern. Does that all sound about right?

>>16062793

Sure, but voxel-GI is already kinda standard in the big engines. Even Godot has it. Scene voxels are another thing. Automontage's demos show a small square of land at most. At high res, you'd run out of memory pretty quickly, so you can't have high detail and a large map.

>>16062793

It's basically just going to be platform-specific extensions for these new features, kind of like how the oldest shader languages are all platform specific, like the ATI_fragment_shader OpenGL extension….

Right now for that new ray tracing, they added a bunch of new shader stages so that you can write a shader that generates rays, a shader that runs when a ray hits something or misses something or intersects with something, i'm not sure how it all works, but , it's interesting, I wont buy the new Nvidia card but I hope they emulate it on older cards, not because it would run fast enough for games but just so that I could access it to program for it, without buying the new cards.

Raytracing right now is just shown as new fancy lighting effects, denoising filters, etc… that can all be dropped in and integrated into existing technology easily because it doesn't disrupt how any other part of the system works. If voxels can do something similar, that would be great.

But yeah, hopefully AMD can respond too it.

>>16062830

>At high res, you'd run out of memory pretty quickly, so you can't have high detail and a large map.

Vid related. And procedural generation.

>>16061964

>Well, how many projects do we have this time?

No idea, but last time I was surprised by how many there were. Took a good while to play & write a short review for each of them.

>>16057705

>muh patents

Just ignore it lol.

>>16060883

If you actually read the image, you will see that it's different from what you described.

You describe Structured as ANYTHING that ever uses if/then/else and while/for

Wikipedia describes it as extensively using mostly those instead of mostly using methods and objects.

The google search also states that OOP is considered a TYPE of structured programming, meaning that although you can't OOP without structured, you can structured without OOP, meaning that OOP is optional

>>16060880

You are just vaguely complaining that the tools everyone else uses are bad without ever going into details of why and how and neither telling what you use (the supposedly not shit language) and why.

You sound like an autistic kid complaining because the neighbor is playing with a toy you don't like

>>16063867

I am using a vpn, so my ID keeps changing. I was hoping it wouldn't, but it ended up changing

>>16062505

In unity, I am using XML to declare items which can link to scriptable object instances, so it's trivial to attach arbitrary behaviors to items

I can declare a "tag processor" anywhere in my code and use reflection to find them, item properties read from processors are saved in a id->properties dictionary, so it's trivial to add more processors and more properties on demand

Item: https://pastebin.com/8w5THrAC

Example of a processor: https://pastebin.com/J7QhGR8c

>>16063927

So you can change behavior on a whim, but the behaviors still exist in your codebase and need to be compiled ahead of time?

>>16063963

yes, you can also easily reuse them, for instance, the 3 seeds use the same script, but the processor can clone the script instance and use a different parameter seed for each of them.

It's better than inserting code in a xls/xml file

>>16063867

>If you actually read the image, you will see that it's different from what you described.

No, if you read it you will see it is *EXACTLY* what I described.

>Wikipedia describes it as extensively using mostly those instead of mostly using methods and objects.

No it does not. It doesn't mention methods or objects at all. It predates OOP, it was introduced in the 50s. Structured programming is using ifs and whiles instead of goto, exactly like wikipedia says. You have no clue what you are talking about. The opposite of structured programming is not OOP, it is unstructured programming: https://en.wikipedia.org/wiki/Unstructured_programming which is using goto instead of high level control constructs. You could learn so much if you quit being an arrogant twat, stopped pretending you already know everything because you read the first 2 pages of "PHP for braindead niggers" and asked questions when you don't understand something instead of trying to argue out of stubborn ignorance.

>The google search also states that OOP is considered a TYPE of structured programming, meaning that although you can't OOP without structured, you can structured without OOP, meaning that OOP is optional

That's what I told you dumbfuck. OOP and structured programming are entirely orthogonal. What you said is nonsensical, it is like saying "just don't use OOP and use dynamic types instead". OOP and procedural are subsets of imperative programming. Structured vs unstructured is just an old way of saying high level vs low level. In unstructured programming you have no loops or if statements, you have to make those yourself out of goto. It has nothing to do with imperative, procedural, object oriented, functional, declarative, etc.

>You sound like an autistic kid complaining because the neighbor is playing with a toy you don't like

All you did was ignore the useful information provided for you and repeat idiotic nonsense. If you want to learn, ask questions. Don't spout stupid crap that makes no sense and then cry when you get corrected.

>>16064298

Ok, let's say you are absolutely correct about everything you stated, you might as well be because I never formally studied the name of the things I use.

Everything you did was shitpost about how OOP is bad without ever giving a alternative.

>If you want to learn, ask questions

Then I will ask a question, what is the alternative to OOP you are using right now to make your game?

>>16064652

>Everything you did was shitpost about how OOP is bad without ever giving a alternative.

I did nothing of the sort. You made a completely nonsensical post full of fundamental misunderstandings, and I corrected it. I am not the guy you were arguing with before that.

>what is the alternative to OOP you are using right now to make your game?

https://en.wikipedia.org/wiki/Functional_programming

>>16062109

Where do you get the sound effects? The tank treads sound really good.

>>16064807

I got them from SoundDogs.com

What they are specifically are two different recordings of a tank from different sources. One had the engine rumble drowning out the other noises, and the other was recorded from a distance so the squeaking and clatter of the tracks was more audible. I combined both in Audacity, and then cut and looped them into three separate sound files that last about a 3 seconds each, but loop seamlessly with eachother.

>>16064803

vols deleted your shitty thread before I could reply to you.

>The point was to be directionless, everything going towards a big puzzle/reveal.

I haven't played your game but, in case you haven't realized, those are mutually exclusive statements.

for something to be truly directionless, it must end in roughly the same place where it started– that is making you say what. a basic human trait is that we often try to make sense of things even when we lack all the information required to make an accurate prediction, the moment one sees something wacky and weird it's easy to arrive to two conclusions, a) this has no point at all, or b) there's actually a hidden meaning to all of this. unfortunately for you b) has been overdone in both video games and other mediums to the point that anyone can see this coming, and the thing is that you don't need to predict what's going to happen, just that something will happen.

but what I think you wanted to do was to pretend to be directionless in order to subvert the players expectations. have you ever played The Witness? it's also a puzzle game, and while it doesn't exactly subvert the players expectations(except for the ending, maybe) it does have plenty of gotcha moments, it does this by literally hiding stuff on plain sight.

managing players expectations is important, not just giving them direction(like the other anon said) because non-retarded humans often think ahead.

>>16059641

Honestly I just jumped on it because it's the cool new graphics meme that isn't OpenGL. I haven't had time to work on it in the past few days, but eventually I would like to create a basic framework for my applications/games which most likely won't do anything interesting beyond antialiasing+mipmapping (you never know though). The framework entails a GUI system and thus rendering text.

My main problems with it:

>Verbose API (I know this is the point, but it still hurts my fingers and my eyes, even after switching to Vulkan-Hpp)

>Terms that you think you know the meaning of ("Framebuffer", "Image") don't mean what you think

>No clue what all the state structs actually do and what are you allowed to do with them (What are good uses for secondary command buffers? Should I put all the non-changing stuff in them and just re-record a primary command buffer for each frame? How do multiple subpasses work with command buffers and pipelines?)

>New terms like "attachment" as opposed to "render target". What is the distinction between an "ImageLayout" and a "Format"?

>Creating handles to hold handles to the handles to the resources with some extra description for each layer for no apparent reason (Why do I need an ImageView that points to an Image to create a Framebuffer? Why do I need an AttachmentReference AND an AttachmentDescription to create a RenderPass?)

>Can't figure out how you would do an effective abstraction; seems like the most effective strategy is to just make a template program and modify it for different use cases since there's a lot of boilerplate that won't change for most use cases

>>16062910

What's the cache performance on these things?

>>16066479

Your explanation does clarify somewhat. And yeah, I think allocations for memory resources are the only things that stand out as things that you could easily abstract.

>>16066635

That depends. Can you have faction specific building placement graphics? If so, save that digital green one for the Huxley boys, use >>16066691 's suggestion of rough ink sketching for Golding, and make a blueprint paper looking tile graphic. Blue paper, grid lines, indistinct white schematics of piping, Regelbau bunker top views, or analog mechanisms.

Did you miss the last earlier response? I feel I backseat-design rambled a bit too much, so I'm not sure.

>>16066744

Thinking more on it, desaturated and maybe a little bit scuffed or with extra napkin sketch notes on top of the blueprint grid and bunker schematics? But not nearly as messy as the Golding/Mad Max doodle tile graphics.

keep up the hard work AGDG, may all your endeavors be successful

>>16066305

Point taken. This game isn't for you. What kind of games do you like?

>>16064803

>It's another abstract, directionless walking adventure game

oohh boy.

>>16064760

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

So you are using all of those languages at once?

I am interested in knowing what language an expert on the subject like yourself is using, it must be such a good alternative to C# that I will swap the second I see it and all of it's wonderful features

>I did nothing of the sort (…) I am not the guy you were arguing with before that.

from >>16060570

<OOP is just a brain damaged subset of imperative programming

<It isn't grouping them together, it is coupling them together. This forces horrible shit like inheritance and prototypes. This is why you have ugly and fragile "design patterns" like this: https://en.wikipedia.org/wiki/Composition_over_inheritance to try to work around the language intentionally crippling you

Sounds like shit talking OOP to me, plenty of shit talking about OOP here >>16060880 too where it's clearly compared to a big smelly turd

>>16065241

UE4 for medium/big teams or multiplayer games, Unity for Small teams making single player games

>>16067363

>So you are using all of those languages at once?

You asked what alternative to OOP, not what language. I use haskell obviously.

>it must be such a good alternative to C# that I will swap the second I see it and all of it's wonderful features

No you won't, you will continue to throw a hissy fit because you don't want to improve you want to justify wallowing in ignorance.

>Sounds like shit talking OOP to me

Which is not shitposting without giving alternatives, you know, like you claimed. I gave you alternatives repeatedly but you are too dumb to read what you reply to.

>>16067403

>I use haskell obviously.

Nice, which engine are you using? How is your game coming along? Should we expect a demo in 2/2?

>>16067444

Don't be passive aggressive. People shit talk OOP because it's so easy to implement that too many people get too comfortable for it without expanding into other paradigms like ECS. One should strive to learn other patterns in order to know what best fits a problem. Settling for OOP dogma will end up in a horror show of unmanageable hierarchies of inheritance with undecipherable functionality.

>>16067361

Cmon man

>>16067483

>Don't be passive aggressive

Nice seeing you dodge the most important questions and instantly going defensive

>People shit talk OOP because it's so easy to implement that too many people get too comfortable for it without expanding into other paradigms like ECS

The existence of pajeets in no way shape or form invalidate the benefits of OOP, even if "It's super easy to use" happens to be the only one. Also, ECS and OOP can be used at the same time on demand, some stuff is insanely hard to do in ECS without forfeiting it's insane performance.

>Settling for OOP dogma will end up in a horror show of unmanageable hierarchies of inheritance with undecipherable functionality

Your argument relies on the programmer being incompetent. It sounds like a personal problem.

I might not know all the useless formalities, but I know that using OOP and a engine like Unity or UE4 is the only hope of ever getting a competitive game in the market.

>>16067583

>going defensive

I'm not the one you've been talking with.

>OOP is the only hope of getting a competitive game on the market.

No.

We don't actually disagree much, OOP has it's uses. I was merely explaining why you'll find so much antagonism about it. The worst programmers that made the Pajeet stereotype flourish are the kind that only know OOP and use it for everything.

>>16067621

>I'm not the one you've been talking with

fucking hell, I need to read the fucking IDs more often

>>16067496

I don't exactly want to sound like a complete fucker, but you did ask for first impressions. That's exactly what I felt when I watched the gameplay. I do like the colorful illustration and character designs and would like to see something come of it, but it's just a re-skinned Yume Nikki clone in my eyes from initial viewing.

It's a collage of random stuff smashed together that you just wander around, interact with and look at. I know some people like this sort of thing, but calling this a game at this point in time is a stretch. You can have abstract visuals and concepts in a game, but when there is no perceivable sense, progression or challenge to it, I personally don't see it as a game anymore; just a interactive visual application.

>>16067444

>which engine are you using?

No.

>How is your game coming along?

Meh, I'd guess I am about 1/5th of the way done. So in reality I am probably about 1/20th done.

>Should we expect a demo in 2/2?

In what year?

>>16067583

>Your argument relies on the programmer being incompetent

No, actually his argument relies on the programmer being human. OOP is defective by design. It adds complexity rather than abstracting complexity. This is why large OOP programs are always broken shit, without exception. There is no human being in existence than can handle the imposed complexity of the broken design of OOP. Programming languages are supposed to help you generate a correct program, not make it harder.

>tfw been too busy working at a game company porting a VR game from android to PC to work on my own projects

>tfw I "have a demo" for that but it's not a project I own so I probably shouldn't post it on cantonese slideshowsites

Wageslaving was a mistake. Work itself's been nice though

>>16067583

>but I know that using OOP and a engine like Unity or UE4 is the only hope of ever getting a competitive game in the market

Not only are you wrong about having to use an engine (Stardew Valley, Bastion, Axiom Verge just to name a few off the top of my head), Unity is currently working on implementing DOP pipelines instead of OOP ones under the leadership of Mike Acton and Andreas Fredriksson.

>>16069535

Those 3 examples are still using XNA, which isn't editor enabled, but is still way faster than writing from scratch with nothing but windowing libraries and an GL layer.

>>16069547

Yes, and? What's that got to do with the fact that the statement "you need to use an engine like Unity or UE4 to be competitive" is wrong?

>hurr you didn't design the CPU for your game from the ground up yourself so you might as well use Game Maker Language instead

Finally going to start working on setting everything up for a game I want to make. What's a good programming language to learn? Typically people use Python but I'm not sure.

>>16069555

The only actual difference between Unity and XNA is the Editor. If you have a prebuilt tool, some experience with format interpretation to use an existing tool for a different workflow, or are comfortable designing levels without tooling, then there's virtually no difference between XNA and Unity/UE4. Don't spazz on me if you don't understand the difference faggot.

>>16069859

Depends on the engine you are using.

>>16070268

you are missing the point, the only real reason to use a game engine is to save yourself a lot of time and effort (or to use engine features that are beyond your skill to develop yourself)

there are also reasons why you would avoid using a game engine such as:

- you are making an extremely edgy game and are worried about being shut down by the corporate overlords that own the main commercial engines

- your game has experimental features that are difficult to implement into the existing framework of an engine

- you already have an established workflow and set of skills that will let you make a competent game with XNA/your custom engine/whatever and you don't want to waste time switching to an engine for little gain over your existing work

- you don't want to pay royalties/subscription fees

basically the quality of your final product depends a lot more on the skills and workflow of you and your team than it does on what engine you choose

>>16070730

I could have sworn I saw your game before on some obscure chsnnel that showcases unique games.

>>16070286

Probably going to go with Unity or Unreal

>>16070808

i doubt it, i haven't done any marketing yet and my youtube showing progress only has a few dozen subs

i did use an asset pack that's pretty common though for the playermodels and many environment props (gonna change most of it later its just for prototyping)

>>16070839

Mind linking your channel?

>>16070811

If you're going with Unity you'll learn C#. This is something you can do while you make the game.

If you're going with Unreal you'll either have to use Blueprints or learn C++. If you go C++ you'll need to first learn C++ outside of Unreal, then go into Unreal and learn the engine from a C++ perspective. Blueprints is only good if your game is very simple. Anything too complex and C++ becomes faster, easier to read, and easier to follow than Blueprints.

In both cases look through the sample games to see how certain things are accomplished/handled. Spawning objects, physics, ray-casting, animation, input, text, and so on.

>>16070894

https://www.youtube.com/watch?v=IB43RakhGJI&t=1s

here's the latest vid

>>16070922

Thanks for the insight and yeah I'll look into them a bit more in depth to see which is better for the thing I intend on making.

>>16070969

If it's a simple game, consider Godot. It's free and it uses a scripting language that's arguably easier than C#.

>>16070925

Looks competent. Keep it up dude, there's always demand for passionate work.

>>16071007

thanks man, its actually my first game (only done modding before this)

so even if it sells like shit i think the knowledge and experience gained will be valuable at least

>>16065241

I prefer Unity personally. The engine gets shit on because a lot of garbage gets made in it, but you don't blame the canvas when a guy takes a shit on it and calls it art. When I was first starting to play around with gamedev, I found it much easier to find resources to help learn in Unity than in UDK and it seemed like a lot more work to get it set up for something that isn't a FPS whereas Unity is pretty much a blank slate. Admittedly I haven't touched UDK in years and a lot of my issues with it are likely stemming from my own lack of experience at the time rather than a problem with Unreal itself, but the whole thing did sour me on it.

>>16071005

I will second this. I'm currently using Godot and enjoying it. 3.0 is very capable and I'm exited for the complete release of 3.1, especially the addition of CSG. It will seriously improve 3D level prototyping from where it is now.

>>16071508

no

>>16071508

I spend all month playing games instead of making them

>>16071508

I need to fix a bug and get it running then I'll have a demo to post.

>>16071508

just make the thread ya lazy faggots I'm all fired up

Is anyone else working on the thread? If not I'm going to do it.

DEMODAY THREAD

Come post your demos

>>16071775

>>16071775

>>16071775

>>16071775

>>16071775

>>16068643

>No, actually his argument relies on the programmer being human. OOP is defective by design. It adds complexity rather than abstracting complexity. This is why large OOP programs are always broken shit, without exception. There is no human being in existence than can handle the imposed complexity of the broken design of OOP. Programming languages are supposed to help you generate a correct program, not make it harder

What a incredible large amount of OPINION

Sounds like a personal problem

The biggest issue with anti-OOP shitposter is that if they manage to fool someone, it will inevitably create a nodev hipster that will never accomplish his goals or have to settle with downscoping it to hell and back until his game is a console only text adventure.

The fact that shit devs manage to shit out trash is a blessing not a curse, it means that OOP is so fucking good that even retards can manage to finish something with it

>>16072010

and my ip changed again thanks to the vpn

>>16072013

Your "personal problem" stab already identifies you, anon. :^)

>>16069535

>engine like Unity or UE4

>like

This word was a very important part of the sentence

Competing against those standing on the shoulder of giants without doing the same is crazy. Sure it's possible, but it's an unneeded added challenge.

Also, stardew valley was successful because of the gigantic hole in the market for a farming sim

>>16070755

>is to save yourself a lot of time and effort

Some people can't afford the time and effort of not using a engine, not everyone is born with a silver spoon in the mouth

Depending on your circumstances, using a engine is a requirement for success. I will accept that some people can go the long way instead, although I am hard pressed to agree that they could get the same "money/effort" ratio as using a engine would give. I am talking about the average, not stating that no example contradicting me exists

>>16070969

>>16070811

If you are inexperienced and making a single player game, go Unity

If you got a team or are making a multiplayer game, use Unreal

Unity forces you to use their multiplayer service without exceptions, so I'd argue that being able to choose your own multiplayer solution gives UE4 an advantage in this regard. In Unity, you can't even host your own servers

>>16071005

I'd argue against Godot for now, but it's certainly something to look at in 3-5 years. I will possibly not use Unity for my next project depending on how much they keep fucking up their users

>>16071508

Last 3 months was about inventory, interface and reading XML, caring about modding made it take even longer than i'd like. Too much engine code and almost nothing gameplay wise, so it's almost indistinguishable from last demo and most of the new stuff is half finished, so we are going to skip this demoday

>>16072010

Facts don't become opinions just because you don't understand them. Those retards would finish something faster and it would be better if they weren't stuck trying to manage the increased complexity of OOP. You are literally doing this:

>OMG dumb people can't handle programming

>I know, lets make it harder, that is helpful!

But you are too stupid to realize it.

Anyone knows any good gunfire sounds for rifle, machine guns and pistols? Doing some WW2 game and I would like to borrow some of the sounds before i find a proper replacement.

>>16072364

>OOP too hard

brainlet

>>16072364

OOP doesn't inherently make projects more complex. OOP only makes projects more complex when you get retards who force it everywhere they can on the project with no thought onto whether or not it's needed.

>>16073005

We've already been over this, read before you reply.

>>16073085

>Read more code style sperging which has plagued this thread for 3 years now

Nah. The filter option works much better.

>>16073107

>programming paradigms are code style

You're a retard.

>what other people said is a waste of time and ruins the thread so I will ignore it

>me repeating what other people already said is useful and improves the thread though

Yep, definitely a retard.

>>16073085

my ip probably didn't change this time, sorry

might have, hopefully not

>>16070268

>The only actual difference between Unity and XNA is the Editor.

And this is the dumbest thing I've seen in this thread so far. Back to your loo.

>>16073169

How am I wrong?

(Hint, I'm not.)

>>16073176

What do you think the difference between a framework and an engine is? It's not whether they have a UI or not. Unity has had? multiplayer built in, an entire animation system, and a shitload of mobile features like ads to help you output your endless supply of shovelware.

>>16073604

Not a bad idea in of itself, but you will most likely not find any information online to help get that going since it's a very outdated technique

Old games like Megaman legends and some other ps1 capcom games used to do that, could work and could end up looking good at the same time if done right, especially on the joints

Might or might not be harder to animate, depending on the engine

>>16073604

>and the rest of the game is a generic skyrim ripoff

0/10 apply yourself. At least try and rip off some half decent combat system.

>>16073252

The mouse graphics look very nice. I'm not sure how 'busy' they look ingame, in terms of potential distraction, but they are highly communicative. I actually like the grungy feel of them as it is now and the contrast between saturated and desaturated parts. It's fitting for example that the Move icon is more mutely colored than the one for Attack, which is a reddish-orange-hot metal color. I'm not sure what the frame by frame artifacting on Select is though, as the others are very clean. Is it just that, or an attempt at radar noise that the exporter messed up? Though there's also one small artifact on the upper right corner piece of the Enter gif. This tiny floating diamond shape.

>I do like the asymmetrical balance of games like Generals. GLA had no airforce in that, but compensated with tunneling and stealth tech.

Thing is, the GLA not having air was maybe not quite as big a handicap as having no navy at all would be if you end up with water-heavy maps. Generals had no non-campaign naval play without being modded; everything is overland, and air units go over the top of that. And while tunneling and AAA can bypass lower direct assault ability & no air cavalry, what could allow Max to surmount being trivial to blockade? Speedy transports? Subs? Expanding so quick after landfall (if you can micro it) that they just ignore the naval game after making a beachhead?

To use another C&C faction for example, Yuri has all of two naval units. The generic amphibious transport, and the expensive all-purpose Boomer submarine. That's it, that's that. For Max/Golding, you can maybe have a similar dynamic and have two basic units amalgamating the previous concepts, though as initially cheap and low-quality instead of expensive do-alls; the PT boat and the tanker ship. Both are cheap, respectively HP and speed poor, modularly upgradable and/or use scrap to get GLA style powerups, and have the main initial purpose of getting a certain number of units where they need to go. Glorified taxicabs with popguns. Stick enough battlefield scrap on or infuse enough cash into them (either fills the upgrade capacity bar) and you end up with ersatz hybrid versions of the other factional naval units.

>>16074533

The PT boat starts bloating up, lengthening and changing hull style to take on crudely imitative aspects of riverine gunboats, E-boots, proper destroyers and WWI submarines, going from a glorified squad carrier to becoming a glass cannon speed demon with a slow sneeki option. With a 'late' upgrade to give it Cybran Salem class land capacity with a set of tank treads, allowing it to join land battles directly; when it has both this upgrade and the submersible one, the PT gets the Atoragon drillbit. You can rig them to blow on death as fireships.

The tanker piles on industrial equipment & reactivates the holds to become a limited mobile production base for basic units and a resource refinery, guns guns guns to compete on the battleline or at least deter interdictors, or crude air servicing equipment to vomit out shitty gyrocopters that don't do well in air to air and aren't nearly so fast as proper air units but have the advantage of being comparatively cheap as dirt hovering wherever they like, making great ghetto helicopter gunships if the enemy neglects their air defenses. The 'late' upgrade could be a Katyusha style rocket array firing V-weapon scale suicide piloted boomtubes, expensive and incapable of hitting the broad side of a barn in spite of the speed freak guidance system, but being a potential base flattener on its own through sheer output compared to more precise and slower enemy firing equivalents. Like a Scathis on the water.

These options make these two units more versatile than many in the mostly WYSIWYG Huxley fleet, and though the upgrade options are of lower quality/stat gain than similar Orwell upgrades, have the benefit of not directly competing with each other with no either-or options; you can pile absolutely all of them onto a unit if circumstances allow. It's probably a stupid idea under many circumstances to, say, make your double-sub mega PT boats into inevitably chain detonating suicide bombers by sticking the fireship upgrade onto units you've invested so much into (for the Max faction standard at least) but you can do it.

So you go from cheap do-nothing chaff ships you use for sneak attacks, expansion and intel gathering to cheapish to moderate do-all units that are deceptively effective if used right. And 'used right' means either sneaking around or unsubtly snowballing the shit out of your inflexible opponents, who have to worry about things like 'fleet composition,' and 'draft limits,' or 'naval construction regulations.' Carbarians don't give a shit about any of that and reap the benefits of completely unsafe uncontrolled overbuilding. Plus, you get the excuse to go hogshit wild modeling the many, many modular upgrade graphics, with no worries of possible unit redundancy like the Orwell tankettes. Because it's all one unit per two weight classes.

>>16073140

>>16073085

>crawling in my skin.mp4

>>16072056

>In Unity, you can't even host your own servers

>Unity forces you to use their multiplayer service without exceptions

They corrected their EULA and made it so you can use any multiplayer service (including that one service the multiplayer fiasco was about), and made EULAs binding per version of unity; so they can't retroactively fuck you over with an EULA change.

It's a change for the better imo, but that EULA fiasco was a real shitshow; so distrust is warranted.

>>16075180

If you have one big vertex buffer, this will be faster. The ideal is to have as few vertex buffers as possible, it might be more difficult to write it this way. I found a Vulkan-related discord server that had a lot of people talking on it, I am not sure if an IRC chat for talking about it exists. Using only one Vertex Buffer will probably be hard to write especially when you have to put multiple pixel formats, etc, into one buffer. At some point, aren't you doing the GPU's job of managing multiple vertex buffers in one memory space?

Also if you want to change what is in this vertex buffer beyond the setup, it will be rather slow, because you will have to change it's memory property, ideally you want everything to be in memory that is VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT but to change what is in the buffer, it has to be VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT which is slower.

However needing to change what is in your VBO is really something that, I think people say that you would need to do, but isn't something that you really need to be doing. Just design everything correctly and it will be great. It's like people who allocate memory in their game loop. It's not necessary but lots of people do it when they don't need too.

>>16075180

>vkCmdDrawIndirect to draw all my geometry

I'm curious, how would one apply transform matrices per-object if vkCmdDrawIndirect is used? Currently for transform matrices I'm using a uniform buffer filled with an array of matrices and sending an index in between draw calls using push constants. Is there some kind of invocation id with indirect draws or something else I'm unaware of?

>>16073604

Nice and easy for non-organic characters like robots. Though figuring out how to orient the raw individual parts can be a pain. I wanted to do something very similar with parented parts and robots, but I couldn't figure out the gameplay so I never got far. Good to see someone is making robot game for me.

>>16075802

Instead of messing with the rig and animator, leave the bones intact and create a new object and contain the detached part in it.

>>16075989

>Though figuring out how to orient the raw individual parts can be a pain

the actual parts are prefabs, i just equip them once, move them around until they look right and then copy the new values onto the prefab's transform

>leave the bones intact and create a new object

yeah i was just hoping i could use the existing bone, would have been the cleaner solution

>>16075809

Quite frankly I have no idea what I'm doing, but I found an example from Sascha Willems:

https://github.com/SaschaWillems/Vulkan/blob/master/examples/indirectdraw/indirectdraw.cpp

>Per-instance transforms go into vertex attributes

>Separate vertex buffer with different binding used to hold per-instance data vs per-vertex data

>>16074813

oh, so they are smart enough to backtrack if they step on shit?

That ease a lot of my worries

>>16075989

>>16076003

>leave the bones intact and create a new object and contain the detached part in it

By far seems the easier solution, but you could also create entire robots, separate parts of them, and then change the skeleton they are following to be the same as the center piece.

It's literally the same tech for equipping clothing to characters on the fly, but instead of shirts on a humanoid, you are attaching arms and legs to a flying chest piece. Might be easier to model the robots and test their animations, even if the implementation is a bit harder

>>16076524

tbh i already have prefabs of the base bones for the sake of reattaching them later, guess i'll just instantiate those and re-equip the items to them when detaching

Are there any good tutorials for getting started with UE4? Some of the official ones look a bit outdated.

Shmup anon here,

Work finally wound down, and I had some cold/flu last week.

Slowly getting back into my current task: explosions.

I had a formula maybe working from a point and propagating to other, but when it was time to implement it…

It can't possibly work in the fragment shader since I can only read from any location, but write only on the current fragment. And from what I have is from the current point propagating to others.

I'll have to rework this and have a lookup from a certain range around the current fragment.

And if this doesn't work… Back to the drawing board.

I'm a fool, I didn't see this thread was still up and posted my progress in the demo day thread. Here it is: >>16078103

>>16075492

This is sexy, let me know if you need some voice work for anything that sounds like it comes from a shitty headset (like a pilot or mech unit). This is ANUBITEK dev.

Demodays have officially reached Recap Monday status where whodevs you've never seen before (except maybe previous demo days) go to shill their games.

>>16049287

gib demo now

How long should I get into the online lua reference before jumping into Love and shitting out some prototype game? I'm reading a lot of really neat tips and tricks but at the same time I can very rarely relate any of the things I read to something I might use later.

>>16080984

Start your game as soon as possible. Getting experience is the best way to learn.

stupid question. but

love zbrush, having a blast in it, but $500 is steep at this moment in time (just had some big vet bills) and my license ends end of february

reading up on it, zbrush cracks seem scarce or super sketchy

do i just bite the bullet or is there a actual cracked version out there? hate to be spoonfed but i've been looking for like five weeks.

>>16082145

cgpeers, registration opens on 15th of every month