/v/ - Video Games

Vidya Games

[Index] [Catalog] [Archive] [Bottom] [Refresh]
Mode: Reply
Options
Subject
Message

Max message length: 8000

Files

Max file size: 24.00 MB

Max files: 5

Supported file types: GIF, JPG, PNG, WebM, OGG, and more

E-mail
Password

(used to delete files and postings)

Misc

Remember to follow the rules

The backup domain is located at 8chan.se. .cc is a third fallback. TOR access can be found here, or you can access the TOR portal from the clearnet at Redchannit.
Upgrade success! 8chan.moe is now running on the LynxChan 2.4 engine with our custom frontend. Please report bugs on /site/.

8chan.moe is a hobby project with no affiliation whatsoever to any other "8chan" site, past or present.

/agdg/ - Amateur Game Developemnt General Anonymous 05/17/2020 (Sun) 17:06:06 Id:2b24c2 No. 8707
Resources: >#8/agdg/ via irc.rizon.net >https://matrix.to/#/+agdg:matrix.org via matrix programs >Dev resources: http://8agdg.wikidot.com/resources Links: >Wiki: http://8agdg.wikidot.com/
(41.69 KB 421x421 not smug.jpg)
I genuinely cannot figure out how C works I'm learning it before C++ so I don't get overwhelmed by the bloat Am I retarded or something? I've been attempting this for a good 8 months now but it's just not clicking. It's the same shit with math as well.
>>8735 How bad are you at math, anon?
>>8738 Well, everything. I don't understand it. I can repeat the instructions in a tutorial and I can listen to what the guy says, but none of it makes sense. I'm not retaining any information. >>8745 I'm literally worse than a 6-year-old. I can't multiple 7 by 7 without using a calculator. Division is absolutely impossible for me. It's not like I haven't tried to get better, I have, but it goes the same way as code. It's hard to explain, but I can't automatically do math like a lot of people can. I have to manually go "7 plus 7 is 14, plus 7 is 21, plus 7 is 28.." and so on. It's hard to even count to 100 without going accidentally jumping from 46 to 74.
>>8749 Interesting. Tell me anon, are you clumsy at all? Do you have trouble telling left from right, do you find yourself unable to tell if you're in somebody's way or not, do you have a poor sense of your surroundings, etc? This might sound irrelevant, but it's quite important.
>>8755 Yes to all of those. We're starting to derail the thread a bit so can we take this back to code? Do you know me IRL or something? What you listed is so bizarrely specific yet so strangely accurate, I'm trying to figure out if something I posted might have given away who I am.
>>8707 Every project I start solo ends up dead by way of feature creep and eventual burnout due to said feature creep. Any tips to avoid this fucking blight. >>8772 >spoiler Sounds more like you may have a condition or something as Dr. Anon is clearly diagnosing you based on your issues.
>>8772 Oh dear. Anon, I'm really quite sorry to tell you this, but you perfectly fit the description of an actual neurological disorder. https://en.wikipedia.org/wiki/Dyscalculia
>>8707 reminder that Phantom Path is finally out https://store.steampowered.com/app/1180970/Phantom_Path/
>>8830 So l should have a clearly defined end goal but also have a number of smaller goals to ease towards making it? Now that I think about it, I feel like setting ambiguous/vague goals before is what was likely giving me room to feature creep myself to death. Thanks for the advice anon.
(51.09 KB 800x800 untitled.gif)
Hello my niggas. Long time no see. :) You know already who I am. :)
(13.27 KB 217x320 shiajust.jpg)
>>8797 >>8805 just fuck my brain up fam So where do I go from here, what do I do with this information? Can I ever learn to code or am I cursed to forever be a nodev?
>>8897 As one of the other anons mentioned, programming is tied to logic and not only math. However, both sides may be related in your case. Not really a rocketbrain scientist. If you are on the younger side, you may be able to get help from a doctor. Apparently there are apps out there to help you get a handle on it. Even if you end up not coding, there's plenty of other stuff out there you may be able to do, like drawing or music.
>>8897 learn to draw. or learn basics of music. write good stories.
>>8880 I require more belts on this character.
>>8847 didn't know speedbot dev finished his new game. I'll check it out.
(361.96 KB 823x1012 Literally10MinutesOfCode.png)
>>8735 >falling for /tech/ memes Don't listen to turbo programming autists anon. Listen to turbo gamedev autists. If you want to get started with game shit you can just download godot and follow a tutorial in youtube. You'll have a shitty mario clone in literally 2 hours.
>>8897 Just like going to a gym, you don't start with the 500 pounds weight. Try to make a finished pong, like music, a level system, some dificulties, nice particles, a background, some shaders, basic 2D/3D graphics. It's clearly doable for a noob. If you really think pong is too easy, try to remake a game I invented and use as my hello world to learn any new engine. https://www.youtube.com/watch?v=NY3y4YI_ZRE >>8907 She's from an SNK fighting game.
would it be a good idea to code something like a calculator or a tic-tac-toe game with C# using visual studio? because i've been doing just that in short bursts for the last 3 or 4 days
>>9111 to start practicing*
>>9111 Sure, that's fine to learn the basics of coding. but why not use godot?
>>9128 i dunno. whats the difference? i haven't coded in like 10 years and forgot how to almost entirely. besides basic concepts.
>>9111 I suggest writing CLI applications before you start doing any GUI stuff, just to get the basics of the language down. Then either start doing GUI stuff or you can jump straight into game programming.
>>9132 It's much much easier, It's like using blender vs making a 3D model on notepad.
Trying armor paint on ubuntu.
>>9128 It helps me exercise self-control, if that makes any sense. With Godot, Unity, Unreal, Game Maker or whatever, I can write performance issues and things I don't like off as "engine limitations". At times, it's possible to work around it. At times, it's not, but I remember being frustrated with issues of various engines and saying "fuck it, that's just an engine limitation, I can't do shit about this". It's a terrible attitude to take, considering the player doesn't know that. They don't know whether it's an engine limitation or incompetence on my part, all they care about is whether the game works fine or not. Essentially, with my own engine, I can never write anything off as "good enough", it has to be perfect and if it's not, that's my fault and my fault alone. It's a good kick in the ass to get things working well. On a more important but simpler point, I just enjoy the process of making an engine, probably more than I enjoy making an actual game. >>8897 I started with a "make game in OpenGL/C++" youtube tutorial that used GLEW and a bunch of other libraries of that sort. Mostly copied shit and googled errors until I got something working, and through months of practice, I got a hang of things. Of course, that meant that the whole thing needed an eventual rewrite but that's not a terrible thing.
>>9180 Do u want to make games or write an engine? Making a game using a fucking engine like unity and godot is still way shit ton hard, don't make it more retardely hard.
>>9191 >Do u want to make games or write an engine? Loaded question. While I thought I wanted to make a game, making my own engine has been far more enjoyable than using someone else's to make a game has in the past. I want to do both and there are many things I can do with my own engine that I cannot do with Godot or Unreal. >Making a game using a fucking engine like unity and godot is still way shit ton hard, don't make it more retardely hard. One is hard, the other is harder but gets easier. By the time you write an engine, you have full knowledge of everything, from programming language to the interworkings of your engine, so doing anything you need is much easier than it would be on a premade engine. I really don't see why you so aggressively try to push people to use other people's engines, Godot especially. Comes off as shilling and I'd hope it's not that. The whole point is to make a game, regardless of how you do it. Wanna "waste" ten years and make your own engine? Go ahead. Wanna use RPG maker? Equally as valid. You could make this argument for anything. Why use your own 3D models when there's many you can get for free? Why make your own music when you can get royalty-free shit that's probably better?
>>9202 even godot was written by a spic ginger that had 20 years of coding experience, and took him near 4 years of working full time on the engine, plus hundreds of people helping him to make godot. Imagine now competing with unity or unreal.
>>9209 >Imagine now competing with unity or unreal. It's very easy, actually. Those are written to be good at everything, with Unreal perhaps having a slight focus on FPSes. Due to the ridiculous bloat of those engines, a custom engine can get far better performance, something I have tested quite many times and it's very obvious. Then there's smaller things, such as bugs and inconveniences associated with Windows' default behavior. Those are things I've discussed at length with Sigmadev (hope he finds his way here) and we've found many workarounds for these issues - workarounds that are impossible on Godot or Unreal. That's on top of Godot requiring me to learn ANOTHER fucking programming language, because I don't know enough of them yet, and it being some mess styled after Python which I think is worse than hot aids. And Godot breaking old code with new updates. And Godot's devs cucking out with that Master/Slave - Master/Puppet thing (if I recall that correctly). And many other issues. You know what sparked it all, though? The very simple fact that I'm making a game styled after late 90s platformers. In my mind, if the game is styled after late 90s platformers, it should be able to run on late 90s hardware and older OSes because fuck you, why shouldn't it? There's no answer, only excuses. If I can make it work, I should, and I am. Can't do that with Godot, Unreal or Unity - not because of performance reasons but because fuck you lol, our engines are shit. I'm not "competing" with Unreal or Unity or Godot in a traditional way. I'm making a specialized engine for a specialized game, because I can make it run on more and older OSes, I can address issues with those OSes that other engines don't care to address, and I can make it run better. Why would I not do that? Because "it takes too long?" Fuck you, I'm not in a rush, am I? I have to write >60 songs without much knowledge of music. I have to make thousands of models without much knowledge of 3d modeling. I could go to incompetech.com and free3d.com right now and save myself YEARS of effort. But I could also buy a Unity pack for 15 bucks and make an asset flip in a few days. Taking less time to make is not always a good thing. You do you, but stop shilling your methodology (and your product) to others, especially if they have numerous legitimate reasons for doing things the way they are.
>>9222 There's already engines even for the NES and later consoles.
>>9222 I think people are merely saying you don't need to grow, pick and grind the wheat yourself to make bread. Some people are quite time-poor and have to make hard choices about this stuff even if they do find it interesting. Maybe you have good reasons to do that, but at the end of the day it's just bread, and it's most special to you and noone else. If you want to sell bread to make a living, like I think many at AGDG do, then you need to make some hard choices about scope to get a realistic roadmap to deliver. If you're just playing around, sure, whatever, have fun. Then there's a point about force multiplication. Sure, you can take the 300 hours to learn FL Studio or Famitracker or whatever, some rudimentary music theory and take the time to make something, but it may be cheaper for you, in a cosmic sense, to work for less hours at $dayjob and pay someone else to do this for you for less than it would have cost you. Shit like Birdemic doesn't become a better movie because it was done all in-house. Also, the alternative to mastering everything isn't going straight to Kevin McCloud and getting free tracks. There is a middle ground.
>>9234 >Maybe you have good reasons to do that, but at the end of the day it's just bread, and it's most special to you and noone else. Don't get me wrong, there's reasons to do whatever. Like I said, you can find perfectly valid reasons to use RPG Maker or whatever. That was my initial point - do whatever makes the most sense to you. But if someone's considering making their own engine, I highly doubt they care about time. Nobody who decides to make their own engine assumes that it's easy, unless they're those delusional retards who come in saying "hey guyse I'sa wanna make an MMO XDDD", but I doubt they even are aware it's possible to make your own engine. The other anon in particular has been aggressively pushing against making your own engine in any capacity, to any end, regardless of the circumstances. I have named dozens of reasons for why I'm making my own engine when I only needed one - it's fun, and to the other anon, it's unacceptable for some reason.
>>9240 If I might offer one closing thought: people tend to try and realize the most perfect complete version of their overall ambitious vision first, noone wants to start simple and work their way up. Particularly amateurs. This tends to lead to burnout as people bounce off technology with very little rewards to show for it as they learn the ropes. Knowledge, of course, is not tangible so the brain tends not to reward us chemically for it. I would propose this as a reason why people may be cautioning others to start with conventional and established tech.
>>9392 Looks neat for throwback stuff.
(8.24 KB 384x704 ClipboardImage.png)
I'm working on tile collisions for my game right now. I made my own Physics component system that right now just retains a bunch of positions and velocities, and doesn't see things like tiles, solids, etc., so the logic is in my main game object for the time being I am having trouble trying to resolve an entity entering a tile it shouldn't. For example, if a character moves X so fast that it hits a wall, it should revert to the previous positions. However, for the direction that wasn't embedded, it should still retain its velocity in that direction. In very specific diagonal cases, there is ambiguity for which axis it should zero out. Do I then poll adjacent tiles to see which edge case it is? Another difficulty is that since it is physics based, I need a way to cheaply indicate that each tile is solid, and not all entities would be stopped by the same things (eg a passwall monster).
>>9722 I'm no programmer but I know that collision checking between 2 pixels that do not collide all the time is retarded so put walls on their own layer and create a mirror sprite (collision silhouette) on said layer that is bound to the graphical sprite by some pixel and have the graphical sprite reset to the collision sprite's position in check if computation gets wonk. Those that do not interact with walls don't have a sprite on this layer or you simplify the collision and perhaps use a different collision sprite than graphical.
>>9794 Oh, I'm working with tiles or abstract units in this case, not pixels. I know you can offload stuff to the graphics card for performance gains, but for my simple case and limited size, the overhead of it would probably cost more than just writing it out by hand.
>>9722 You either have to store a surface normal for each tile (calculated ahead of time, when the wall or tile is placed) or you will have to check surrounding tiles to determine the surface normal at the collision. Storing on wall creation would make more sense and you could easily handle odd angle diagonals.
>tfw tempted to make a fetish RPG Maker game >but it would probably be a colossal failure and take literally years to make anything worthwhile, or at least 1-2 years even if I wanted to do something akin to a tech demo spinoff of what I had in mind it sucks honestly. Does anyone have any tips when it comes to game dev discipline, or how to keep the ball rolling once you start? One time I started learning RPG maker to give it a shot and it wasn't necessarily going badly but I just stopped and forgot about it.
>>9805 Well yes, fetishes themself are a waste of time and soul degenerating bullshit. Thus you will never program anything. Go on nofap and burn up the impurities and parasites that make up over half your weight and volume.
>>9798 Using Terraria as my base, that's 11.2m tiles on a medium world. Chunking helps, but storing a normal seems a lot more expensive than just checking one or two additional things on a per-collision basis. Even then, I could store a byte for a lookup table for the normals. >>9805 >Does anyone have any tips when it comes to game dev discipline Just like make game. If you're afraid of making getting your hands dirty and doing it the right way, you won't make anything and you'll stagnate. Just make shit for yourself to enjoy and then share it with others. Nobody cares what you make as long as it's fun. You will always have an audience.
>>9812 Maybe, but other people have made fun H-RPGs so I know its at least possible. Plus RPG Maker seems doable and doesn't involve much programming at all. You have a fair point about the soul-degenerating bullshit though. >>9814 Fair enough, it can't hurt to give it a try I suppose.
>>9805 Discipline is tough. I usually take it easy until I get in the mood to work on a particular aspect, such as music or coding. I'll then work on that as hard as I can, sometimes for a few months, until burnout hits as by then I'll have collected enough inspiration or ideas to work on another field. There's pretty much never a time when I don't want to work on some aspect of my game. Plus, it's RPG maker, it's not that hard once you get the hang of it. Starting will be hard but you just have to get in the swing of things.
>>9814 >Using Terraria as my base, that's 11.2m tiles on a medium world. I think in that case you just need to detect which side of the block you are striking, since you are using square blocks. No need to check adjacent tiles or do a pre-calc. There isn't ever really that much on-screen in terraria. You could use a pre-rolled system like Box2D.
>>9722 You are dealing with very simple and cheap primitives. If you want to support high accuracy collisions; use a sweep instead.
(553.80 KB 768x432 ClipboardImage.png)
Unrelated to anything, but I clicked on a random bookmark I've had for years, and hopefully someone finds it interesting or hlepful http://www.elopezr.com/the-rendering-of-middle-earth-shadow-of-mordor/ http://archive.is/HAYsU
>>9850 Thanks
>>8906 Got any good free sources?
(385.43 KB 2048x1302 music-theory-gay-image-1.jpg)
>>10623 here is an image for music theory. You don't need to know all of it but the basics is that certain notes go well together to set a certain type of mood you want. Higher pitched music sounds more lively and lower pitched is more dark and sadder. Your melody is going to be very simple with only 4 keys being used. The beat is even simpler since you can do it in 1-3 keys depending on how complex you want it to be. The lead is the most complex part and the biggest part. You can choose to have it in the same key as the melody or you can make it different if you want the lead to clash with the rest of the song.
>>11217 I think I'll stick to free use tracks.
(32.97 KB 571x896 sweepAlgorithm.png)
>>9722 Use sweep algorithm like what >>9846 said. It's perfect for collisions like this.
>>11217 Look, unless you're planning to do really fucky kinds of stuff with music like King Gizzard, you really don't need to know music theory to write it.
>>11436 Oh my shitty picture was worth saving? Anyways I'll do that and I'm not sure why I ran into issues on something I soft solved already. Im actually working on a simple octant based raycaster for lighting but which the collision system can use for sweeps
(1.53 MB 1171x862 ClipboardImage.png)
Planting grass, or rather, clovers
>>11452 This, you'd be surprised how simple some good songs can be. Just an example that I was thinking about today, Secret of the Forest from Chrono Trigger hops between two chords for the majority of the song and only gets at all complex towards the end of the loop. When you get into music theory breakdowns of songs, you'll notice that relatively-simple sounding sections of Mario songs will go through seven chord changes because they have to categorize every single note that falls even slightly out of scale. To the laymen like you or I, we'd just describe that as "it sounds nice there so I did it", whereas they'd label it as a F flat major seventh going to G sharp minor fifth or whatever else. When I write a song's melody, it mostly falls down to instinct - part humming it, part hearing the song in your head and trying to lay it down on the sequencer. But to the musical theory-scientist, they'll go and break down every single note in a heavily-improvised jazz solo because there's exact mathematical science to why it sounds good and belongs there or evokes the exact emotion that it does. In other words, music theory helps quite a bit, I'm sure, but you can fumble your way through it just like you can with anything else. But unlike with coding where bad performance is much harder to quantify, you can very easily tell what sounds good to you - just fucking listen. Worst case, post it here and have anons insult you for it. That helped me out quite a bit. Instrumentation and chord structure are the more essential things, the former you will have to figure out by ear and the latter there's a bit of simple science to. It's probably as hard to figure out as anything else related to gamedev but you can certainly get the hang of it if you hammer away at it.
(3.73 MB 400x300 RashiGimmickTest.webm)
I hope no one minds if I start posting my progress here. The last character gimmick is finally done. Rashi Montel is the son of the corrupt and murderous rabbi of Bel Canto. For the character I wanted him to be religious, so I tried to tie that into his gimmick. His gimmick is something I call "the calendar." In short, at the beginning of his first turn, he chooses 6 "laws" to follow, one of which is applied each turn. On his 7th turn, it is the sabbath, so he takes no action, but regains all of his spent MP. Afterward, his calendar clears and he chooses another 6 laws. Mechanically, he is a melee, single-target Healer/Damager. He is the only character who has no direct control over his MP regen. As such, he does not need to spend turns performing MP-gaining actions and will regain all 100 MP every 7 turns. Because of this, his abilities all have a heavy MP costs and are intended to force the player to consider on what turn they should actually use his abilities. So far I only have two laws in the system: "Goliath Principle" gives Rashi 50% more Power (damage/healing multiplier), but he can only damage the enemy with the highest HP% or heal the ally with the lowest HP%. "Two by Two" gives Rashi +2 movement, but he can only move an even number of spaces. I intend to add a lot more, but it is easier said than done. It should be noted, to get these laws working, I had to create some new systems. Namely, status effects can now alter the pattern of viable attack and movement spaces. Hopefully I'll be able to use this system to create new and interesting things.
>>11664 Oh shit, are you the one doing that Uncommon Time sequel/whatever thing? Nice to see a familiar face.
>>11669 >Oh shit, are you the one doing that Uncommon Time sequel/whatever thing? I am the anon who was make the Uncommon Time tactics game that is a weird prequel/alternate timeline. Not to be confused with the anon who is making a rewrite of the whole game, or Duet Anon, who made a sequel to one of the endings and a rewrite of the middle of the game.
>>11436 Keep in mind that sweep is great for 2d top-down games (as most movements will be linear) but mediocre for 2d games a la Mario (where a lot of movements will be parabolic due to gravity). >>11471 Looks good, how is performance?
>>11710 >how is performance? so so, it's pretty bad on the editor, but it's mostly 60 on the built client, it also my GPU is fucked up and I need it permanently underclocked so it does not melt. It's a GTX 1070 for laptop
>>11750 Keep in mind that drawing tons of overlapping stuff like that can easily eat tons of frametime if you don't figure out some way to optimize the process, even with pure sprites a la doom it gets to a point where you're spending far longer resolving the overlap chain than actually drawing to screen.
To understand video games you may need first to develop a working tabletop game at least once in your life. I did.
>>11769 Mind sharing?
(523.28 KB 2046x1198 output.webm)
I'm trying to do weekly progress reports. So, since last week: >Campfire Developed pipeline how to add objects >Bus Damn, it was harder than it looks >Toilet For the sole purpose of working toilet door I've added interaction system >Firs level nemu >WASD as some anon suggested
(11.36 MB 1624x816 2020-05-22 15-16-04.webm)
>>11768 It's instanced rendering already, there is no efficient way of ordering that, at least for now. I am planning on rewriting (yet again) the memory pattern of the grass, but that will have to wait until I actually have a game
>>11874 Nice! Building systems as you need them is the best way to do it. I look forward to next week's progress.
>>11887 >more field vids >planted plants also deform I'm rock hard anon
(37.30 KB 800x180 arrow.jpg)
By the way, are there any examples/tutorials/source code for rhythm games? I've got something vaguely half assed based on overlapping rectangles but I feel there's gotta be a better way (maybe).
>>12100 Every arrow input has a timestamp and you run the game in terms of integer ticks with 60 ticks per second, and then check to see how far off the timeline you are. Arrows shouldn't use collisions, that seems weird
>>11887 >plants move around Aaand my optimization idea goes out of the window. Good job tho. >>12100 You can cheap out and just check the difference in coordinates between the last hittable arrow and its box: you need to make sure you're aligning things properly (and not, for example, check the coordinate of the left side of the box against the coordinate of the right side of the arrow) but it's faster. >>12106 >Arrows shouldn't use collisions The entire point of the game is to hit a key when two things overlap on screen, collisions are "what you see is what you get". Timestamps are an option but you're detaching the game logic from what you're actually showing on screen, and that can lead to really unpleasant bug hunts down the line as you can't trust what you see.
>>12244 >hit a key when two things overlap on screen, collisions are "what you see is what you get" This is how you get O2Jam bullshit i.e. lower BPM(original O2Jam and most modded clients)/scrollspeed(some modded clients) songs have bigger timing windows >>12106 >Every arrow input has a timestamp and you run the game in terms of integer ticks with 60 ticks per second, and then check to see how far off the timeline you are This is good, but be careful how you implement it to not get Osu bullshit i.e. if (BPM * note value) / (amount of ticks per minute) is not a whole number, chart will be inherently mistimed. Also 60 isn't nearly often enough, and while we're at it remember to NOT tie tickrate to FPS because again, that's how you get osu bullshit.
>>12273 I'm a retard, the last sentence should really say "make sure tickrate lines up with musical timing". Look up how bmson(file format for rhythm game charts) does it.
(1.83 MB 1280x720 2020-05-23 20-35-58.webm)
I made a bouncy ball type thingy. yay
>>12097 >>12244 You could also watch the grass grow if you're into that. The hang ups are the physics complaining about me running at 4x speed up (fucked up and the webm ended up 22 mega) https://a.uguu.se/gHNcMo32O7XD.webm
>>12347 Whatever thing unity comes with. I didn't make any changes
(202.74 KB 1268x367 Just Like Make Game.png)
Where's my nigga shinobro at?
(13.35 KB 808x627 ClipboardImage.png)
Welp, somehow I fucked up my raycast logic. At least it's partially implemented now
>>12273 >lower BPM songs have bigger timing windows Only if you tied the arrow speed to the song BPM, and there's no reason to do so by default. Using collisions (or any other overlap) means that timings are based on arrow speed and size (and target size), things you can control easily. >...chart will be inherently mistimed 1 millisecond discrepancies literally do not matter, it's a computer game not a CIWS. As long as you keep the original data and round in a sane fashion (not rounding the same data twice, for one), you'll be fine. >remember to NOT tie tickrate to FPS Good advice, but if your rythm game is dropping frames there are bigger issues.
(54.34 KB 657x527 1578062696198.jpg)
>tfw discover about butterfly farming >tfw that idea for a game is amazing >tfw is a female hobby wtf, why do all the best ideas right now are female hobbies?
(97.36 KB 500x571 perho finland.png)
>>12821 Stop whining and become the best butterfly farmer on earth. Become a proud gay man, instead of a faggot.
(906.91 KB 669x446 ClipboardImage.png)
>>12821 Do it anyway. Make money and get the adoration of butterfly bitches everywhere. Women (if I remember right) were mostly mobile players though so keep that in mind.
>>12584 I like that you are using the classic theme on Windows. I see it being used so rarely these days.
(56.37 KB 323x301 tank.png)
>>12825 >>12832 >>12837 Right now I'm working on making a roguelite version of Battle city. Maybe next game.
>>12821 >letting hobbies be gendered if you let that shit slide men wouldn't be allowed to play Style Savvy, which is a pretty good combination of store management and color theory puzzle game just do it and see if it's fun. At the very least it sounds like it could be comfy
(1.82 MB 1150x714 Unity tank.png)
>>12956 Finished my quick test. Unity graphics are so much better than anything godot can do. :( >>13006 I guess you're right.
>>12956 couldn't you look up a better design? At least fix the turret it looks like it got squished together.
(370.33 KB 378x756 THE GAY.png)
>>12821 >not getting into a hobby because women are in it That's some unironic incel-tier shit, anon. Gaming is a mostly male hobby but that sure as fuck doesn't stop women from engaging in it. You shouldn't be afraid of making a butterfly game.
(4.14 KB 256x224 42540651.gif)
>>13194 underage.
>>13202 Nigger, look up at any C&C mods there is tons of gud tonk design you could base off from, and now you got the qualms to call me underage?
>>13210 I'm remaking battle city.
(1.80 MB 1148x690 Unity tank.png)
pls rate.
(525.64 KB 459x242 dropped-items.gif)
Added the ability to drop items and pick them up again. Faster and cheaper than storing items in pots, but less secure since anyone can pick them up. I'm hoping it will help players trade with each other, move their stuff around more easily when crafting or looting, and maybe give gifts to strangers they encounter.
>>13218 A Black abyss of out technic colored tank.
>>13218 Reminds me of 3d battle city clone from 2004 or something Tank-o-Box. I've played the shit out of it as a kid. Tanks are as blocky as yours but they worked on textures a bit more.
(1.76 MB 1144x686 Unity tank.png)
(29.28 KB 1018x1198 2015-05-25_100527.png)
(52.72 KB 1289x1018 2015-06-21_105813.png)
(113.89 KB 1288x752 2015-08-01_114459.png)
(316.10 KB 3844x2120 2016-11-14_063355.png)
(126.41 KB 890x1417 2017-02-25_221007.png)
I just realised that as of a couple of days ago, I've been working on Hellas for five years. Here are some early screenshots to celebrate.
(16.16 KB 1037x636 Godot.png)
>>13276 Godot quick test.
(62.13 KB 1038x637 Godot.png)
>>13316 Actually even if godot has worse 3D I like godot much more.
(22.79 KB 300x300 hired.jpg)
>>13276 >>13279 >>13316 >>13336 You have already created results? I'm stunned. It's time to cheer you on.
(256.29 KB 466x420 tank.gif)
>>13361 working on making a quick twitter bait.
(1.77 MB 647x530 tank.gif)
>>13363 Could really be improved with a background image or tileset. Also so font work or HUD mockup.
>>13279 Looks interesting, and you've obviously put work into it. What are you writing this in? C? 2016 image looks comfy. What are you using for your GUI code? I've been using imgui and exploring Nuklear. I've also seen FlatGUI but haven't tried it.
(2.80 MB 5624x5379 2020-05-17_034307.png)
>>13703 C++, with SDL for graphics. The 2016 shot is pretty close to how the game looks now. Pic related is the latest tech tree. I rolled my own UI system, which took about a month. Basically an Element class that has zero or more Element children, and a bunch of subclasses like Button/Label/Picture that know how to draw themselves. Mouse events and draw calls propagate through the tree. I think most UI libraries work the same way, which means ultimately I probably didn't benefit much from doing it myself. That being said, I started out doing this for the fun and challenge of development.
>>13279 >>13259 Your current tile transitions look very nice, anon. How does water and paths compare to the 2016 screenshot?
Are there any really good books that help teach SDL? I've searched for some books and looked at the ones the SDL Wiki., the reviews for some of them on Amazon said they were either mediocre in teaching it or the code it shows off is buggy as hell. I'm aware of Lazy Foo's tutorial on SDL but I want something that takes a little different approach. An alternative that I can switch back and forth between the two. Also, been working an a texted based adventure game to brush up on my C++ skills.
(781.54 KB 800x600 2020-05-26-1212-06.mp4)
(2.57 MB 640x640 2016-03-24-1046-40_1.mp4)
Well, my raycaster is working now, so hopefully I can catch up with where I was 4 years ago
>>13808 >Raycaster Have you done the standard edge optimizations? Or did you have to make something different given that you're working on a grid?
>>13819 I just copied this implementation and reworded it so it was much easier to read http://www.adammil.net/blog/v125_Roguelike_Vision_Algorithms.html Basically, it's octant based and for each octant, it draws a slope with rise/run of (1/1) and (0/1) and checks the midpoint of each tile.
(3.00 MB 1033x629 battle city.webm)
>>13698 made the basic dungeon generator.
>>13822 Yep that's optimized, and far better than what I expected.
(85.67 KB 1288x747 2020-05-27_044712.png)
>>13778 Water looks the same, though recoloured a bit. Roads are now built by players and grant a movement buff.
(248.70 KB 1504x1472 dawn-pipeline.png)
>>13779 Lazyfoo is how I learned. My advice is to start very simple, and learn how to solve the immediate problem you're facing. If there's something complex I want to do, it can take some tinkering but eventually it turns out to be possible. Pic related: how I got stone carvings to overlap without being see-through (in SDL1.2)
(49.44 KB 199x214 substance.png)
kek, learning SD.
(819.70 KB 1036x635 godot.png)
(12.84 KB 570x543 ClipboardImage.png)
Rather than just doing a radial burst for raycasting, I wanted to specify a min/max angle so I could have cones of different widths and lengths, but this has been proving rather difficult. For example, if I had a min angle of 45 and a max angle of 290 (pic related), then I'd have to find those points on a circle, and then somehow map them to the 8 octants, figure out which octants it actually intersects, and then generate a slope for each octant. For some reason, it's just not coming to me. I made some progress with trig functions, but I feel like there should be another way as well. And even if there isn't, the octant mapping is killing me
>>14092 Work in spherical coordinates (radius, angle around the circle, angle above the plane of the circle) so that you don't have to do so much trig. It's been over a decade since I've done any real math, though, so I can't coherently describe the details. Surely there's a library out there with a solution for raycasting into a cone?
>>13834 >>14048 Reminds me of that 3D NES emulator.
>>14179 I don't think that's too uncommon. That's my experience as well.
>>14179 Sounds like everyone in the AAA industry.
>>14179 Yeah my hatred of current year video games is what helps drive my interest in making them. If no one else makes shit I enjoy, might as well make my own.
>>14092 Are you >>13808? If yes, working with tiles makes things a bit trickier than continuous game worlds, but nothing too weird. Don't bother finding points on the circle, just keep the angles. Then, do a first optimization pass: figure out which octants the angles are in (this is trivial), which octants are outside the cone, and which are outside. Then do the fully in and fully out octants as usual, so you'll only have to check if raycasts are within the angle for the 1 or 2 octants that contain the edges of your cone.
(180.44 KB 1920x1080 ebin neder.jpg)
>>14179 >>14186 I seriously hope this is a coincidence and not shameless samefagging
>>14379 >The reason I find the point is so that I can find the slope. It's smarter to find the angle given a slope and then check if the slope is on the right side of the angle: all you need to do is the tan() function, as you're working in octants you only need to worry about 0 degrees angles (flat slopes) as edge cases but those should be easy to turn into "is within octant" checks. All octant woes shouldn't be a big deal either, if I understood correctly just do "tan(slope) > angle" for half of them and "tan(slope) < angle" for the other half.
>doesn't even know how to program and uses visual scripting >makes shitty game simulator >somehow this gives him a job in the industry >makes another shitty walking simulator that makes $150,000 >now a full-time "game" developer How does this make you feel? https://www.youtube.com/watch?v=Y3Rs1z7it5M
>>15124 Don't care. I will never hire him.
>>15124 >How does this make you feel? Good, it's strong evidence than willingness to learn and hard work can make up for shortcomings in creativity and talent. >>15140 Exactly, and they make it really obvious when someone replies to himself.
While I still haven't found a solution for a tile-based conic raycast, I did figure out a clever way to do a polygon-based conic selection (eg for entities). >Get the min and max slope/angle for the cone from the user, as well as the range/distance >There's a thing you can use to see if a point falls above or below a line >Apply it to both slopes/line segments >Now have a pruned list of things that fall within an infinite triangular ray >Discard anything where distance squared is too big
(599.95 KB 1128x722 output1.webm)
(215.43 KB 970x606 output.webm)
Had an ambitious plan for this week but failed miserably. But, still, progress >Added hand to indicate interaction >Implemented walking into water >Made tutorial for art guy how to make levels. Last one took a lot of time
>>8735 You shouldn't learn C before C++. Look up "Kate Gregory - Stop Teaching C" on Youtube. It's some speech from CppCon that explains why modern C++ is different, and that learning C first is probably a detriment to your ability to learn C++ effectively. Also, do not worry about not remembering everything right away. You can't and won't remember absolutely every detail of a language. It's good enough to have a bird's eye overview of a language's features + being good at Googling shit.
>>17755 I've watched that video and generally disagree. I guess, ask yourself the question: do I want to be a decent programmer or do I just want to learn C++? C is a great first language. It's C++ minus all of the class, reference, template, overloading bullshit. Also, you have exceptions turned off right? You're not a try{ harder(); } faggot right? The only better language to learn first would be 68000, 6809, or i386 assembly. Those are not relevant. However, you're first game won't be relevant either so who cares. With a macro assembler and a few framebuffer and audio library routines (ROM or linked library) you can write a simple game in assembly quite quickly.
>>17878 >C is a great first language It's a very low level language with a lot of odd practices to work around its barebones features. That alone is enough to say it's not a great first language, and then there's the fact that even when you know C you'll use anything else for most of your code. >It's C++ minus all of the class, reference, template, overloading bullshit What you call "bullshit" is mostly really useful and convenient, especially for beginners. C++ has issues due to its sheer size, there's a million ways to do the same thing so you may easily miss the perfect tool for your job. >The only better language to learn first would be 68000, 6809, or i386 assembly. Confirmed LARPer, learning how ye olde processors worked isn't as useful as you think it is and it's surely not a good idea for a beginner.
>>17884 well let's agree to disagree. Let me address a few points >It's a very low level language almost a meaningless term, and also not relevant to the discussion. Games are just as easy to write in a "low level" language as any other. Use libs and ROM/OS calls. In fact, it's probably easier because at least your variables are typed, which I assume by your off-hand slur on low-level languages you believe high-level languages would be better suited for games. >What you call "bullshit" is mostly really useful and convenient, especially for beginners. this is self-evidently false. Beginners do not find templates, references, overloading, and the insane class and inheritance system convenient or useful. MAYBE references, but beginners fuck these up universally and would be better off restricted to pointers only. You know, like in C. >Confirmed LARPer I learned C and assembly on an Amiga 500. >learning how ye olde processors worked isn't as useful as you think it is and it's surely not a good idea for a beginner It is, and I'm just trying to head off future retarded posts like yours.
>>17923 >let's teach noobs how to iterate on loops manually and write tons of boilerplate just to get hello world on screen Retard, this is how you get YandereDev.
>>17924 He has a game. A shit game and he's a moron, but a game
>>17923 My post was shitting and mean, sorry. One of your good points: >even when you know C you'll use anything else for most of your code. True, I write most of my actual game in Lua using LuaJIT embedded rather than C++ or C.
>>17925 This is unironically my excuse for having shitty code.
>>17927 I have some code from 2013 or so, and boy is it bad
as long as the code works it's fine
>>17976 Not completely true. One must struggle for perfection. But shitty code is thousand times better than no code. I've refactored my tiny project in a serious way two times. And the code is pretty good for now. The pace would've suffered a lot if I didn't. On the other hand I write shitty code regularly to keep things going and rewrite them when I see better solution.
Going to repost this here since this is the more active thread. I'm thinking if making a game I could possibly try to sell for real money. >going to be a ff clone with shit from smt (mainly smt 2 and 4) and paper mario with real time combat. >player mp caps at 100 and regens 2-5 per second. All attacks and spells use mp so bigger spells like healing put the player at a disadvantage and if they spam the attack they won't have enough mp to heal to block attacks. >game will be very short overall. 4-5 bosses only with no normal enemies. Story based overall. >multiple endings. Going to start out with 4 but I want to add more if I can later on. >I have to put down some designs down first that I have in my mind to see if that is good enough so people will buy it. I want to have paper mario like character design but with more adult themes. I have a bunch of animations for backgrounds from my past games I can reuse. >I have some of the soundtrack done already. It's going to be mostly consisting of church organs, choir, harmonic guitar, with some synths. RPGs are still sort of niche enough that it is mostly overcrowded with rpgmaker games so something original would stand out more.
I just want to edit the dialogue in a Unity game for fun. Is it as simple as, or close to to as simple as, going through all the .assets files with AssetStudio until I find text files, extracting, editing, recompressing, and replacing the original .assets files? If not, I'll have to resort once again to autistically rearranging letters in gimp like the tech illiterate I am. Figure this is the place to ask.
>>17926 Don't worry, I don't hate C: it's a great language when you need performant code, it's just not the first thing I'd teach to a beginner. >>18247 >I'm thinking if making a game I could possibly try to sell for real money. Start with a short free game using similar (but not identical) mechanics, themes, and aesthetics, then if that goes over well do the paid game: a bit like Dead and MiniDead, if you want examples. There's a LOT of new games nowadays, so even a free indie game needs significant effort to stand out, and making money is extremely hard when you're competing vs AAA titles on sale for 10 dollars.
>>18247 Well, good luck. You using Unity or Unreal or similar? What are you designing the 3D models and animations in?
(17.18 KB 396x396 1583673420955.jpg)
I don't know if the Keyreal is here but I'm playing through Phantom Path and it's bretty gud. Speebot was also good but it didn't really hold my attention as well. I guess the whimsical robot land isn't as interesting as a creepy forest that slowly reveals it's secrets to you. >>18697 >it's just not the first thing I'd teach to a beginner. I don't know if I agree or disagree with that. I was using C++ for a long time then I switched back to C recently. It requires you to think differently. It made me think in a way that tailored custom containers/functions to the situation rather than always going back to general things like std::sort, std::vector, std::map, etc. The difference is absolutely staggering. On one hand there could be an argument that this would help beginners think based on the context of the situation rather than what tool to use, but on the other hand most programmers just want to get shit made.
>>18872 speaking of getting shit made, I switched from C++ to Godot and GDscript to actually build game prototypes. It was fun writing an opengl + C++ core, but it was so much work. I'd rather spend my time designing levels and characters in Blender and building actual game logic in GDscript. probably the most useful thing I learned was GLSL shader coding, as that is useful anywhere with any engine or framework.
>>18872 Lucky you, C++ was the first language I was taught and it was a whole year of "here's how you do for loops, here's some pointer math, let's pretend std doesn't exist". The one semester python course that tried to get us to write an actual program (a simple calculator) was infinitely more useful for understanding how to work with code. >>18926 Keep in mind that GLSL is a slightly unusual shader language, so you may see oddities when using other tools.
(172.53 KB 237x309 banana.png)
So I am attempting to teach myself C++ in order to design software, and also "Video Games". My current resources include "The C++ Programming Language (Fourth edition)", "Beginning C++ Through Game Programming (by THE Mike Dawson)", and "Accelerated C++". Are there any further books or resources that you might recommend? I spent two years at University studying this language as well, though had to bow out due to severe anxiety and other gay shit like that. I say this as i was considering using CodeAcademy when i sort my various medications out, and am wondering if this is worth doing? Apologies for the fact this is pretty much a shitpost. I'd very much like to properly get into the swing of (re)learning to program, as i think once I'm able to get past the mental block of starting and get into intermediate territory I will be fine, and able to progress. Then i will be able to just, like, make game.
>>18956 In 2008, took a 2-year programming diploma course at college. It took 4 years to complete because my grades slipped in the final semester, and I had issues filling a second co-op placement. Despite being a specialization, the actual material was very basic. We learned a variety of languages, but had less than a semester in each. It does not take a semester to teach Python when we've already learned the universal programming constructs from 3 other languages. We had two semesters in Java and it didn't even cover generics, just web shit. Point is, don't feel bad about school. It's shit. I learned more in the 7 years dicking around after graduating as a hobby, than what I paid for. Once you understand the basic logic and each language's quirks, the big thing to learn would be different algorithms and how to implement different things effectively
How fussed are you about compiler warnings? Do you obsessively fix each one or do you just let them stick around if they're not a bother? I just went ahead and fixed a few lingering ones, a few "Not all control paths return a value” ones and some "potential loss of data", neither of which were a problem in their applications. Considering how complex this shit will get down the line, I'd rather take care of everything down the line. I wasn't really bothered until a small change gave me a hundred "potential loss of data" warnings all of a sudden.
>>18956 I would strongly recommend the Scott Meyers series of books. His early stuff has now disseminated into every intro and intermediate book, so you'll get most of that information there, but once you understand how to write/use STL data structures and you're ready to move on to C++14 and newer features you should absolutely use Effective Modern C++ as a reference. Modern C++ has a lot of great tools that make software cleaner and more reliable. The abstraction layer is a little different from other languages, but it gets you the most static optimizations possible. That'll bring you up to early-modern C++ pretty fast once you understand stuff like move semantics. As an aside, I'd like to shill for >>>/t/. We have a programming thread that's pretty dead right now, but we'd be happy to answer any questions or recommendations. I'd like to see more general programming traffic over there. If you write any toy projects to test out language features, maybe drop by and tell us about it.
>>19161 >do you just let them stick around if they're not a bother The major compilers will only give you compiler warnings if you really are doing something wrong. >"potential loss of data" "Potential loss of data" probably means that you need a cast somewhere: https://codereview.stackexchange.com/questions/2610/avoiding-the-possible-loss-of-data-compiler-warning Having casts like that be explicit rather than implicit should clarify the meaning of the program anyways (i.e. converting a graphical floating-point coordinate to an integer pixel location). If it doesn't then check whether your types make sense in the first place. >"Not all control paths return a value" Ignoring "Not all control paths return a value" sounds like a good way to forget an edge case. It's preferable in most cases to use an if-else chain with a terminating else to ensure that you really do return a value, even if that value is NULL. Alternatively, if your function can't reach the end and you want to preserve your control flow you should instead use abort() or assert(false) (these will also get rid of warnings). If you have mutually-exclusive conditionals you should almost always be using if-else chains for those. Doing otherwise risks the possibility that you execute two conditional blocks when you should only ever have one. >Considering how complex this shit will get down the line, I'd rather take care of everything down the line. I wasn't really bothered until a small change gave me a hundred "potential loss of data" warnings all of a sudden. Compiler warnings are largely intended to catch bugs, not to enforce any particular coding style. Think of cleaning up compiler warnings as less "cleaning up your house" (which can be postponed for a long time with relatively little consequence) and more "not leaving food on the counter" (which usually does attract bugs, smells and other annoyances eventually). Also, there's no risk of missing a new compiler warning in a list of older warnings if you don't have any older warnings to begin with. Most warnings should take between a few seconds and a few minutes to deal with, so just do it. If you're using an IDE, you have even less excuse, since you can just double-click and be taken to the source of the warning.
>>18975 Upper-level programming and software engineering classes, though, are exactly where you'd be learning algorithms and how do to things effectively. But it's easy to fill time with the basics of a bunch of different languages, and I guess most people aren't able to teach themselves very well, so that's what the lower-level courses are going to spend their time on. Once you know how to logically implement a certain data structure, for instance, it's just a matter of drudge work to go on and do it in Python or Java or C++ or whatever language you want to use it in. I mean, sure, each language has quirks that make the implementation more or less efficient if you do or don't know them, but that's where experience with the language would come in.
>>18273 Generally, the standard practice for dialogue heavy games is to contain all the strings all in one place and have localization files. if you're lucky, you can search for something like "l10n" and find most of the strings in the game translated to every supported language.
>>18975 I don't so much feel bad about uni, so much as feeling bad that i know i could easily pass the course if not for my anxiety and (alleged) ADD. I was never particularly planning on using the degree they'd have given me, I know i'd not suit working in a company. My intention is to work for myself producing gaems and most probably software as well. Thank you, though. >>18975 So you'd suggest i acquire 'More Effective C++', 'Effective STL', 'Effective C++' (Third edition, I imagine second and first are obsolete), and 'Effective Modern C++'? 'Overview of the New C++' also looks useful, but it seems like starting with 'Effective STL' and 'Effective C++' are my best bet, right? Thanks, i appreciate the informations. I'll be sure to stop by /t/ with my shitty text adventure projects or whatever garbage i come up with to experiment with.
>>19486 Finally found the dialogue. It was stored as a bunch of .json files in a folder called "local" in the main game folder, not in the "data" folder at all.
(336.70 KB 492x376 typing words.gif)
>>19220 >Scott Meyers series of books http://gen.lib.rus.ec/search.php?req=Scott%20Meyers&column[]=author >Effective Modern C++ http://gen.lib.rus.ec/search.php?req=Effective+Modern+C%2B%2B&open=0&res=25&view=simple&phrase=1&column=def I am a backend programmer and work with golang, kotlin and php. Initially, I wasn't interested in game development, but why shouldn't I have a look. Thanks for the recommendations. Can you recommend a C++ book for experienced programmers that doesn't start with explaining the creation of the universe? By the way, I find rosetta code convenient for direct syntax comparison. http://www.rosettacode.org/wiki/Rosetta_Code >I'd like to shill for >>>/t/ I'll drop by.
>>20594 >I know i'd not suit working in a company You might want to try anyway, if only so you can get a better handle on how to stick to a schedule while maintaining quality. One of the downsides of working for yourself is the greater risk of not getting things done on time. You see it a lot with many of the great game devs, that they finally get away from being ruled by their companies and can make their dream game only to find that it's a lot more difficult making everything run when nobody's checking up on you. I'm not saying that you can't do it on your own, just be mindful that things can pile on quicker than you think.
>>20779 It's likely i might learn some things, sure. I am very antisocial though, I really don't enjoy interacting with people if ever i can help it. I suppose that's not hugely uncommon among spergy imageboard users. I absolutely cannot handle people looking over my shoulder at something im working on, either. This is admittedly something that might be helped by working for a company, but I'd rather not bother. Maybe i'll change my mind in the future.
>>20684 "local" for "locale" or "localization" I guess. Everything's obvious in hindsight.
(30.40 KB 322x292 output.webm)
Weekly progress >Added small splash animation >Continuous integration Dumped a lot of time to setup actions on GitHub. But now it automatically deploys to itch. >Pioneress now can do emojis Hope, CI will payoff in the long run. It took a lot of effort.
>>22766 What's your game even about? I can't really pin down the mechanics
>>23013 The initial idea was to make a game like Party Hard where you as a green insect-lizard alien go on a stealthy killing rampage.
How can you implement proper Vsync in a WinAPI/OpenGL application? Using wglSwapIntervalEXT ties your game logic to your refresh rate, which is unacceptable. WinAPI itself has a similar setting with identical issues. Using QueryPerformanceCounter and some math, I am able to set the window update rate to either the monitor's refresh rate or a custom value, that part works perfectly and doesn't limit game logic otherwise. However, the draw will have a consistent screen tear in the same spot - the game draws at the (seemingly) perfect rate but it has no way of knowing when the blanking interval is so it starts drawing whenever it can, making it offset from the getgo and resulting in the tear. The only way to fix this, as far as I can figure out, is to know when the vertical blanking interval is/when the monitor starts and stops updating the screen, but there's no real way to figure that out in Windows easily. DirectDraw/DirectX has a way to retrieve it but I'd rather not have to require DirectDraw just for something so small as Vsync (it's also not available on some computers and apparently the function I need doesn't even work on properly on Windows 10). Is there a way to figure out when the hell your window should start drawing? How do modern games and game engines do this stuff?
>>24352 >Using wglSwapIntervalEXT ties your game logic to your refresh rate, which is unacceptable. Use threads, faggot. Long version: you should have one (or more) threads that only do game logic and don't touch the framebuffer, and those ones don't wait on Vsync, then you have one (or more) threads that only do rendering and don't do a single bit game logic, and those wait on Vsync. Everything in the second paragraph of your post is a bad idea, don't do it.
>>24377 That'd work if OpenGL's Vsync was at all workable which it hardly is. Its driver implementation is questionable and it will not work on many random drivers (mostly integrated graphics chips), not to mention older OSes where it just won't work outright.
>>24381 If OpenGL doesn't work on your drivers try DirectX, there's not much else you can do. No, reimplementing DirectX yourself is not a smart idea. >not to mention older OSes where it just won't work outright How old are we talking about? Because at a certain point you've got to accept that stuff is going to break.
>>24386 >How old are we talking about? Windows 95 and 98. >Because at a certain point you've got to accept that stuff is going to break. Haven't needed to so far. I'm making a game in the style of games from the era, there is no good reason it shouldn't support those other than lack of effort on my part.
>>24394 If you want to support win95 you're really, REALLY going to need experience, it will still take a stupid amount of effort as you won't be able to use tons of modern libraries, driver support will be a colossal issue and will require ugly workarounds at best, and then there's the whole unpleasantness of old Windows itself. >there is no good reason it shouldn't support those other than lack of effort on my part. Your effort is valuable, to yourself at least: don't waste it on supporting dead systems if you can, if you really want to at least start by writing stuff that works on modern software and then adapt it backwards.
>>24352 vsyc on PC is not a simple or reliable thing. It's something PC has done a shit job of in standards, video hardware, and OS software. Wait on vsync in your main thread using some abstraction library like GLFW. Run your game logic in a spawned thread and synchronize positions and orientations for rendering using locks/mutexes. Use another thread to monitor rendering and logic thread, coordinate syncs of data, as well as monitoring high-precision timer in the CPU to limit framerate in case you don't get a vsync signal. You will need to track delta-time between frame updates and interpolate properly the orientation and translation data of all objects for partial frames.
(28.27 KB 1080x677 ClipboardImage.png)
(20.21 KB 727x495 ClipboardImage.png)
Hello Anons. Gigadev here. Last time on 8chan I was telling you guys about the platformer I'm doing alongside GigaMaidens, which I'm still waiting to hear back on my funding application. In case you don't remember, this is a video of what I have thus far: I haven't updated much because I hit a bit of a roadblock. I've been going back and forth on the sprite size I'll be using in the final game. In the current build of the game Kasha's sprite was drawn at 68x68 resolution (not its actual dimensions, just the resolution it was drawn at). I really like how this sprite looks, but the problem is that Kasha is not the biggest character in her game by any means - there are going to be characters larger than her, and keeping her sprite at this size will exponentially increase the workload for the game as a whole because larger sprites require more animation frames to not look choppy. I redrew her at a smaller size, 48x48 (again, resolution of the canvas, not the actual sprite size). We could go forward with this look, but I feel it doesn't look as nice as the bigger sprite. On the plus side the workload would decrease so I could potentially release stuff faster. For now I'm going forward with the smaller sprite and drawing a character sheet for each size to get a feel for how they would look in both sizes. Any pixel art anons around here? Could use some feedback.
Last month has been busy for me so mapdev keeps making cool additions to last demo map while I finish irl stuff that needs doing. A week ago, we released a trailer for Peripeteia though, so check it out guys https://www.youtube.com/watch?v=O60Pd7bDKQQ
Weekly progress >Finally made a menu >Arts guy made decent looking level >I had to refactor this unmanageable mess with few dozens of objects On the bright side I was forced to learn to make tiles that are bigger than one tile.
>>37572 Гадаю Video Game зайве в меню.
So what ever happened to Sim loli?
>>37849 Приємно бачити людей, які знають мову. У мене руки з дупи і я просто приймаю те що маляр намалює. Але маєте рацію. Передам йому.
(496.80 KB 300x126 cant_deal_with_it.gif)
Finally picked up one of my projects I started maybe over a year ago, to remake some flash game without the flash (because flash is trash and slow and dying). It's just too bad I still have to work while this gay virus has shut down the rest of the world. So I'll only have a few hours during the week and my weekend to squeeze dev time into my regular duties.
>>38897 feel ya do you get the pandemicbux
>>38910 nope, cant even be laid off because I'm part-time, in case I end up catching the my sharona virus I want to at least be able to say I tried finishing this project
>>24352 >How can you implement proper Vsync in a WinAPI/OpenGL application? No. >Using wglSwapIntervalEXT ties your game logic to your refresh rate, which is unacceptable. You can simulate multiple (or no) ticks of game logic per actual frame render. You should be doing this anyway to support systems that can either not render fast enough, or have the ability to output at a framerate higher than your game logic's tickrate. Event handling, however, is bound to the actual refresh timing. There is no way to avoid this, as event handling can also trigger things like window resizing. It's messy. Don't manually re-implement vsync. You might think you can easily do it if you can just figure out the initial timing, but that goes out of the window when technologies like freesync introduce a variable refreshrate. My recommendation is that you make your game support both borderless fullscreen and vsync exclusive fullscreen. The former is "free" vsync that suffers from timing issues, the latter has no timing issues but blocks. If you don't like either of them; too bad. You don't get a choice.
(17.51 KB 706x500 ClipboardImage.png)
Glowing up some sprites as I can.
I have a problem. <Over the years I have somehow gathered all of these programming-related materials thanks to friends, relatives, and random giveaways because I've had somewhat of an interest to get involved with game development. This is also leaving out that I have a copy of Clickteam Fusion, and I do have an actual Kaypro 10 to accompany the user guide. However, I'm not sure which of this I should actually start with (Not to mention the uncertainty of which ones are and are not useless). Anyone have any suggestions?
>>39785 First and foremost, you can forget Visual Basic. Some people still love it but until you want to write scripts for Excel there's no reason you should start with it. HTML5 and CSS is likewise not very helpful for writing actual software, but is fine for learning to design websites. If you want to learn the basics, the books in the lower left of your images are acceptable. I don't know if they're optimal I, too, was gifted C++ for Dummies and Computer Programming for teens by relatives who knew I "liked computers", but I skimmed the first chapters and and that was it. If you want a more computer science oriented experience, SICP (either the original LISP version or the Python adaptation) are good for understanding algorithms, recursion, and the Apply/Eval loop. Fundamentally, if you're starting out with no experience, you need to pick a language and learn the following: >Variables >If/Else (Flow Control) >Arrays >Strings >Loops >Input/Output >Functions >Classes and Objects If you can do those things then you can do basically anything. If you'd like more resources for advancing beyond the basics of a language, or perhaps you want a project to work on, you might drop in to >>>/t/9 which is pretty slow at the moment but has some resources available. For games development, C++ is probably more helpful, although C# is used by Unity and it's basically Microsoft's version of Java despite the name, so it's not the end of the world if you start with that.
(423.23 KB 1200x545 ClipboardImage.png)
>>39794 >>39785 Don't forget!
>>39921 Whatever helps.
>>39958 Anon two guys cant fuck, thats gay
>>39785 Start with an online tutorial on Python/Lua (they're both "simple" scripting languages that see a lot of use in games), then learn some java and C++ to get some experience with more complex languages, then look at some Unity tutorials to get an idee of how games are built.
>>39785 I agree with another anon. >Basic Forget about that >Java Sorry, but your books are horribly outdated. I don't think there are point in reading those
(67.67 KB 2034x338 fuckc.PNG)
>Following a tutorial online on Unity >Get a known bug with Unity where it displays 4 blank errors. >Tried fixing it, nothing worked >Start from scratch with a new version of Unity >Open the tutorial again >Open a new fresh project of Unity >207 errors What is my fucking luck
(165.12 KB 300x270 parry this.png)
What's the most optimized way to do timed hits? Specifically like Nero's EX-act moves, where if you time it right it forks into a different animation.
(4.30 KB 256x256 aadfhlin.png)
pls rate this shitty pseudo tileset I whipped up in under an hour. Is it comfy?
(689.31 KB 350x272 refreshing_taste_of_bepis.gif)
Made more progress on my project in a week than I have in the 2 years since I started it. >started using an IDE instead of trying to just work with a text editor with syntax highlighting >cleaned up imports so I don't flood the namespace >implemented a rudimentary level loader so I wont have to deal with autistically positioning objects for the level with manual coordinates and sizes >implemented a rather simple event that spawns objects into the level still trying to decide on how best to structure and organize code, like physics functions and entities, haven't started on sprites or scene transitions yet >>41368 Wew, I don't even know what to say about that. >>41490 What's the green thing?
>>41492 It's supposed to be an enemy, a crab. I only had enough time to do this much and not any coding
>>41493 Ah I thought it might have been a phantom, otherwise it's not bad, can pretty much tell what everything else it supposed to be.
>>41490 The smileyface is pretty jarring in the context of the rest of the tiles. I don't know what to say about it other than that.
>>41490 I like it. stolen. might make a digdug clone or something with it.
>>41368 Your Unity install is borked, at best. At worst, your OS is borked.
(524.93 KB 1000x774 Sylvarant_(ToS).gif)
(111.13 KB 2048x2048 alefgard-2.png)
Anyone experienced with RPGs mind giving out some general advice? I've already decided on all the main characters and the major story beats, so right now I'm trying to finish up the world map. It's going to be more or less based on the geography of the real world, though I'm stuck on two things . 1. I need to decide if I want it to work like the world maps from the Tales and DQ series where the player actually moves across it, or like the map in Pokemon Colosseum where you just select your next location before loading in. I'd given some small consideration to how the Dokapon Kingdom map's put together, too. 2.I'm worried (likely way too much) about having a bunch of unused space. The story as is doesn't have the player going to the equivalents of North/South America much, the first major dungeon's in the equivalent of Mexico but you leave for another continent soon after. One idea I had was to add in a few quests to finish before the party member you need to get into the dungeon would help out, though I could realistically only use that as an excuse to explore either north or south, not both. Another idea was to say fuck it and have another story in the same game that mostly focuses itself there.
>>41886 The first problem directly corresponds to the second. Ideally, you'd be able to walk through the whole world without skipping sections (which makes me think of Xenoblade), but if there's too much empty space, that's not really ideal. It's fine if you don't have to go to a few regions of the world, as long as there's something cool to do there. Perhaps optional dungeons and sidequest-strings, optional party members to find and so on. If you can fill the world with enough to do, you should. If you can't, a map screen would do you better.
>>42008 I think I could easily fill one of the continents with sidequests and the like, the game's loosely themed around ancient civilizations and one of the secondary characters is an archaeologist. Ideally I'd have at least one story-mandated location on each continent, and there are a few mini dungeons I haven't placed yet. I could also try scattering some optional party conversations around, I think Tales of Symphonia indicated them with glowing rings on the map. I think as long as I make sure two consecutive story events aren't on opposite sides of the globe before fast-travel unlocks I'll be fine.
Weekly progress >Implemented dialog system. It's simple AF, but took a lot of time to figure out.
I was making a game about two years ago in Godot, but I gave up because it just overwhelmed me in addition to all the shit I've dealt in my life back then. I also had no musicfag and the drawfag eventually left. How should I get back to it? Still use Godot? Did it get better since then? I used to have quite a few annoyances with it. What kinda important features has it received? Do you guys think I could recruit a drawfag / musicfag if I showed them a prototype? (need to record vid, still have the project of course).
(3.58 MB 1280x720 2020-06-23 00-34-26.mp4)
Fully working signal connections with Qodot. This means triggers, buttons, doors, and etc. This was a huge bitch to pull through.
(40.80 KB 1306x743 ClipboardImage.png)
Hope it looks good. What do you guys think?
>>44240 The art is inconsistent. Ignoring the leftmost character and going from there: The rightmost girl(?) is by far the most detailed, much more than the other two. Her body is decently detailed but her hair is quite plain by comparison. The middle girl (red haired one) has some plain spots on her body (that red shirt section above her hands). Her hair seems to mesh better with her body than rightmost girl's does. Redhaired girl's hair has very bright highlights that do not match the highlights that rightmost girl has. Her white clothes also have different shading than the other girls do, the shades don't contrast enough from one another. The left squirrel-looking girl looks goofy. Her body's a bit pudgy for her head, her feet look like they don't belong in this artstyle at all, her hairband doesn't look like it's wrapping around her head circularly and the outline on her head is inconsistent with the other two - her darkest shade is several pixels thick towards the bottom which is not the case for the other two girls. The first girl would match the third fine enough if you fixed the few things I mentioned. The second looks like she was ripped from a different game, due to the differences I listed. As for rating the character designs themselves, I have no fucking clue, really. Not something I'd go for but I'm far from an expert. The rightmost girl is the most interesting design, certainly. The other two aren't all that interesting.
>>44243 Agree with analysis. And I'd add that big one feels dirty. I think that your problem is too many colours and too many details. I'd pick 16 colour pallet and see how it goes. If absolutely necessary, I'd pick 32. But I'm no artist, I'm just an anon.
(525.64 KB 459x242 dropped-items.gif)
Items can now be dropped. It allows easy sharing with others, and bypasses the need to place containers. The downside is that anybody can stroll by and pick them up. This is a stop-gap until I add direct trading between players. (More formal and asynchronous trading can be done through merchant objects, but it isn't very lightweight.)
>>44791 I honestly thought your game would have had that already? Nice to see you continue your work, though.
>>44801 Before this you could "drop" to delete items, like in WoW.
(11.43 MB 1280x720 spongebob.webm)
If i were wanting to create a (relatively simple) roguelike as a small project to test how well i'm grasping C++ (As i am currently teaching it to myself) would you suggest I use a library/selection of libraries (And if so, which might you suggest?) Or attempt to do everything completely from scratch? In the latter case, correct methods of drawing/redrawing the screen/information (Beyond system("CLR")) would be the main thing to learn, I think.
(5.99 KB 640x400 rogue.gif)
>>44871 If you're trying to make a simple (for now) game, and are still learning C++, I'd suggest starting by making it text-based. Starting out with graphics might mean trying to focus on too many things at a time, and will distract you from the more game-y elements that exercise your C++. That being said, if you're a beginner programmer then making the transition from text to graphics later on might be impossibly hard, as your code probably won't be clean and modular (which takes months or years of experience to learn how to do). So this decision might haunt you for the entire project. My general advice is to start very simple. It's easy to add more later, and extremely rewarding to have something playable early (within a day ideally). See the first screenshot in >>13279 for an example of this.
(43.02 KB 822x505 ClipboardImage.png)
>>41490 Still messing around with tilesets and stuff. I think the best approach would be to do it sort of how DF handles it, that is, have a foreground and background color for each drawn image, and that way I can have a generic picture for water or sand, and depending on the color, it denotes what it actually is (eg water or lava). I was trying to keep everything to 3 colors+transparent, but doing a simple grayscale could take advantage of that anyways, and some things like tiles and special objects could have actual colors to them. Anyways, which staircase looks the best? I'm thinking the third one, simply because it doesn't touch the wall. I think having a firm black outline on things is helpful to reduce visual clutter.
>>44879 nice. stealing these too. 3rd staircase is best. needs more sneaky lizards. >>44877 take a look at Brogue source code. It's simple C and gets it done. The game is pretty good too.
(13.26 KB 490x738 ClipboardImage.png)
(4.92 KB 352x438 ClipboardImage.png)
>>44243 Working on some changes. I removed one of the dark tones from the big girl. I'll work on adding another tone to her hair. Regarding Kasha herself, I did adjust her headband a little bit and gave her some visible legs. I think you are right that she looked too stumpy earlier.
(37.16 KB 832x448 ClipboardImage.png)
Actually did a bit more work on my engine today, and the drawing is almost where I want it to be. No gameplay yet, but this should hopefully reduce the time it takes to make big changes to what I'm drawing >>44980 >>44879 I've been trying to keep everything two-toned where I can, and using black and negative space if possible. I added windows, the squares that sit on the walls here. The open window actually draws the )( shape in hard black on top of the underlying wall tile, to help imply you can look through that tile, without it just being a floating square like the closed window. I'm not sure if I want the doors to be placed on wall or open tiles. They feel better on the walls.
(21.80 KB 1052x751 ClipboardImage.png)
Small update. Want them to look as good as possible. Still haven't changed much about the big girl's hair yet. Mostly been working on fixing Kasha's sprite. Incidentally, their names from right to left: Kasha, Sachi and Milchi.
(10.15 MB 1280x720 2020-06-26 00-34-42.mp4)
Just finished sliding door implementation.
(8.71 MB 1920x914 2020-06-28 11-49-08.webm)
In the middle of reworking the grass system to use ECS for as much as possible, a fun and interesting effect emerged, so I decided to make a webm of it before fixing it I locked the fps to 30 so I don't have to hear my fan going at 100% all the time. It's currently capped at 90 due to the character animator, would be 140 otherwise
(19.28 KB 465x459 ClipboardImage.png)
(28.55 KB 808x627 ClipboardImage.png)
(273.00 B 32x16 ClipboardImage.png)
>>45230 Been messing around with more art assets, rather than the programming. I've been making a few more draw helper things that make eg rendering tiles easier. Again, I want to use grayscale on the tile, but then have fore/background colors. The first pic is a mockup, the second is ingame. The third shows how the doors appear on the spritesheet - anything solid white gets converted to transparent (magic pink in this example), while solid black is used to block out the shape of the object.
>>49987 That's pretty trippy dude.
How retarded does a Tom Waits inspired 2d platformer sounds?
(196.08 KB 500x500 untitled.gif)
WIP vivian running animation. Probably gonna add hair and headwear physics, color and add/fix details in the design later.
>>44877 For C++ programmers, please start with graphics instead. Maybe it's because I was underage and retarded at the time, but it took for fucking ever for me to make the switch from text to graphical programming. In all fairness, the real issue was that I wasn't really sure what to look for so I kept on looking up "How to display pictures in C++" when I should have been looking up "C++ graphics library". And then I fucked myself over AGAIN by thinking that I needed to use OpenGL, but I was too retarded to actually know how to even load up OpenGL for my project, much less even use it. If you're just starting out and want to make 2d games in C++ Use text based games to at least learn the very basics of the language. After that GET THE FUCK OUT OF THERE AS FAST YOU CAN, TEXT BASED GAMES WILL FUCK YOU OVER AND YOU WON'T LEARN ANYTHING FROM THEM. First, download SFML. Look it up, it's stupid fucking simple compared to anything else out there and it's actually useful for professional projects as well. Step 2: /viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxjcGZhc3Rmc2QyMDE3fGd4OjRhOTVkYThjMGY2NjM0Yzc Huh, what a strange string of letters and numbers I just typed. Wonder if there's a Google Docs page that matches it.
(1.78 MB 1601x1149 ClipboardImage.png)
This is from Factorio v0.1 Always remember to keep working on your game, because it might turn into something great.
(19.05 MB 800x600 StoryUpdate9.webm)
>>11664 It's been a long time since I've been able to work on this project. This is another story update. Hopefully the writing doesn't feel off here. Those cut in images I have will most likely change. I was attempting to give a sense that as the days went by the numbers were supposed to show Teagan's diminishing mental state, but I don't think the effect worked at all.
(4.38 MB 1020x428 2020-07-06 17-11-52.webm)
Bugged ass shit grass is not properly working after 5 weeks. It is almost feature complete, but only works on editor, it silently fails to render in the player. Not to mention the constant crashes on the editor If anyone think can give a pointer on what could be wrong, more info here: https://forum.unity.com/threads/graphics-drawmeshinstancedindirect-not-working-on-the-player-but-working-on-editor.925988/#post-6058079 On another note, the new UI system is pretty dope. If I can painlessly make nested tooltips, it's a done deal and we will slowly replace it all with the new system
>>55514 The UI looks cool, but i would be worried that it's taking up too much space and generally feeling cluttered.
>>55530 one of the best parts of the new UI system, it seems UI scaling is trivial to place in a config window so you can just shrink it if you think it's too big. Also, that's the size of the elements on 1920x1080 too, so it uses significantly less space when maximized On the other side of things, it seems this fucking thing also does not work on the client, only on the editor
>>55514 >Attributes and unsafe contexts Working at a more advanced level than I am, so not sure where to even begin looking. When was the last time you had grass rendering correctly? What changed? What are you able to do to make any grass render in the player? Does that include testing it in the Unity player, or does it still fuck up with a compiled game exe too?
>>55611 >When was the last time you had grass rendering correctly? What changed? What are you able to do to make any grass render in the player? No idea what exactly was the change, which is the biggest issue. >Does that include testing it in the Unity player, or does it still fuck up with a compiled game exe too? Exclusively does not work on the compiled exe, which I was calling 'player'. Which is why I didn't notice the issue until it was 'too late' to figure out exactly what broke it
>>55691 I see you're using a lot of #if preprocessor directives. Is the game being compiled/played with a different build than what the Unity tester is using?
>>55696 the Begin/EndWrite are a bit buggy on the player, so I was attempting to switch to SetData instead (which is also bugged). I kept both codes instead of commenting/deleting one of them, in case I want to try again. those directives switch between which will be used
(137.82 KB 454x603 oop.jpg)
I'm beginning to understand why python is a meme language. What little I've built using pygame uses ~100% cpu, the majority of that taken up by pygame's own functions. At least it's still faster than flash games...
>>55838 >What little I've built using pygame uses ~100% cpu Are you sure you haven't left framerate uncapped?
>>55863 Nope, my framerate should be capped at 60 if the docs are to be believed.
>>55838 Probably not the problem but check you aren't using a while loop that's waiting for input or some other sort of undefined end without adding a (very) short delay. That can cause cpu usage to spike similar to what >>55863 suggested. >>55898 Never believe what the documentation says, always check for yourself. There's so much pajeet-tier code out there it's jaw-dropping even in what are otherwise popular modules. And while you're at it do some proper debugging over your own code to make sure it's not something you fucked up, even if it looks like it's mostly pygame functions it could be to do with how you're calling things. You probably shouldn't be using python for anything like a serious 3D game anyway, at least not at the low/engine level, but if it's a basic prototype it's fairly unlikely to just be a language issue unless you're on a complete toaster so find out what's going on.
>>55907 I actually did run it through a profiler and the most time consuming call was the engine's own clock object. At most the time spent on any of my functions was less than a tenth of a second (the most expensive seems to be iterating over a list to check if an enemy should be added to the level), but here's the top most time consuming snippet compared with my function: ncalls tottime percall cumtime percall filename:lineno(function) 1455 21.520 0.015 21.520 0.015 {method 'tick' of 'Clock' objects} 1455 0.028 0.000 0.045 0.000 level.py:77(spawnEnemy) Where tottime is total time spent on the function, percall is the quotient of tottime divided by ncalls, cumtime is time spend on this and all subfunctions from invocation till exit, and last percall is quotient of cumtime divided by all primitive calls. Now nothing is threaded, so when it uses 100% cpu it's only on one core, but that still seems ridiculous. So now I'm wondering if I should delve into a different language or at least finish this so I'll have a base of reference when I do pick up a different language. It's only 2D game by the way.
Could someone explain to my why Python replaced Basic as the "go to" language if you want to start learning? Aside from Basic being (Well) "basic", the only problem I've actually heard that people have with the language is the abuse of the "GOTO" function making the code damn-near unreadable (Which has gained such widespread infamy that anyone with the most simple surface level knowledge knows to avoid it like the plague). Meanwhile, with Python, the problems I've heard of with that language is that it makes things "too simple" on the programming side and side-steps a couple dozen functions that the programmer should be aware of (With the potential of spaghetti coding having a much higher probability).
>>56224 I just don't like Basic because it's syntax is shit. I think Python is good for data science and utilities but it's not scalable. Reason being is that Python's programming standards (it has a name forgot what it's called) literally discourages people from making empty declarations of variables, instead you declare it when you need it. That might sound good on the surface but it actually makes the code hard to read if the program gets too big. When you have a header file declaring variable names, it gives you a bird's eye view of the application structure and you see all the variables that are available to you. If you're using a good IDE then most of the time you don't need to consult documentation. With Python,even with an IDE you NEED documentation to know how many members are in each class, what "types" you need in function parameters, etc. I think larger programs should use C#, C or CPP, while python should stick to minor stuff. Too many idiot CS graduates think that Python and JS are the end all be all.
(497.53 KB 500x369 WHAT_I_LEARNED_TODAY_IS.gif)
>>56224 Probably because python has lax syntax and is pretty forgiving. For as long as I've been using it I still barely understand some core programming concepts, but it's fast to write a python script and there exist modules for just about anything you want to do with it. As I'm finding out though, probably not the best choice for games with all the overhead.
>>56224 >not too hard >perception that it's quick to develop in >became part of university curriculums for comp sci, mathematicians, scientists and for engineers training a generation in it's use >has a large amount of third party libraries
>>56084 >15ms on clock calls This is suspiciously close to 16.6ms (aka 1/60th of a second), so I suspect that their framerate cap implementation is literally "check if enough time has passed, if yes start working on the next frame, if no check again": this is of course the dumbest possible way to do it. >>56224 Because it's fantastic for quick scripting. You want to do some simple manipulation on a few text files? You'll have it done within a few hours, even if you don't know the language at all.
>>56660 >What makes that a bad way to handle fps? The fact that it hogs 100% of the cpu to essentially ask "are we there yet?" nonstop. In more technical terms, it's called a busy loop and generally it's something you want to avoid.
>>56224 Python has very easy syntax, an expansive built in library and actually gets used out in the real world for more than a few purposes, mostly purposes where you need to program but don't need to know how anything deeper works like data shit where you anyway just grab some library that's more or less a wrapper around embedded and hyper-efficient C code.
>>56574 According to pygame docs, Clock.tick() actually uses SDL_Delay to get the milliseconds passed since it was last run in order to keep a consistent fps. Using Clock.tick_busy_loop() is more accurate for using pygame.time.delay() but actually uses up more cpu time.
>>56727 Try limiting FPS to a few other values (30,90,120...) and see what happens: if the game remains at 100% CPU while holding all those framerates, the frame limiter is to blame.
>>56674 Oh, I thought it was implied that there was a sleep being called in there somewhere
>>57314 Well like the profiling showed it was the game's Clock object at fault, so setting the fps value to anything else only changes how fast or slow the game updates, cpu time remains at 100%. A better solution would probably be to not use pygame at all.
>>56248 >>56281 >>56513 >>56574 >>56701 So, summing it all up, Python is the new Javascript, except people take it much more seriously.
>>57552 How did you reach that retarded conclusion?
>>57511 >so setting the fps value to anything else only changes how fast or slow the game updates, cpu time remains at 100%. Have you actually tried that? There's also a significant chance that the Clock.tick is legitimately being called too much or running abnormally slow, and in that case changing the FPS lock would show noticeable differences (lower CPU usage on sub-60 fps, failure to meet target above 60 fps). >A better solution would probably be to not use pygame at all. That's a stackoverflow tier solution, "welp just use something else" is a last resort not the first thing you do when you find a problem. >57552 >look mom I mass replied No (you) for you.
>>57576 Sorry, I mean Java, not Javascript. As, for the reason I drew that conclusion: >Easy as shit language to learn for the most retarded >High cross-platform capabilities with a low barrier of entry >Schools and institutions championing it as the "one all end all" language for programming >An absolute joke of a language for work that requires more than simple CS charts and equations, and/or proofs of concepts >Gained most of it's popularity thanks to India
>>57585 Java is "okay" but C# is a proper implementation of its core concepts.
>>57580 I did try it, and that's my result. It calls clock.tick() once per frame, any number passed to tick() makes sure it never exceeds that many frames per second, so tick(60) limits it to 60 fps, using SDL_Delay. Not that it's causing a noticeable slowdown when I test the game on my machine, but I imagine on slower systems it could be a detriment, "works on my machine :^)" isn't an excuse I want to give anyone else if it performs poorly.
>>57585 Java is not easy nor does it have a low barrier of entry. Also Python is interpreted and Java is compiled, so not comparable. Python is more akin to Perl.
>>57588 >Java is "okay" but C# is a proper implementation of its core concepts Java can effortlessly hotswap debug mode code for instant testing C# needs a 10 second domain reload (or it's a Unity issue, rather than a c# issue) I miss being able to instantly swap code during runtime from my java days
>>57591 Try looking at github.com/pygame/pygame/issues/331 and github.com/pygame/pygame/issues/304, there seems to be something fucky with the clock. Also, which version of pygame are you using?
>>58003 I was going to riff on pygame having such a fundamental issue, then I realized I'm using the C# binding for SFML and it has a fucking typo in Vector3f subtraction code so (Though in its defense, its a 2D library and doesn't really use v3f's for anything)
(195.94 KB 700x700 1559846175201.jpg)
After five years of development with en effectively singleton Client class and class statics everywhere, I'm finally adding player grouping to the game. Which means regression tests that need multiple clients online at once. So now it's time for a long-overdue refactor. Pic related is how I feel about this.
>>58003 Latest version, and I'll be damned it was the sound module causing 100% cpu, I haven't even started adding sounds yet and already it sucks. Reading through issue 331 revealed there was an issue with sdl_mixer so quitting that module after initializing pygame practically nullifies cpu usage. Supposedly fixed in pygame2 but I haven't tried building that yet.
>>50554 sound great to be honest. All the albums could be a source of interesting shit. Though in my opinion, the themes would suit a horror or adventure game more
>C# >Made a simple Coord struct that holds an XY position expressed as a byte >Remember reading some shit about how custom structs should provide their own equality comparisons and override Equals and GetHashCode (and optionally == and != ops as well) >Decide to actually profile this shit >32,768,000 inserts (65536x50 repetitions) took about 10 seconds >Strip it back down to the default implementation >65536 inserts took over a minute before I got bored What the fuck, I didn't realize it could be so slow with the defaults.
>>58198 Well, glad to see you found the issue and a fix. >>58495 Default hashing and equality checks in C# are very slow because they have to work in general, custom structs can use much faster checks that would however completely fail on different data types. Hashes in particular are fucked, the default implementation is both slow and likely to work poorly in realistic use cases. See here for more info: devblogs.microsoft.com/premier-developer/performance-implications-of-default-struct-equality-in-c/
>>57552 No, Python is actually good at what it's designed for. Treat it like an expansive scripting language which, frankly, is what even 'serious' programmers need 3/4 of the time unless you happen to work in a specific field and what retarded diva-programmers working at a start-up need constantly and it works just fine. Try using it for what I'd classify as engineering problems like a (3D) vidya engine or whatever and it falls over as expected, and as with any language give it to a pajeet and he'll write inefficient code Python will never be blazingly fast but a lot of the time speed issues in production code are as much down to terrible pajeetcode as they are inherent features of it as a language. The real problem is as >>56248 says which is when people treat it as the only solution and more generally in tech nowadays nobody ever specs a problem then picks the software to that spec instead they just pick whatever's trendy or whatever they know and beat it into the shape they need after the fact and then if there's a further issue they'll just eat up venture capital to buy more hardware and more employees to cover it up. Now that said even if your spec says you need a scripting language you'd not necessarily pick python, if you wanted to embed a scripting language for game logic into an engine written in C/++ you'd maybe go with Lua or something for speed reasons. Define the problem first then pick the tool, don't try and hammer a nail in with a screwdriver just because you heard Linus really using likes screwdrivers as cheap buttplugs.
>>58885 pleasing digits. Python gets used a lot in the scientific community, also neural networks and AI stuff. Mostly because specialized packages written in asm/C++ do the heavy lifting while the researcher glues it all together using Python's fairly flexible syntax and data containers.
Fresh progress >Working on AI Damn, it's harder than I thought >Implemented action list >Still have no idea how to implement GoToPostion action Hope to resolve it until next week's report. Also I think I need sensor system.
>>59918 >Still have no idea how to implement GoToPostion action You mean in terms of pathfinding or in terms of fitting together with the rest of the code?
>>59925 In terms of overall architecture. What should be responsible for what and how to ducktape it together.
>>59932 As I don't know your codebase, I can only give very general advice, so take this with a grain of salt. You might want to have an Pathfind function which calculates the best path between the current position and the goal position, and produces an array of intermediate steps as a result (they might be adjacent tiles in a tile-bases system, points that can be reached in a straight line by the unit in a classic RTS, or whatever else you need), and then you have some simple ActionMove that takes your character from one step in the list to the next (for example, the character first walks a tile down, then when it reaches that tile it walks one tile left, and so on) and handles all animation-related stuff: that way you can reuse the ActionMove for things that aren't properly pathfinding-related, such as cutscene animations, auto-kiting, and the like. The idea there would be to separate what the character does (a simple sequence of simple movements) from what the character thinks (the path he found to his goal), that way you can change one without messing too much with the other: however, this approach requires either the character to stay still until the pathfinding is completed (not an issue for small maps and small unit counts, or super-fast pathfinding code), or the pathfinding code to be very robust to allow for the character to start moving before the optimal path is found (and not risk getting stuck in a loop).
>>59286 Yes, and that's a great use of both Python's strengths and the strengths of whatever other language the expensive work is done in.
(2.60 MB 360x205 its_a_feature.gif)
>implement a new feature >player can pick up and carry objects >Oh right I need to do collision check when the player throws it so it doesn't fall through the level >add collision check after update to y position >throw object >it lands on ground and decelerates until it stops moving >add collision checks after update to x position >it should stop moving when it bumps against the side of a platform >it accelerates to infinity in one direction regardless of direction it was thrown and loops back around from other side of the level and just keeps going Alright.
>>60481 you are dividing by zero
>>60481 Well check your math then. Not sure why it goes in one direction only but if it loops it hits +2^32 and overflows
Hyperstimulation
>>60481 You probably fucked up the friction math.
>>60913 I finally did figure out what I was doing wrong, I always end up overcomplicating my conditionals, and usually it ends up fixed just by reducing to one or two and using ternary here and there when there's just one value I need to change within the code to be run. Same thing with my collision check, I completely overcomplicated the conditions to check because nothing seemed to be working, then I finally realized I didn't need any of it once I realized the problem, and now I've got half the lines of code I had before and it actually works. Before I had twice the code, and collision would break depending on what it hit first (hitting the side of a platform and it would fall out of the level, hitting the top of a platform and it would end up inside other platforms).
I have unfortunately decided to try making a text adventure game. What do?
(497.11 KB 846x1200 1589415267332.png)
Man I'm seriously blackpilled. I just got done binge watching former developers from Looking Glass Studios play through System Shock 1 with commentary and talk about what gamedev was like back in the day. It sounds so day and night different from gamedev nowadays. In culture, work ethic, what you could get away with, trust, etc. Apparently most of the studio was frat kids from MIT who played sports and stuff, but also had weird hacking hobbies like lockpicking, along with stuff like D&D. It sounds like they all kinda just got to do their own thing and everyone just threw what ideas they had into the game almost like a modding community. It's the type of thing that couldn't exist today, because 90% of this industry is composed of faggots who only care about shoving their politics into everything and it only takes 1 retard to shit up the entire thing. Even the office is so 90s it seems like it's from another time all together. It's hard to know this and many other game studios like this once existed, when I'm trying to go on in full hermit mode to crank out a game.
>>61058 We're all the shitheads and /v/ is your office, except you're the only one capable of doing work.
>>61058 Then stop fucking around and get some /v/irgins on your team.
>>61058 Yeah, I was thinking about professional gamedev. But it's business. And it's not that fun. Guess, I'm amateur forever.
>>61057 Practice your writing, because you're going to have to write up a lot of content. >>61058 The main take away of this I think should be why older games were so great in the first place, many of the developers weren't just obsessed with the idea of creating a game, gaming wasn't their only hobby, they didn't go to college to learn how to make video games. They had life experience in a variety of areas, they had other hobbies and skills that lended to their creativity in creating games. Inspiration often comes from the least likely of places.
>>61058 >I just got done binge watching former developers from Looking Glass Studios play through System Shock 1 with commentary and talk about what gamedev was like back in the day. Care to link it?
>>61057 Write an i/o loop, sort of like a REPL. Your game will parse user input the way a programming language does, though the complexity of it will depend on the needs of your game. Consider, for example, if you wanted to implement Final Fantasy I as a text adventure. What are the verbs available to the player in that game? There's "go" to which you can attach a location, or a direction. You "talk" to NPCs, and in some sense, to shopkeepers. And of course, you spend most of your time in combat, where your party members fight, cast, use, or flee. Since this is all in text, you of course need to "look". So you loop over whatever "get string from STDIN" function your language has, split the string by spaces into a list of tokens, and then check that the first token makes sense in context. By "context" I mean that you need some sort of game-state object that stores the player's location, possessions, plot-flags, and NPCs. You can't "fight" without something to attack, you can't use an item you don't have, etc. That's basically what you do. There are a lot of questions you could then ask, like "with what data structure should the game world be represented?" since it's effectively just a graph, where the vertices are game-world locations, and the edges are connections between those locations, like how you might go directly from your bedroom to your bathroom or kitchen, but not to the White House. There are many data structures and algorithms for dealing with graphs. Or, you could just use one of the existing interactive fiction engines. The obvious choice there is INFORM, of which version 7 is the newest. I myself preferred 6, since 7 is so amazingly user-friendly and convenient that if anything doesn't work perfectly, you're fucked. It's open-source, and there's inevitably defective tooling available for various editors. https://inform-fiction.org/ Obviously, writing your own is more fun.
(1.81 KB 160x160 ClipboardImage.png)
(1.89 KB 160x160 ClipboardImage.png)
>SFML/C# >Use vertex buffer for debug drawing >Treating vertex data as quads or tris? No problem >Treating vertex data as lines or line strips? Everything is offset by one pixel And actually, even if I do the same A-B-C-D-A coordinates and stuff, the lines overlap at the origin and don't quite meet at the opposite far corner (regardless of whether Y flip is applied). Kinda retarded
>>61174 Or put another way, quad/tri ends up with pixel perfect drawing (eg 0..1 inclusive) but using the line buffer uses 0..<1, 1 being exclusive
>>61081 This is the playthrough I initially was referencing: https://invidio.us/watch?v=-LDiTZ89j-Q https://invidio.us/watch?v=wMXguHtfEs4 It's fucking long and I'd use it as background noise, although some of the devs are in the chat channel asking questions. I know nothings2 is one, but there are others. The devs will randomly say some dubious things though, like how Garrett in Thief should have had Dishonored's teleport ability. How Deus Ex 2 is better than 1 because the dialogue around whether or not you played stealth or lethal couldn't detect if you were just bad at stealth in 1. Really shows how the devs changed for the worst imo, and how they ended up producing turds like Bioshock Infinite.. It's still worth listening to just to see what their thought process was back in the day. There's a ton of these though. SS1 & SS2 with original devs and the new studio doing the remake of 1. I haven't watched this yet. https://invidio.us/watch?v=4IzNzVAxk8E https://invidio.us/watch?v=iWFylN8bRdQ Thief Gold with 1-3 of the devs. There is a moment where one pops in for a brief period, then 2 of them drop off. This is unlisted for some reason. It has one of those corrupt GG faggots heading the interview, but he actually does a surprisingly good job. Terrible playing skills aside. https://invidio.us/watch?v=mHITDn2qnhg https://invidio.us/watch?v=6AcmmDeJS1k There's probably others I listened to but I forgot.
>>61069 Seemed pretty fun back in the day. I'd rather do what they did for a day job instead of what I'm doing now. 100 hour work weeks and all. >>61073 Nintendo actually specifically hires people who don't play games for this reason.
(963.68 KB 1902x899 ClipboardImage.png)
(910.29 KB 1894x894 ClipboardImage.png)
>>61177 real good shit man, thanks. Here are some cute doggo pics for your trouble and some Dude Sex restrospectives https://www.youtube.com/watch?v=oaJ18p4QH2Q https://www.youtube.com/watch?v=tffX3VljTtI https://www.youtube.com/watch?v=don-oWbjq3E And one cool Splinter Cell thing https://www.youtube.com/watch?v=TMhgNi7X53c drop by peri discuck sometime
>>61607 I watched all of those except the Harvey Smith one. I never found Warren to be especially interesting to listen to, mostly because I he sounds like he's selling the game rather than talking about what they went through. Going off my memory of him talking and just what the SS1 devs said, it seems like he really mostly talked to the business people and also kept an eye on what features made it into the game. It seemed like 90% of the ideas came from the devs. Hell the SS1 devs even slipped in features without him knowing, like how the stars render instead of being a texture, which they put in after the game was locked off and nothing but fixes were suppose to go in. I know he had the original premise for Deus Ex in mind for years, but what else beyond that?
>>61637 yeah, pancotti was definitely the more interesting one to listen to there. The Harvey Smith one talks about Invisible War so that was an interesting insight into how that game came to be I wastched that one ages ago so I don't remember everything about it Basically it would be best to find little nuggets of insight and information in these things and write them down in a list or as timestamps. I like stealing scrapped ideas from devs too, if they are mentioned if you want SS dev content there's this series too http://gambit.mit.edu/updates/audio/looking_glass_studios_podcast/
>>61641 Nice. I actually had 5 of those downloaded already, but somehow missed the other 5. Looks like they update it still when they can, since some where posted in 2011 and others 2017. I'll have to remember to checkup on that. I think the only cut feature from SS1 was a terminal puzzle. Warren Spector didn't approve it because it would've been too jarring to go between 3d and text. Pretty much every single idea they had made it into the game surprisingly. I don't know how true that is for 2.
>>61637 >I watched all of those except the Harvey Smith one. I never found Warren to be especially interesting to listen to, mostly because I he sounds like he's selling the game rather than talking about what they went through. Going off my memory of him talking and just what the SS1 devs said, it seems like he really mostly talked to the business people and also kept an eye on what features made it into the game. It seemed like 90% of the ideas came from the devs. Hell the SS1 devs even slipped in features without him knowing, like how the stars render instead of being a texture, which they put in after the game was locked off and nothing but fixes were suppose to go in. His incompetence when it came to actually making the game is the reason the Deus Ex EMP cheat is named after him.
>>61058 Be the games you want to see
(142.31 KB 1280x1024 008.jpg)
(145.61 KB 1280x1024 007.jpg)
(146.20 KB 1280x1024 023.jpg)
>>61901 It's not the games they made. It's the work environment. Going full NEET alone on a game is one thing. Being in a company that is essentially a frat house full of people who eat, breath and sleep the game they're working on is another. This type of thing only exists today for startups since all the angel investors are in silicon valley and only care about funding the next Twitter.
>>61915 ok, be the comfy you want to see then
>>61915 >This type of thing only exists today for startups since all the angel investors are in silicon valley and only care about funding the next Twitter Did id Software make Doom with someone's money or by getting together on their fucking garage and coding away? What you fucking forget is that you don't need a huge ass budget to make a fucking game. Just Like Make Game. There are tons of fucking great engines out there that are free to use, the only thing stopping you is yourself
>>61988 >Did id Software make Doom with someone's money or by getting together on their fucking garage and coding away? Not Doom but they actually did their early games by 'borrowing' high end hardware from their then employer and arguably doing a fair bit of other stuff that bordered on taking money for a day job while working on their game on the side. I mean that's very different from taking investment directly for a game and there was still plenty of coding talent around but they hardly just set up and did it.
>>61956 Studios need money. AAA publishers are nothing like they were back in the day and would destroy a studio like that (as seen with Arkane). Indie publishers only support their friends and are generally corrupt as hell (as seen by GG). Kickstarters that absolutely explode are still barely enough to fund whole games unless they're in the top 10 funded projects like Prison Architect and Scam Citizen. Hell, Warren Spector had to go to fucking Tencent of all people to get System Shock 3 funded after his Underworld Ascendant kickstarter project imploded. >>61988 Doom wasn't Id's first game. According to IMDB it had a budget of ~$1 million, which is about what System Shock 1 had, however there isn't really much evidence to support or deny this claim. Maybe they did that on their earlier games, but you could make a tiny side scroller and get away with it back then. Now 99% of indies are side scrollers and are way larger and more detailed than Commander Keen was. For something else you have to be a full NEET. Hell even Notch was practically a NEET since I doubt he would have survived any of his game projects without Swedish gibs. >>62016 I doubt you could get away with that shit today. Corporations monitor their hardware like crazy now. Gamedev these days isn't really about hardware either. It's about time and energy.
>>62034 >I doubt you could get away with that shit today. Corporations monitor their hardware like crazy now. Gamedev these days isn't really about hardware either. It's about time and energy. It was more a small company, but yeh. There was apparently almost a lawsuit about it after the fact too. >Doom wasn't Id's first game. According to IMDB it had a budget of ~$1 million, which is about what System Shock 1 had, however there isn't really much evidence to support or deny this claim. Maybe they did that on their earlier games, but you could make a tiny side scroller and get away with it back then. PDFrelated covers most of the interesting history. I mean I assume it plays up a lot of the interpersonal drama but ignore that.
>>62054 I'll check it out. I've been looking a lot at how these small studios of 5 dudes back in the day operated. It seems like they all had at least some sort of budget. The bottom line always did their big game fulltime and it was during a time when they could make enough money off such a small project to stay on it fulltime. Thinking about just quitting my job and going all in myself for a while, so I figured it'd be smart to at least do some research. /agdg/ part time is fun and all but I just don't see it as a practical way to make anything aside from tech demos and clones of DOS/SNES games.
>>62080 >I'll check it out. I've been looking a lot at how these small studios of 5 dudes back in the day operated. It seems like they all had at least some sort of budget. The bottom line always did their big game fulltime and it was during a time when they could make enough money off such a small project to stay on it fulltime. Thinking about just quitting my job and going all in myself for a while, so I figured it'd be smart to at least do some research. /agdg/ part time is fun and all but I just don't see it as a practical way to make anything aside from tech demos and clones of DOS/SNES games. As I see it there are 3 things about id that you wouldn't have today if you just tried to set up a few dudes in a room. Firstly they were pushing tech hard as well as making actually good games. That means they were getting huge publicity/free marketing just from being the first in terms of tech and had almost no competitors (at least not outside of consoles, an entirely different market) so standing out was easy. How would you do that in the current wave of indieshit? Mostly you're down to blind luck some e-celeb or journalist picks you out. Secondly the internet culture at the time allowed for shareware to be a viable model, and valued quality and innovation because it was mostly other autists and not normalfags. Business culture meant people could also play the game at work and pass it around like that. Thirdly, and this is the most underrated and yet the most important one: they all ended up working shoulder to shoulder every day at an old-style tech company before they went into making vidya together. Each had individually worked on some stuff before but it was them all being in the same room day in day out that made it and that's mostly down to stumbling into tech companies after making a name for themselves in the early anarchy of home computers. Nowadays you can't really, as a 16 year old, just sell software by post so autists who are amazingly skilled but can't play the university/college and office politics games will likely not get a chance to even get in the door and even if they do get in they'll be in a soulless office job full of diversity hires and retards with compsci degrees not next to other autistic but creative programmers who actually know their shit. Carmack was key to the success of id but nowadays a Carmack probably wouldn't be looked at twice because tech companies are run by business graduates not techfags and even if he did get the job he'd not have spent most of his day interacting with Romero and whoever but stuck in a shitty position next to someone who's pronouns are Xe/Xim/Xir and whose job it is to change all the occurances of blacklist and whitelist to blocklist and denylist and even if he managed to avoid that cancer it'd still just be repetitive back end tasks. In the past tech companies (vidya included) were like outgrowths of the original university hacker communities where those more experienced in one area would feed back into teach others and vice versa and also fags would meet and have productive downtime maybe spinning off naturally into their own company. That's missing now and in general in today's society good luck having a casual male only group of workers try and set up anything by themselves. This is really key for a small company like id were because when you have so few employees they need to be very competent and to work together perfectly and that can't really grow out of picking 5 random faggots you know online and instantly throwing them into a major project together. You can certainly try and aim for that outcome but you first need to try and replicate the conditions that turned them into an effective studio. It's hard to get across entirely what I mean by this third element but when you look at what really killed id in the end it was the cohesion dying when they failed to maintain this atmosphere. Fags will tell you you can't start a 5-10 man studio nowadays because assets are more expensive and games need bigger world and more code and that's certainly true to some degree but really it's because you'd have to either know 5 people really well who also had all the skills needed already or you'd have to find some way to get everyone to work together for a bit first while they got up to speed and you can't just make Commander Keen nowadays and expect it to stand out in the waves of other indieshit so what do you do to get the ball rolling?
(1.88 MB 1106x486 code.webm)
>>62114 That's a lot like Looking Glass, except instead of them being from the same office they were MIT graduates who all knew eachother. They also pushed the hardware to new heights a lot like Id did, and I think Id actually took inspiration from Looking Glasses Ultima Underworld game if I remember correctly. I feel like this is a common theme with a lot of these early game studios. I think even some indie games started out of universities. The Magicka devs and the Pnumbra/Amnesia devs did at least. Rare also had that insanely competitive culture but I don't know their origin story. These days all the competitive people go work for silicon valley companies. SpaceX, Facebook, and Amazon all specifically hire college grads to burn out and churn out. Or they will join a startup that has angel investor funding. >Fags will tell you you can't start a 5-10 man studio nowadays because assets are more expensive and games need bigger world and more code and that's certainly true to some degree but really it's because you'd have to either know 5 people really well who also had all the skills needed already or you'd have to find some way to get everyone to work together for a bit first while they got up to speed I think it's possible to find them. Maybe at game jams. The problem is you can't convince anyone to join AND STAY you unless you have money. There are free collaborative projects but they usually fizzle out, because it becomes too hard for everyone to juggle their job and their project at the same time. Imagine pulling 80 hour work weeks on your game engine and still having a day job. It's impossible. Remember the horror collaboration on here? That died. I think there is a collaboration for a cyberpunk FPS between the diablo guy, kowloon sim dude, and the guy who made a Deus Ex game way back in the day. Maybe they'll pull through. I know the Risk of Rain devs managed. Point is it's rare as fuck and you gotta be all in. Even then I still don't know how many of these people are doing 80 hour work weeks like LGS had several employees do. >and you can't just make Commander Keen nowadays and expect it to stand out in the waves of other indieshit This 100%. The fact people think something that worked in the early 90s would work today is why I always take info from this board, and even these threads, with a massive grain of salt. There's tons of faulty logic like this. >so what do you do to get the ball rolling? Honestly, most indieshit has no substance at all. It might look nice. It might have 1-2 good mechanics. However, the whole campaign or meat of the game is soulless, is missing something or you've seen it already. Within the first 30 minutes of the game it feels like you've seen all the game has to offer. It's like this for almost every indie game. I hate to admit it but even Keyreal had this problem in spades all over Speebot and Phantom Path. Dusk exploded because it took the old games but also had so much substance to it but also industry connections most likely (I'm going off what I heard, haven't played it).
>>62150 >That's a lot like Looking Glass, except instead of them being from the same office they were MIT graduates who all knew eachother. They also pushed the hardware to new heights a lot like Id did, and I think Id actually took inspiration from Looking Glasses Ultima Underworld game if I remember correctly. I feel like this is a common theme with a lot of these early game studios. I think even some indie games started out of universities. The Magicka devs and the Pnumbra/Amnesia devs did at least. Rare also had that insanely competitive culture but I don't know their origin story. These days all the competitive people go work for silicon valley companies. SpaceX, Facebook, and Amazon all specifically hire college grads to burn out and churn out. Or they will join a startup that has angel investor funding. The cultural elements in general are very important. A lot of the reason bong developers were decent during the 90s and early 2000s, for example, was a result of shit like the ZX Spectrum being semi-accessible to poorfags but requiring fags to get a basic ability to code (even just typing shit in is a start) to actually play vidya and resulting in clubs for poorfag kids to share shit around. >I think it's possible to find them. Maybe at game jams. The problem is you can't convince anyone to join AND STAY you unless you have money. There are free collaborative projects but they usually fizzle out, because it becomes too hard for everyone to juggle their job and their project at the same time. Imagine pulling 80 hour work weeks on your game engine and still having a day job. It's impossible. Remember the horror collaboration on here? That died. I think there is a collaboration for a cyberpunk FPS between the diablo guy, kowloon sim dude, and the guy who made a Deus Ex game way back in the day. Maybe they'll pull through. I know the Risk of Rain devs managed. Point is it's rare as fuck and you gotta be all in. Even then I still don't know how many of these people are doing 80 hour work weeks like LGS had several employees do. Sadly there's also the issue that nowadays a lot of people treat mods and community projects as portfolio stuffers and drop them the moment they've turned it into a job offer. I'm not saying it's impossible to get something like the old style of company going mind just that you can't expect it to work like the 80s or the 90s: it's like how fags will point to smart men who got rich from nothing in the past as if that's still practical now when so many of them did something like work in a factory first and learnt their industry in and out that way or started a mail order businesses despite being only one man and the former is impossible now it's all outsourced to shitholes while the latter is strangled in regulations and fixed costs because the same fags who got rich doing it have pushed government legislation to make it impossible for competition to harm them. You can still succeed but you need to acknowledge that the world has changed and plan accordingly.
(18.23 KB 150x150 1470835590147.jpg)
>>62176 >You can still succeed but you need to acknowledge that the world has changed and plan accordingly. This is the #1 thing 8chan/bunkers/alt-chans never got about games. It's changed. Discord, Kickstarter, etc are all tools that everyone else is using to get ahead. Pissing off the SJWs and burning bridges with 95% of the industry is not a recipe for success outside from that 1 edge case that happened during the height of GG. The #2 thing is that games require time. Doing it by yourself is never going to be up to par with games made by a studio people crunching like hell (which the "low effort" indie pixelshit is). Doing it in your free time is dividing what little manpower you already have into basically nothing. Oh yeah, you can make "a game" by yourself part time, but it's not the game you or anyone else wants to play.
>>61073 >The main take away of this I think should be why older games were so great in the first place, many of the developers weren't just obsessed with the idea of creating a game, gaming wasn't their only hobby, they didn't go to college to learn how to make video games. They had life experience in a variety of areas, they had other hobbies and skills that lended to their creativity in creating games. Inspiration often comes from the least likely of places. I was actually listening to another dev interview from LGS on SS2 and they touched on this. Half the people at looking glass had PHDs in virology, a masters in computer science a liberal arts degree or something completely unrelated to games. The SS2 dev was basically saying this added a lot of life to their games, because you had people bringing their outside experiences into the game world. Rather than people just rehashing what they already know. https://invidio.us/iWFylN8bRdQ?t=3857
>>62238 I'm not sure I'd be quite so defeatist about it but you certainly can't just puddle about making a sort of decent game for 3 years then release it into the wild without any sort of consideration for getting fags to see it. You'll just get buried in the other shit. For example if you make another 2D platformer you're arguably setting yourself up for failure just by picking a flooded niche. >Pissing off the SJWs and burning bridges with 95% of the industry is not a recipe for success outside from that 1 edge case that happened during the height of GG. You don't have to bend over backwards for them but it's probably better to just say nothing than it is to explicitly engage in that whole debate as a studio, if you somehow end up successful then after your third or fourth game feel free but it's not a great plan from the start. That said if you were particularly competent at handling drama you could use it for free marketing, no such thing as bad publicity and all that, but you'd have to prepare for your name to have a permanent black deny mark next to it for anything but your own company in future.
>>62597 >I'm not sure I'd be quite so defeatist about it but you certainly can't just puddle about making a sort of decent game for 3 years then release it into the wild without any sort of consideration for getting fags to see it. You'll just get buried in the other shit. For example if you make another 2D platformer you're arguably setting yourself up for failure just by picking a flooded niche. The more I read into it the more blackpilled I become. In terms of making a game with any substance I genuinely believe you either have to do it full time (and maybe contract some help on fiverr) or you have to do it with a team. Otherwise you're spreading yourself too thin and this is including scaling it back. A lot of people believe the "Just make a shitty ass small game" meme, which is Extra Credits-tier advice. What's the point of making a side scroller or a 2d game that brings nothing to the table? Whats the point of doing something everyone else is doing when everyone else has a team of autists working on it? How far can you scale it back before it's not a game even you would want to play? >You don't have to bend over backwards for them but it's probably better to just say nothing than it is to explicitly engage in that whole debate as a studio, if you somehow end up successful then after your third or fourth game feel free but it's not a great plan from the start. That said if you were particularly competent at handling drama you could use it for free marketing, no such thing as bad publicity and all that, but you'd have to prepare for your name to have a permanent black deny mark next to it for anything but your own company in future. I didn't say bend over for them or even censor your game for them. Just understand who has the power, what position you are in, the current circumstances and if they can hurt your chances of actually surviving. Right now it seems like, unless you are already established, the only avenues you can take to get around them is Steam and crowd funding. Both of which are being heavily censored. If you burn bridges with the journalists your only hope for getting any sort of funding is getting cuckchanners to fund it like YandreDev, being in a country that doesn't give a shit like Destructive Creations or just having the money on hand like Running with Scissors.
>>62628 >IA lot of people believe the "Just make a shitty ass small game" meme, which is Extra Credits-tier advice. What's the point of making a side scroller or a 2d game that brings nothing to the table? Whats the point of doing something everyone else is doing when everyone else has a team of autists working on it? How far can you scale it back before it's not a game even you would want to play? There's still value in the experience you'd get but if you spend months or years at it and don't get a product you might make money off of it won't leave you any closer to forming an actual company and getting serious shit done. >or just having the money on hand like Running with Scissors. I was under the impression most of them actually have other day jobs as well.
>>62679 >There's still value in the experience you'd get but if you spend months or years at it and don't get a product you might make money off of it won't leave you any closer to forming an actual company and getting serious shit done. This is phrased a little weird. I guess what you're saying is if you don't get a product going it wont make you any closer to a company? Runescape was made by one dude who eventually grew it to be a full fledged game. Notch also got Mojang off the ground after Minecraft started picking up steam. I don't know about the former, but the later was at least making games full time and possibly supplementing his income with gibs. >I was under the impression most of them actually have other day jobs as well. Yeah. The money is coming from somewhere. Postal 3 had a publisher, but 2 and 4 didn't.
>>62628 >A lot of people believe the "Just make a shitty ass small game" meme, which is Extra Credits-tier advice. It's actually good advice, you memey fuck. Small, simple, free games are a good way to train your skills, learn how to manage larger and larger projects, and get feedback,all while not putting much at stake so even a complete failure won't be much of a setback: that's how things worked in the Flash era, and that was a much better time. >What's the point of making a side scroller or a 2d game that brings nothing to the table? Learning the basics. >Whats the point of doing something everyone else is doing when everyone else has a team of autists working on it? Learning how to prioritize, how to focus on the truly important parts, and how to cut what's not necessary. >How far can you scale it back before it's not a game even you would want to play? Have you ever heard of Flappy Bird?
>>62724 >It's actually good advice, you memey fuck. Small, simple, free games are a good way to train your skills, learn how to manage larger and larger projects, and get feedback,all while not putting much at stake so even a complete failure won't be much of a setback: that's how things worked in the Flash era, and that was a much better time. Nah, it's shit without context. It makes sense doing in a 24 hour-week long game jam (even if you're brand new). If you're spending years making a game that's tiny then you're wasting your time. A lot of people take that advice and dedicate their life to making insignificant games no one wants to play. >Learning the basics. Except faggots don't use it for the basics. Almost every single shitty indie game is a metroidvania/snes retro gimmick these days. No one makes the distinction between applying the rule to a shitty practice game made in a few days and actually trying to make their dream game. And if you plan on making 3d games hardly any of the difficult skills transfer. Maybe the insignificant shit of how triggers and other things work (that you will learn going 3d anyways), but anything meaningful that requires a huge time investment to learn will not transfer. 3d art, working with 3d physics, etc. >Learning how to prioritize, how to focus on the truly important parts, and how to cut what's not necessary. Sounds like something to do during a game jam. NOT a year long project, like the vast majority of people are doing. >Have you ever heard of Flappy Bird? A 1 in a million mobile ios game that wouldn't get anywhere today? Same situation with the .io games. You're not going to convince me with a hail mary.
>>62238 >but it's not the game you want to play Yes, you need to make a small game focused on a single mechanic and explore it to it's ultimate conclusion. It should be fun enough for 2000+ more people to buy it on steam, and if you luck out, it will become a new speed run fever and you will get rich. Otherwise you use it as a stepping ladder for a bigger project. Just like you could work to save up enough to feed yourself while you work on the game you want to play, you can also make a smallish game instead.
(47.95 KB 616x353 Neon Struct.jpg)
>>62806 >Yes, you need to make a small game focused on a single mechanic and explore it to it's ultimate conclusion. I'm so done believing this nonsense. I've played so many shitty indie games that feel like you've seen everything in the first 10 minutes. Thief is a game that focused on a single mechanic and explored it to it's ultimate conclusion. However, Thief had substance so people actually wanted to play it. Same cannot be said for the indie clones. Pic related is indie thief. It does everything you said. It's shit. Maybe it had some success, but I found it completely unplayable with how shallow it was. It probably took a fuckton of work to. Doesn't change the fact that it's shallow, empty and boring as fuck. I'm not interested in games with nothing to them. >It should be fun enough for 2000+ more people to buy it on steam, and if you luck out, it will become a new speed run fever and you will get rich. Kek, 2000+? Did you forget how oversaturated steam is? Maybe you'll sell that many if your game goes on sale for $0.50 for a few years. The blackest pill out of that is even if you sell that many you still barely made any money and definitely not enough to live off of. If your games $10 you only made $20,000 not factoring in Valve's cut and the taxes. >Otherwise you use it as a stepping ladder for a bigger project. Just like you could work to save up enough to feed yourself while you work on the game you want to play, you can also make a smallish game instead. The latter does nothing unless you basically win the steam lottery or are really cranking them out like at game jam speeds. Only option left is the former.
SURE IS A LOT OF POSTING AND NOT A LOT OF DEV NOISES
>>62832 I at least can say I got value from this because I learned how older game studios with no budget got off the ground. Better to know where I'm going than grind away at a project that isn't going anywhere without realizing it.
>>62849 Honestly, I think you'll know when you are able to make a game from start to finish. It's like someone asking what engine to use. If someone asks that, then need to just use anything so they can build experience and form their own perspective
>>62786 >Nah, it's shit without context. No context can change lack of skill: if it takes someone years to finish a tiny beginner-tier project, they aren't cut for game dev and no advice will change that. >You're not going to convince me with a hail mary. Progress IS a long series of hail maries, each pushing the boundaries a bit further. Also posting retarded maymays isn't going to help your point.
>>62823 > If your games $10 you only made $20,000 That's two years of salary in most parts of the world. More than enough to bankroll another game. Maybe your standards of living are too expensive?
>>62681 >This is phrased a little weird. I guess what you're saying is if you don't get a product going it wont make you any closer to a company? Yes. It'll help you grow as a developer but it won't help you get the resources together to put a serious game out there or to form a full time vidya company with even a handful of fellow autists. That means you should aim to make something that will take maybe a few months, half a year at most, for you to finish not get embroiled in an autistically complicated project built from the ground up that'll suck years away from you. Treat it purely as a learning exercise and be prepared to drop it unfinished if it's taking too long or a better opportunity shows up I'd classify even a shitty entry level job in non-vidya coding as a better opportunity, it'll pay badly and be mostly grunt work but there's value in seeing how coding in a team is vs doing it alone. Again not trying to be a defeatist faggot but you need to be realistic if you want to make a living out of vidya and not just expect that things will magically fall into place. Treat any minor one man project as if it'll sell no copies at all and make sure it'd still have been worth your time to do it if that happens (experience etc): that means you also need a plan for how you'll move into more 'serious' projects if you've spent 3-4 years making games like that and none of them have even broken even, even if that plan is to give up on vidya as ever being your day job and just doing it for fun on the side instead. >>62906 If you assume $10 asking price and 2,000 sales minus Valve's cut (30%) you're at $14,000 not accounting for whatever local taxes there are on top of that and any other administration costs you might hit. How far that'll take you depends entirely on your living situation. If you're a 21 year old man with no attachments and no other costs living in rural Poland possibly in a house your family already owns great, you can stretch that out for 2-3 years needed and use EU-funded welfare because the Polish government throws gibs at basically anyone. If you're a 29 year old stuck in Bongland with rent or a mortgage to worry about or worse yet you're old enough to have to deal with family costs that wouldn't last you 6 months. This does not account for funding for the time it took you to make a game that's able to grab 2,000 sales of course or any spending you might have to do on contracting in faggots for music and art if you're a one man show. And even if you get 2,000 sales, which according to a quick google is about 150% of average sales for an indie on steam right now, you're unlikely to get them all in one lump so it'll be cash trickling in over time. Now in some regards that can be beneficial for tax reasons but on the other hand if some unexpected cost shows up you're fucked. So again something you can definitely shoot for if you're young and have nothing to lose but not something you can risk if you have any obligations.
>>62823 Neon Struct isn't bad. Slayer Shock is. It had L/S stealth which is extremely rare and appreciated, even if it is simplistic. What other games have used that since 2015? Only ones I'm aware of are Gloomwood, Filcher and two agdg games, including us.
(297.17 KB 500x314 programming_is_hard.gif)
What's agdg's opinion on this? >Run function to get rarely needed data only when it's needed during runtime, using up extra cpu time occasionally or >Declare the data as a variable right from the start of the game, keeping some memory used up for something rarely referenced
>>63939 In purely hypothetical terms the former since you could just do it as needed (when you load the level anyway, when you know it's going to be needed up ahead anyway when the player is in a part of the level that's known to be less CPU intensive). I'd guess you could probably just automate that too. Then you could let go of it when it's no longer needed and won't be for a long time if it's that much of a concern. A fair bit of handling real time programs is not so much controlling how much memory you use as long as you're not retarded about it as it is when you allocate and deallocate it so a bunch of work isn't done all at once slowing things down. That's why automatic garbage collection is not acceptable for such purposes. Not something I've ever done serious enough vidya work to encounter though, and it seems like it's probably fairly academic on modern hardware.
>>63939 Depends on the circumstances. How much RAM does the data require? How much disk space? How expensive is it to load it at runtime? Can you load it async or would that cause issues? Can you load the data during a loading screen? Can you calculate the data from a smaller seed?
>>63939 Memory is cheap. Ideally do a lazy load, or else async load and grab a few things during idle cpu time, and take it all if it needs it sooner. Frontloading stuff all at load or laumch leads to bad feels for modding
been struggling with clean overall game state logic lately for a MP arena type game. been trying to use the finite state machine pattern but it has a lot of pitfalls and if just one variable outside the state gets set then not reset it fucks everything up. Probably worse without it though.
Stupid question because I'm bad at pathfinding If you have tile-based game, and a navmesh at arbitrary points, as long as you guarantee the navmesh is correct as tiles change, you can just A* between two nav points, right?
>>64323 If performance isn't an issue, that should work (assuming tiles don't change mid-pathfinding, but that's retarded so don't do it). If performance is an issue, read factorio.com/blog/post/fff-317 for inspiration.
(4.64 MB 1280x720 2020-07-18_05-41-38.mp4)
(936.01 KB 1280x720 2020-07-18_18-57-00.mp4)
We have huge fans
>>64663 >Read latest FFF >Kovarex took a break from 1.0 devving and decided to play WoW for 3 weeks months, and then got burnt out and depressed that all his time raising a tank to 60 didn't matter for raids because tanks are easily acquired by then >Almost considered selling or closing the studio due to the depression WEW Anyways, I'm glad they're back on track
(267.72 KB 432x454 JCDenton2.png)
>>65115 Is this supposed to remind me of deus ex?
(15.27 KB 561x184 ClipboardImage.png)
>>65190 maybe
(1.07 MB 400x400 kurwa motur kura.webm)
>>65412 >polish KURWA
>>65438 I see your point, but dev time doesn't work like that: new tech means writing code nowadays is much faster than it used to be (mostly because you can reuse fantastic tools made by others instead of having to implement everything yourself), and artistic choices can have a massive impact on the time you spend working on assets.
>>65427 tfw no one noticed the voice clip at the end https://twitter.com/i/status/1272028382566498305
>drown in uni work after 8chan goes under >almost forget about it with all the time passing >try to find back home >everything is now fractured across dozens of boards and bunkers >posting links to other boards isn't even taboo anymore what is going on anons, where is everyone
>>66351 I didn't really spend much time on 8kun so I might make mistakes. After 8chan went down I pretty much only go to /agdg/ or the odd thread that is talking about games that I actually find interesting. Essentially: >8chan comes back as 8kun >Not everyone comes back and there are altchans >Mark makes an open letter to Ron on the board >Ron removes Mark as the admin for /v/ >Ron also bans loli because his lawyer says it might be a good idea >Ron also did some other shit to step on people's toes >There is a mass exodus from 8kun to the various altchans >The old altchans get taken down and new ones come and go No one's going to bat an eye if you mention another board. For all they know they might end up there in a few weeks. I think Zchan just went down a few days ago. What's hilarious in all of this is you still see the contempt the altchans have for eachother. Similar to how 8chan/vg/, Julay and others did. When all these boards are basically the same now just with a fraction of the traffic, because everyone's hiding in their own little bunkers. None of these places even feel like video game boards anymore. The only threads that get any traffic are ones bitching about the state of the industry, drama, etc. All the generals from 8chan (Stalker, Deus Ex, SS13, etc) are gone.
>>66624 Isn't fracturing and splitting userbases how alphabet agencies operate? Fuck everything was much simpler even 2 years ago. The worst part is, we'll probably miss our existence here in a few years
>>66636 The shitty part is that if you say "there needs to be some consolidation" then someone will go "REEEE, no, the more the better" and call you a D&C shill. But there is clearly a middleground where having redundant infrastructure is good but you also are consolidated such that there's really only one of each board. The current state is the absolute worst. Nevermind the constant deplatformings; it was inevitable that some sites would hold up better against attacks than others. No, the real issue is that the biggest boards are all split among 5+ websites, sometimes have 2-5 "major" bunkers with some level of activity. And the small boards? They're just gone. This is a natural consequence of everyone looking out for themselves and those who like the same stuff and not looking out for other anons.
(365.11 KB 400x328 another_all_nighter.gif)
>trying to make a level editor using the same libraries used for making the game >have a functional grid that I can add objects to by clicking >no idea how to make a menu for selecting objects to add in because it would offset the position of everything on the grid >no idea if there's even any functions to handle clickable objects without iterating over sprite lists to see if mouse was at position of sprite at x,y when mouse_button1 was clicked >>66636 >>66687 There's also the issue that too many people were praising the webring as a safety net, like as long as the webring existed there would always be a place to go, when really the webring was a hit list for whoever went around getting every site deplatformed with 1 complaint to the host/cloudflare/dns/registrar/etc. We're still seeing attacks to alternative sites which just goes to show nothing was really solved.
(27.93 KB 512x512 1588740952453.jpg)
>>66636 Those things existed even back then with 8/vg/, because people were getting fed up with every single thread being spammed with bait and shitposts. The only difference now is the shitposters got bored and went somewhere else or stayed on 8kun. Also more people are on the alternatives after Mark started mass-deleting posts and banning people like crazy on 8kun for some reason. >>66687 I don't even think consolidating would save the 8chan community. It's so obvious most of the old community is gone even when you factor in all the bunkers. Everyone left is so pessimistic and catty I don't see it lasting. All that's left is /agdg/ and a bunch of cancer. At least /agdg/ has a chance of surviving this dumpster fire.
(15.30 KB 227x255 1459757927943.jpg)
>>66698 It's not just about consolidation. It's the idea that one basket per egg isn't a community, but the more eggs you have in a single basket, the better your basket needs to be. There is massive duplication of effort that goes on. 10 sites means 10 admins paying for hosting, 10 sets of global volunteers to prevent CP raids and spam, 10 people working to add features to the site in their spare time. If you whittle that down but retain the community and the staff then you could feasibly run 3 sites of better quality and with more crossover discussion happening. Slow sites that are prone to being spammed and DDoS'd are bad for people spreading out and discovering new boards. >>66699 I think it's possible to get many of them back. The ones who went back to cuckchan don't seem thrilled about it. The hard part would be advertising without bringing the plebs. The ones who went elsewhere might have quit imageboards entirely for a bit and are focusing on life. They will trickle in eventually. But right now, half the webring is trying to avoid the "8chan" name and the other half is embracing it. I think giving up the name is giving up on all the users from 2016, since then we become just another 4chan alternate. The webring deserves better than to be another 7chan. But I don't think we can get to that point until anons, or at least the webring admins, can put aside their differences and focus on the big picture. At the very least, it would probably be best for general threads like these to consolidate down to 1-2 bunkers and singular resource websites. Having a board + a thread on 5-10 different sites is spreading resources and anons thin. Especially now that sites like 9chan are being put out to pasture it would be wise to consolidate to the remaining active threads.
>>66721 >I think it's possible to get many of them back. The ones who went back to cuckchan don't seem thrilled about it. The hard part would be advertising without bringing the plebs. Cuckchan's /vr/ board is actually pretty decent. The only downsides is you don't see the Dwarf Fortress, Stalker, Space Station 13 threads. It has it's fair share of faggotry, like unironic console wars and the occasional bait posts, but it's nowhere near as bad as 8kun or late stage 8chan was. I actually bumped into one of the anton & coolpecker guys there. >But right now, half the webring is trying to avoid the "8chan" name and the other half is embracing it. I think giving up the name is giving up on all the users from 2016, since then we become just another 4chan alternate. The webring deserves better than to be another 7chan. Honestly this kinda faggotry is why 8chan is doomed. Too many people sperging out about shit that doesn't even matter. >At the very least, it would probably be best for general threads like these to consolidate down to 1-2 bunkers and singular resource websites. Having a board + a thread on 5-10 different sites is spreading resources and anons thin. I think in regards to /agdg/ the agreement was to post progress across all the bunker threads. So that every thread would be alive. I think the Peripeteia devs are the only ones doing it.
>>66636 >Isn't fracturing and splitting userbases how alphabet agencies operate? You don't need glowies to explain why the aggressively retarded userbase 8chan had in its last days would split up in a hundred autistic shards.
>>66733 >I think the Peripeteia devs are the only ones doing it. us and Yeerk. If I bother to maintain the wiki on life support, I'm gonna make use of it. also, I just want agdg to survive. I updated the archives, btw. https://mega.nz/folder/QlVxiaaY#swgVF22L3ZDc7Dc8n-rJpA I'm still waiting for a hero to make the thread aggregator.
(30.40 KB 500x332 1453050518260.jpg)
>>66367 >>66624 >>66687 >>66699 >>66721 >mfw we're still here just to suffer I just want the simpler times back, at this point im glad to even just be in this thread >>63939 This highly depends on how much extra cpu time the operation takes. If the game will hitch if you don't cache it, do it, if not, don't. If in doubt, do what takes less code and effort for now. >stop prematurely optimizing is a meme especially when it comes to system design, but for oneshot kinda things it's best to do whats simple and straightforward first, ask questions later. >>64228 Not sure what context you're working in but if its C/C++ look into fibers for state machines, you can do neat shit with them, like writing each state like an infinite while(1) even though it constantly yields.
(3.96 MB 1872x1044 aristocrats.webm)
more progress on AI
>>66985 Haven't been following the thread too much, I take it they are responding to the sound based on the location you shot from? Or are they just meandering about and I''m reading way too much into it.
It's been a while since I last posted I finished almost everything in my game but the level design and some of the code. Then all of a sudden the engine forced an update. This update ruined the engine making simple tasks annoying as fuck to do. The faggot also removed all the older versions online "by accident" and the versions I have backed up is over a year old with most of the newer features I use either bugged as fuck or not there. So I am switching engines and remaking my game but now it is going to be a 3d fps instead of a jrpg clone but the story and everything else I created will be the same. I just have to figure out how to take 2d assets and convert them into 3d ones. It might end up being a shitty doom clone but once it is done it is 100% going to be released for cash. I should take one of my older games and add a small update then sell than for cash first since all I need to do is add one more scene to my last game and it would be good enough to sell for 10 dollars on shit like steam.
>>67000 No, sorry. Only reason I shot at the beginning is because Unity Editor doesn't automatically recapture mouse control even if a game event tells the game to, so you have to click the screen in order to get rid of the cursor. Sorry for the confusion.
>>67004 Ah no problem. Was just curious as to what was going on. >>67001 >forced an update >This update ruined the engine making simple tasks annoying as fuck to do. The faggot also removed all the older versions online "by accident" What the fuck. What engine did this?
(166.57 KB 348x342 1468949129423.png)
>>67001 >remaking my game but now it is going to be a 3d fps instead of a jrpg
(19.15 KB 203x219 billy_is_retarded.png)
>>67001 I guess it's true what they say, if you want anything done right, you gotta build your own game engine.
>>67014 Well, there was a point a click adventure that became a beat 'em up on the Genesis, and another P&C adventure that became a horror game on the PS1 and PC.
>>67014 Mc is an archer and there are better tutorials for fps games than RPGs
I struggle to find enjoyment out of video games, so as I struggle to gain motivation to create one. I no longer have time to play games due to work, and the effort spent to learn game mechanics and "skill" can always be better spent at devoting energy to my work and my education. If I was to create an interactive program, or a GUI. It would not waste time like a game. I want to create something to excel productivity, not hamper it.
>>67098 >If I was to create an interactive program, or a GUI. It would not waste time like a game. I want to create something to excel productivity, not hamper it. Then why bother posting here?
>>67342 GUI based stuff and productivity tools is exactly how you go about making modding tools, which I'd say falls into /agdg/ territory, depending on his inclination
>>67098 >I struggle to find enjoyment out of video games >I no longer have time to play games due to work >the effort spent to learn game mechanics and "skill" can always be better spent at devoting energy to my work You're burning yourself out by working too much, remember that working is a way to pay the bills and not the one and only purpose of your life. >I want to create something to excel productivity, not hamper it. That's because you're currently obsessed with work, so you interpret everything by the way it affects your work: now it's games that you see as a waste of time, then it will be books, eventually you'll start avoiding foods that take longer to eat so you can wageslave for a few more minutes. Unless you're really desperate for money, there's no reason to do this: your boss will appreciate your efforts, sure, but only because you'll be the sucker putting in twice as much work as what you're paid for.
(487.06 KB 464x325 1451094536641.gif)
>>67431 >You're burning yourself out by working too much, remember that working is a way to pay the bills and not the one and only purpose of your life. >That's because you're currently obsessed with work, so you interpret everything by the way it affects your work: now it's games that you see as a waste of time, then it will be books, eventually you'll start avoiding foods that take longer to eat so you can wageslave for a few more minutes. >Unless you're really desperate for money, there's no reason to do this: your boss will appreciate your efforts, sure, but only because you'll be the sucker putting in twice as much work as what you're paid for. This is only partly true. I worked a job that was very startupy. It was mostly college kids in their 20s-early 30s, a hip office, beer on the tap, everyone was smart and there was little bullshitting around. We busted our asses. Pretty much the moment you got into the office to when you went home it was just work with no downtime outside of lunch. However, I never really noticed how grindy it was at all. It felt good like there was a future. I got a lot done on my game in my free time then if I wasn't distracted by shit like happy hour. I actually started enjoying my job more than my hobbies/life which was the scary part. I later ended up working at a stereotypical dilbert office. It had an overwhelming stench of stagnation and rot to the work environment. It was almost entirely braindead pajeets and managers in their 50s. The level of incompetence is absolutely fucking astounding. Every problem has to be solved with a meeting where the problem isn't actually solved and people would intentionally bullshit around to use up the entirety of the meeting time regardless of if it was helpful. The managers clearly got their jobs due to playing the corporate game and not because they know what they're doing. I started working like how I did at the previous place, but what happened was the pajeets started piling all the shit they were too retarded to do themselves onto me. My manager realized I was doing all the hard shit, but the most I would get was meaningless lip service, and then everything would be drowned out with nitpicks which were the things that actually mattered when it came to talking compensation. It was clear that I was getting no favors for working hard, so I started slacking hard and I mean working maybe 1 hour a day and then just procrastinating the rest of the day. Literally no difference noticed by anyone. I didn't save any energy doing this. My work ethic had been compromised. I was going backwards not forwards. I was at a company that valued replaceable cogs over talented people. I didn't play games or work on my own game. I sometimes would go home and immediately pass out for the rest of the day. It felt absolutely soul crushing. Holy fuck what a nightmare that was. I'm in between jobs now, but I'm going to try to get something part time. More time to dev, less time wasted on corporate bullshit. Just hope I don't land in a shitty work environment again (like 90% of them are) TL;DR You can bust your ass and still have energy to keep going. You can slack off and feel like everything is a huge waste of time.
(42.13 KB 1089x263 ClipboardImage.png)
(657.79 KB 800x800 2019-08-01-0450-02.mp4)
I'm having trouble figuring out the best way to handle my object behavior for my puzzle game (I've worked on it on and off for the last 2-3 years, some might remember it, but I don't have any current pictures) Anyways, I have an assembly that's strictly gamestate and update logic. It relies on client code to implement "something" and provide methods for object behaviors that it will then use to simulate the logic. Each Actor and Tile object will have an index, or a reference to some sort of Behavior object. The problem is that I don't know the best way to structure the Behavior object. Not every object exhibits every behavior; for example, most will simply have an OnIdle behavior, like moving or falling. Having it as an inheritable class makes shitting out objects easy for the client code; just subclass it, override what you need and shove it into the engine, done. Having it as a collection of function pointers means I can avoid subclassing and also only define methods that are needed (and avoid empty function calls), but the behavior still needs to be defined somewhere which leads me back to the first implementation. Does anyone have any insight on the way to approach this?
>>67649 Inheritable class seems like the smart choice to me, unless your tools don't optimize it well or run into weird bugs with it. >only define methods that are needed You can do that with the inheritable class too, you define the base method but you don't call it by default, so only objects which need it will call it: for example, your base class might have DoCollisions() but derived objects would have to call DoCollisions() in their OnTick() if they want to have collision detection. I'm not sure how avoiding subclassing would be an advantage, but I also don't know your codebase so...
>>67718 >I'm not sure how avoiding subclassing would be an advantage Some anons in here lean towards composition over inheritance. If a method has no commands and immediately returns, is the compiler smart enough to optimize things away? In any case, I ended up going with an interface with the signatures I needed for my engine assembly, and then a flexible base class in my client code, this seems to be a happy medium.
>>67551 Fully agree on this, having strict discipline, proper sleep hygiene and so on is a net benefit. "Work" or energy/motivation is not a resource that slowly accumulates while you're idle and is then depleted once you do something, it's almost entirely the opposite. Training a muscle is a good analogy to this, sure you can overtrain or tear your bicep but if you do nothing all day you'll be weak as shit. >>67649 In general it goes like this: Subclassing aka virtual inheritance has an (almost) unavoidable overhead: Given a pointer to an abstract class, it is unknown at compile time which specific derived class will be referenced using that pointer. So if you doo "foo->run()", that's a double indirection. It looks up the actual function pointer to the derived class in what is called the vtable. So, pointer deref, offset into the vtable, read function pointer, call real function The only exception to this is "devirtualizing" if the compiler can prove that only derived class X can possibly be used here, but that's very rare. So essentially virtual inheritance is nothing but convenient syntax for a raw struct of function pointers. As for your problem, im not sure how to understand it. You can have pure virtual methods in the abstract class, which must be overriden, or normal virtual methods, which have a default implementation. So your abstract class would just need empty function bodies for all of those methods, then you override the ones you need.
>>67728 So I guess the takeaway here is that it's basically unavoidable, it's a needless optimization and that shut up and just like make game™
>>67728 >Fully agree on this, having strict discipline, proper sleep hygiene and so on is a net benefit. "Work" or energy/motivation is not a resource that slowly accumulates while you're idle and is then depleted once you do something, it's almost entirely the opposite. Training a muscle is a good analogy to this, sure you can overtrain or tear your bicep but if you do nothing all day you'll be weak as shit. It also needs to have meaning. Every motivational speech talks about how you need to believe in what you're doing. At the startup it was optimistic, fun, my peers were intelligent and had a future and everything seemed necessary. At the corporate job it felt dead end, dull, my peers were incompetent and were just waiting to retire and it felt like a complete waste of time. There's no way you can be disciplined if you're only half in the game.
(230.20 KB 542x451 ClipboardImage.png)
(190.47 KB 405x477 ClipboardImage.png)
Some sexy new programmer art. Don't go thinking anything too lewd. ;)
>>67857 So what's the story behind this SCP?
(10.18 KB 259x194 alex-jones.jpg)
>>67870 It's what happens when you masturbate too much It's a game set in a dystopian future where the elites in society get mutations to remain genetically superior to the population, while encouraging the working class modify themselves as well, but the mutations available to them are cheaper and lower quality, turning them into hideous, savage freaks. Some middle/upper class people get "specialized mutations" which makes them less hideous, but they're essentially weaponized/militarized against the population. When I make more progress, I'm gonna need someone else to do the models justice because I'll never ever be a competent 3D modeler.
>makefile is too complicated and retarded, it is prone to errors, slow because it recompiles everything and the 'fix' for it is to compile headers and it makes it faster when recompiling any changes that is not in the header file but introduces bugs if i recompile without recompiling the headers maybe i learn how to use cmake
(3.00 MB 640x532 1592179520845.gif)
>>67857 >>67870 >>67893 Kek. It literally has the coomer eyes.
>>67722 >composition over inheritance Interfaces a la Java are very good for big projects in my experience, but they might be severely overkill for a small game. Or maybe they'll be perfectly fine, I've had the displeasure of trying to mod some absolutely disgusting OOP C# games that really should have used them. >If a method has no commands and immediately returns, is the compiler smart enough to optimize things away? Depends on the tools you're using, I know Unity can struggle with empty Update() calls because of the way those are handled. But if a method has functions that are never called, those won't impact performance at all.
>>68212 Dude should learn to use make in general. >it is prone to errors It just runs the commands you would need to run manually. >slow because it recompiles everything It only needs translation units which are not up-to-date. >the fix is to compile headers You don't compile headers at all. Ever. The solution is "separate compilation", which means breaking your program into logical translation units. >introduces bugs if I recompile without recompiling the headers This sentence doesn't even make sense, so I can only assume he's completely butchering his workflow.
>>68389 i finally fixed it, it's no longer shit >no longer making .gch instead just making object files and linking them together instead >also learnt what the fug everything does >only problem is passing flags to %.o: %.cpp, not sure how to do that god i was retarded
Man this thread is so dead now. Even when you include all the bunkers. It feels like only a few months ago it'd hit the bump limit every week. Where'd everyone go? Cuckchan's /agdg/? Some other obscure game dev community?
>>69401 I've been busy with work and never having time to dev as much
>>69401 from what I've seen whenever yet another largest traffic bunker disappears, people get more blackpilled and decide to stay away from imageboards for a while (or forever in some cases, started with doctard after Christchurch). Currently the liveliest place is the one some people refuse to go to (protip: it's not an imageboard). I haven't seen any poster I'd know from 8agdg migrate to 4cuck agdg, unless they crossposted before. Maybe some consider the agdg dream to be dead, the new thread on 8kunt was made by a newfag and there's a shitty asset flip getting more attention than said thread. We were replaced there pretty quickly. Or maybe some just never bothered with the wiki and don't even know there even are bunkers. I still wish this got somewhere: >>66769 >I'm still waiting for a hero to make the thread aggregator. Having to check 7 different places at the same time is a bit of a chore.
>>69429 >(protip: it's not an imageboard). Gunna take a shot in the dark and say Discord. If so I wouldn't be suprised. That's a pretty good way to filter a lot of the faggots, although people are wary of discord for a reason. Aside from the privacy issues it tends to be a hugbox.
>>69566 >tends to be a hugbox Only in a sense that you NEED to prove you have ANY sort of agdg project to be let in. Otherwise everyone's welcome, as long as you're ready to get bullied for being a nigger, spic or a degenerate. We only had a few bans. 1. A tranny loving faggot 4cuck refugee- for the sole virtue of being a nodev shitter, not for his views. 2. A literal jew who couldn't control his innate urge to subvert 3. Some spics who couldn't take bantz, including el atrocidad Chris >>8880
(40.92 KB 118x117 what the fuck.PNG)
>>68235 Get out
>>69595 I was thinking of joining but I don't see how it couldn't become like every other community ever. A strength of the /agdg/ threads has always been the fact that it was semi-anonymous. Might as well will just join one that has some connection to the industry if you're gunna be name fagging. >>69620 >He thinks I want to see his shitty threads Don't get so full of yourself. I'm only here for /agdg/. 8chan has been unredeemable trash for years thanks to spergs like you. Even the alt-chans aren't safe from you faggots thanks to bunkers now being the norm.
(13.71 KB 307x148 152347892489.jpg)
>>69401 >Where'd everyone go? I'm still lurking.
>>69640 >A strength of the /agdg/ threads has always been the fact that it was semi-anonymous in a way, sure, when it's criticism you don't have a grudge to hold, when taking advice it's the same. BUT my counterpoint would be: 1. You already have IDs on most non-faggy imageboards so once you get into a discussion, you know who you're talking to 2. Once you post a project, you are not really anonymous anymore. You at least have a project that people will associate you with so you have >>62150 >the diablo guy, kowloon sim dude, and the guy who made a Deus Ex >>9209 >Sigmadev >>8911 >speedbot dev and all the others, listed on the wiki, for example. Even if you think of them as "that one anon", they likely have some sort of youtube for posting progress (Doctard), or a website (Fables of Laetus), or twatter (Jackalsoft), or you'll play their demo on itch (ika) or steam (keyreal) and you'll end up knowing their nicknames (sometimes only used for dev) at some point. Some will be namefags and attention whores (remember rustfag?), sure. But this is inevitable on here, and it kinda helps keep this place the highest in quality of most threads. >Might as well will just join one that has some connection to the industry if you're gunna be name fagging. You have no idea what a shitshow most (((industry))) communities are. I choose to stick to the one with minimum amount of faggots and people that want me dead.
>>69595 I was banned for nothing. Fuck you.
>>69679 sure you were fgt
>>69679 and who are you, exactly?
>>69429 >>69654 Im not completely autistic about it but discord really is a bad idea. Sure you can track down IDs across thread archives every time you reply to someone but in general these threads are much more anonymous at face value. You can just lean back for an entire thread and never reveal your project for example for one round of full anonymity. Plus discords/IRCs don't have a rep for becoming cliques for nothing. Plus the discord-specific privacy concerns on top of this are also valid. On topic, anyone have experience with low-level networking libraries like Enet or Raknet?
(260.98 KB 618x649 RIP Tetradev.png)
>>69640 >But my Ragefaces Get out
(116.45 KB 942x1018 updates.png)
>>69429 I actually have a working thread aggregator with a really shitty ncurses interface that's only half baked, but I stopped working on it when 8kun turned out to be the ultimate disappointment so I never got around to implementing support for other websites.
>>69783 man, 8kun IS a fucking disappointment but you might become the savior of agdg if you finish this or release source code or some shit, I won't be able to wrap my head around it but others might. I think the only website there might be a problem supporting is 9chan, considering it's a different engine, but everything else runs off vichan afaik. The dream is basically being able to have posts from all the different agdg threads in one place, maybe with an option reply to specific board and shit. There were a few concepts like that but yours is the only one I see taking shape.
>>69697 Just a nodev who wanted to be a part of the place.
>>69804 >The dream is basically being able to have posts from all the different agdg threads in one place, maybe with an option reply to specific board and shit. That sounds good, but if you combine threads you'd have trouble when replying to more than one post and/or referencing other posts to people. If you have different threads on a tab, that's not much different from a browser already, just with the same UI for every chan.
God, i used to care about making games, but, I just don't give a fuck anymore. It all feels so arbitrary.
(32.46 KB 136x228 maranax.gif)
>>69816 Never give up on your dreams anon I would never do that because I have no dreams
>>69809 I lifted some bans, try jumping in. If you can't, you are one of the literally two undesirables and shouldn't be in this thread either. Proof of current game project still required to get beyond containment. and if you're nanon, I will get to you once I get beelzebox out of cryo sleep
>>69813 >if you combine threads you'd have trouble when replying to more than one post and/or referencing other posts to people oh, that would sound like someone schizoposting, yeah. Then the solution could be a reply button that would send the user to the specific imageboard to give replies in, while the aggregator would be read-only, with posts from all bunker threads (and I guess progress generals) arranged by date. now I imagine those anons crossposting appearing as separate 7 posts, kek but that can be dealt with with the usual 'hide post' option
In other news: 1. I started a progress general on the local agdg board >>>/agdg/15 2. A very old bunker got refurbished https://endchan.net/agdg/ 3. The agdg bunker on anoncafe might have changed ownership https://anon.cafe/agdg/res/15.html#287 I'm assuming the BO is still a nodev 4. The short lived thread on PLW might be dead soon, along with the board.
(91.61 KB 888x893 pep.jpg)
>>69654 Yeah you can recognize projects, but the meat of the discussion was largely anonymous. Although, it has devolved to language wars over the years. >>69783 Noice. >>69726 The only reason you even check this board anymore is so you can find something to autistically screech at for (you)'s. You faggots are like cat ladies and largely are why we needed alts in the first place and why this place is basically dead now. Here's a last (you). Pic related.
>>69843 > and if you're nanon, I will get to you once I get beelzebox out I am. What happened to the guy?
>>69899 >this place is basically dead now Y... yeah, we'll go back to that one place that has more posts and projects... Which one was that, again?
>>69938 >What happened to the guy? You mean Beelzebuff? Asleep until I'm done with the Deus Ex clone. NOT dead
>>69970 >Not wanting to go back to cuckchan is justification for running the board into the ground and chasing away all the decent posters >Not wanting to go back to cuckchan is justification for not finding an alternative that isn't bleeding a slow death and stagnating Retard logic. You screech about discord too, yet all your yes-devs are jumping off the board onto it while occasionally doing CPR to keep the threads alive. They even have a barrier to entry to keep all the shitters who ruined 8chan in the first place out. At least they have a future. Hell a bunch of devs like MoM-dev and Keyreal abandoned this place for twitter before discord even took off. Majority of the no-devs are gone too, which you can easily see by the fact the /agdg/ use to hit the bump count bi-weekly not long ago. I know you see the signs. It's like the retarded cancer that ruined the boards are the only ones left with the unwarranted self-importance that's shackling them to this rotting corpse. Figures they're mostly on Marks/v/. I'm not even here to watch 8chan die with the cancer who ruined it screaming on their sinking ship. I'm only here to see what the devs are up to and where they go when they inevitably move on.
>>69429 >>69566 >>69595 >using non-anonymous chat programs This is how you make a circle-jerk and eventually turn corrupt just like journos did with their mailing lists. Identityfags are a cancer. The fact it's a chink controlled dataminer is an extra level of shit but still.
>>70056 >pic Just glad it's not a gif.
(1.84 MB 1230x1037 jap house 3.PNG)
(1.28 MB 1303x1091 jap house 0.PNG)
(2.05 MB 1087x978 jap house 1.PNG)
(1.54 MB 1156x1108 jap house 2.PNG)
Finishing Japanese poolhouse for lolis. Fuck me how many threads we gonna make and cant even finish one before the board is kill. Being aggydaggy is suffering.
>>70056 I'm not going back to cuckchan and I wasn't referring to that. Out of all bunkers currently online, this thread is the liveliest.And all the dead boards are better than anything happening on cuckchan. >occasionally doing CPR to keep the threads alive speaking of: >>70388 thanks, shinobudev, it's an appreciated breath of life. And I'm not the one screeching about discuck. I am there and I'll stay there because it's where the non-cancerous people stay. >>70087 >Imagine using discuck for personal messages >Imagine not using it under a pseudonym It won't fuck you if you don't let it. And what b0c2a2 said >a barrier to entry to keep all the shitters who ruined 8chan in the first place if you think asking a faggot for a piece of slapped together programmer art makes it a clique or a circle-jerk you clearly weren't even made for agdg in the first place. The question "where is your game, fag?" is being asked for a reason. Judging by your posts, you don't have one. You are one of the language war niggers and you were never welcome. You fags were driving the actual devs out of this place even before 8chan died.
(243.47 KB 1920x1080 AMX-30 1.jpeg)
>mfw I finally fixed my HUD script that displays different ammo types so that works on different resolution now About damn time, now I just need to make it work so that it supports different weapon with varying amount of weapon modes/ ammo types too, so that I can add weapons like a gauss cannon. I need to rework on the multi purpose device model again soon since it kind of looks like a piece of crap imo. Also I need to polish the tonk script more so that I can add a baguette AMX-30 tank also with 120mm cannon with less redundancy and using inheritance. The pepperpot model that I have is not working so well because when it is in reloading state the coax-gun cannot be used at all which fucking sucks. inb4 use Zscript
What happened with the farming anon brothers?
(371.87 KB 800x410 ClipboardImage.png)
>>70716 oh, cool.
>>70406 >if you think asking a faggot for a piece of slapped together programmer art makes it a clique or a circle-jerk you clearly weren't even made for agdg in the first place. The question "where is your game, fag?" is being asked for a reason. Judging by your posts, you don't have one. You are one of the language war niggers and you were never welcome. You fags were driving the actual devs out of this place even before 8chan died. Any place with permanent identities becomes a circle-jerk. If you don't understand that you have no business using an imageboard. Just go sit in your discord instead.
>>70649 >look up different tank designs (T-55, AMX-30, Leopard 1 and Centurion 1) >The Centurion has the most armor followed by T-55, AMX-30 and lastly Leopard 1 >a fucking tiger tank has more frontal armor (120mm~) then AMX-30 (80mm~) and Leopard 1 (70mm~) >but I really like the design of AMX-30 ruh roh this is not good. I suppose I could redesignate it as B4 with more armor hitpoints instead. It is going to be a bitch getting the model to look accurate-ish as the blueprint while it is sort of in high resolution it has also tons of little details covering the structure of it as well the treads blocking the view how the frontal area and rear are sloped. >120mm cannon heh I meant 105mm cannon, the 120mm one will be reserved for the MBT generation II types. >>69883 >2. A very old bunker got refurbished https://endchan.net/agdg/ About damn time, its kind of a shame it hasn't been used much earlier on but I suppose better late than never. >>69816 Feelings are of known, I'm hard pressed to keep up my low motivation to make enough progress instead of procrastinating or playing modded games.
>Have two weeks off becaues of coof scare even though it was literally nothing >Back to work on the 1st >The only progress I've made in the last 4 days was moving one file into a different folder
(467.99 KB 640x640 gabolsad.png)
>>73025 Oh I was like that the first month of coofvacation, just don't let it drag you down and try to make developing an everyday thing.
(97.35 KB 261x333 steve_hogarty.png)
finally have the basis of a level editor made for the game I'm working on without using any more libraries than what the game relies on, it almost looks like a real application
(126.07 KB 432x434 output.webm)
Died for a few weeks but I think, I'm back on track. >Added icons for actions >Added GoTo action >Implemented "Speak" action >Refactored Doors >Added first sensor to check if toilet is occupied >Added "Mine" level, artist was working on for a while My AI code is a mess. I need to clean it up heavily.
(80.29 KB 745x1173 realization_setting_in.png)
>>73110 Well I screwed this up somehow. Tried reducing redundancy by using class inheritance, and now it doesn't update properly after I remove the last object from the level. Everything is a tile on a grid, when a tile is clicked it should be removed and there should be a blank white square in it's place, after my changes it no longer redraws the empty space where the last tile was removed. I have no idea why since nothing is left in my sprite list, but the tile remains. Then I thought >why not just have one class and pass the only changing parameters to it? but the result is the same, and now even reverting to my old code results in the same. Then I discovered why it was doing this. I had made a list of the classes and when adding a tile to the screen would call the class from the list by index. Now I'll have to write my new code all over again and see if it works with the single class taking parameters.
>>73336 Phew, not only fixed my mistake but also implemented a feature to just fill the grid as the user drags the mouse around, a lot better than having to autistically click each square.
(342.85 KB 1920x1007 30-07-20-1596105320_scrot.png)
(283.67 KB 1920x1007 30-07-20-1596105332_scrot.png)
(315.03 KB 1920x1007 30-07-20-1596105339_scrot.png)
(187.46 KB 768x247 AMX-30B2.png)
(100.22 KB 1554x2113 amx-30.png)
Is there any sort of a specific workflow or a plugin that makes creating intricate smooth shapes less painfully to model without fucking up the topology? I'm in the process of modelling the AMX-30 turret but it has weird angled smoothing around the front which is for me difficult to cut it via the knife tool, as I ended up with this horrendous mess. Not even modelling the T-55 and T-64 turrets were that difficult to model. I tried also using the nurbs path but it seems to be even more difficult to get the shape around correctly and it doesn't look like I would be able to angle the front turret sideways with those difficult contour lines.
>have great idea for a game >its beyond my capabilities to do well >already on a project thats pushing me and made little progress with
(461.51 KB 640x480 AGDG_2018_August_8th_DemoDay.gif)
Is everyone ready?
(357.87 KB 680x485 spaghetti.png)
>>77233 (checked) Its don't.
>>77233 The last demo day I tried was over 2 years ago
(840.34 KB 1016x1080 1470952049463.png)
>>76918 spit it out
(126.34 KB 365x382 TRASH.png)
>thought I'd crossed the dunning Kruger peak, and been humbled already >trying to set up a state machine >following along in video, dude makes one "easily corrected error" in it, points it out in the comments, and even tells the audience how to fix it >comment has tons of thumbs up, people even thank him and say it makes sense now >can't for the life of me figure it out >keep checking tutorials, none of it makes sense >frustration increases >finally find the manual's and it does, but at this point I've scrapped what I had and started over in the script/project files >confidence shattered >don't even want to touch it for the rest of today On the bright side, now I can safely say I know fucking nothing about this, and it's only uphill from here. God this is going to take forever.
>>77612 I think we've all been there at one point or another. I haven't gotten anything done in a week, my focus is just gone and I can't think of what I'm doing next.
(10.06 MB 1280x720 2020-08-05 02-25-27.webm)
>>77233 not this time, sadly still getting progress done on shit
(245.09 KB 960x642 output.webm)
Some progress >Implemented light switch >Implemented minecart fall >Fixed low FPS in mine level >Fixed collisions in mine >Implemented light flicker

[Index] [Catalog] [Archive] [Top] [Reply] Posts: / UIDs: / Files:
Delete
Report

no cookies?