135c4f No.16349623
No hardcoded shit allowed edition
Resources
>>>/agdg/
>>>/vm/
>#8/agdg/ via irc.rizon.net
>Dev resources: http://8agdg.wikidot.com/resources
Links
>Wiki: http://8agdg.wikidot.com/
>Beginner's guide: >>>/agdg/29080
>Previous thread: >>16296248
Announcements
>QUARTERLY DEMO DAY SCHEDULED FOR May 5th (05/05)
>Please contribute to the wiki if you can
f16fe1 No.16349640
What the fuck is that OP image and why is it not the Tetradev edit?
2835fc No.16349659
>>16349623
this op pic is gay, this thread is gay
135c4f No.16349660
Feels good to remake the old cooking system
Recipes and Ingredients are now xmls with cool effects and traits, super easy to add more content and make mods.
The effects of the recipes will give them function/personality, stuff like soups and stews, both can use the same ingredients, but the way you cook change
Soup will have effects that slows overcooking and even stops ingredients from burning, allowing easy mode cooking, start up the stove, shove ingredients in and go do something else, come back hours later and it is still edible
Stew will start with overcooking penalty, making your food burn faster, but as your cooking skill increases you unlock skills that makes cheaper easy to activate traits of ingredients, and since you gain points to unlock those traits as the dish is cooking you have to balance the risk of burning the food with the amount of points you can acquire and how many ingredient traits you want to activate, so when cooking a stew you want to keep and eye on the pot to make sure you can get the most out of your ingredients.
>>16349640
it was a template
f16fe1 No.16349878
>>16349660
I don't have the Tetradev edit on hand (God knows where I put it) so I made this instead.
361eed No.16349951
>>16349878
You forgot
>thread gets 300 replies
135c4f No.16350075
>>16349970
>>16349951
>wasting time shitposting instead of gamedev
Here your beloved image
f9d8b9 No.16350088
Simple npc moving and following behaviour done. Will expand on this over the weekend.
fbc2b3 No.16350158
YouTube embed. Click thumbnail to play.
While researching audio for my game I found this. It's a method to semi-realistically model sound in real time. It has an open source implementation and a research paper explaining how it works.
The research paper is really good all around, even if you skip past the more intensive math. It goes into how well the system it scales and how it could be further optimized.
http://www.jthorborg.com/occlusion-research/bthesis_soundocclusion.pdf
The better parts of the audio demo happen later in the video, about halfway through on to the end. The black boxes and walls that absorb 100% of all incoming audio I assume exist for demo purposes to isolate certain features of the system.
Links to the builds and source.
https://bitbucket.org/Mayae/soundocclusion.git
https://bitbucket.org/Mayae/soundocclusion/downloads/
Inb4 Aureal A3D
135c4f No.16350220
>>16350158
Is this shit for real? Already implemented for free? I might actually add it to my game (after it is ready since this is just fluff and changes nothing gameplay wise)
135c4f No.16350414
>>16350098
looking sweet anon
4c015d No.16350417
>>16350075
tfw no dev in last month
fa7580 No.16350792
I've been considering for some time the pros and cons behind third-person and first-person cameras, ever since I played Hitman and noticed how much I enjoyed its camera. It did a remarkable job on changing my mind about it. I'd like to hear other anon's opinions on the differences, problems you have with the two, anything that might help me with deciding here.
No pressure, though. This game dev thing and programming in general is new to me and I don't think I'll get too far with it.
361eed No.16350972
>>16350075
Did Tetra dev ever reappear?
47dfd1 No.16351013
>>16350972
Pretty sure he's dead Jim
adf63f No.16351050
>>16350792
It's a design choice like any other. It depends on the experience the you want to create. Some things work better for first person like immersion and intuitive locomotion. Other things are better with third person like spatial awareness and showing off the mc. You have to tailor your decision to the genre and the style you want. That robot it not cute.
f2f915 No.16351249
So apparently I'm retarded. I'm looking for this UE4 marketplace asset that was free for a month, did anyone pick it up: https://www.unrealengine.com/marketplace/en-US/slug/modular-cliffs
f560a2 No.16351285
f2f915 No.16351312
>>16351285
I wish someone would upload these to a torrenting site, they look really nice and save me the trouble of making rocks myself, after all who doesn't like a good asset flip
15b2a4 No.16351318
>>16350158
>he actually went to sonic college
>spend an entire page on a shitty box carrying analogy to explain exponential vs linear growth
>in a paper that casually references complex differential equations
what are those faggots in academia up to?
15b2a4 No.16351320
>>16351318
I forgot my image because academia rotted my brain.
53289c No.16351337
>>16351320
This is undergrads thesis paper, it's very likely he got the majority of work from other people and then just wanted to demonstrate to his instructor that he understood it. In other words, he didn't actually do any of the work, he just wrote the paper, which is sadly all too common.
Its even more evident when you realize half of the equations are just standard formulas (see the entire page dedicated to defining f*g)
15b2a4 No.16351371
>>16351337
>the faggot who wrote this works full time for unity as an audio engineer
7ed7db No.16352573
A* pathfinding done, until I realized it doesn't factor in perpendicular tiles when checking diagonals. Want to add hezitation around corners or terrain types with different costs.
084476 No.16352602
>>16350220
You'll probably take a performance hit if you have a lot of sound bounces and you don't optimize the system the extra bits that were talked about in the paper. Knowing you and your brother you two will find a way to make it work.
135c4f No.16352682
>>16352602
It will take quite a while before I bother trying, but damn will be a nice thing to add later on
58f9bf No.16355123
>>16349623
Player rotates independent of forward velocity FINALLY. I hate being a mathlet but every time I try to look up game math info, I get stuck looking at formulas I have no background with. All of the player movement code makes sense to me now, so expect another gif of progress soon.
Also I think I already posted this, but gibs working like a week ago.
>>16349660
Even if I don't recall what this is for, the dedication truly is inspiring anon.
854357 No.16355221
Remade her shirt and skirt to match new body topo and weight paint.
9b7750 No.16355234
>>16355221
Delicious Shinobu.
854357 No.16355239
>>16355234
I've done so much fucking shit for this project man. Why the fuck. Literal years of nothing but autistic weeb shit. It's fun but like sometimes I step back and say "Fuck".
9b7750 No.16355392
>>16355239
kek
You doin' good, man. Your progress is inspiring. Have you think on working on other projects so you don't burnout yourself?
4c015d No.16355405
>>16355239
Even if it's autistic, and not something you can openly always show off, it's still something tangible. Be proud of it.
846f0e No.16355415
>>16355239
At least you made progress.
854357 No.16355417
>>16355392
I don't get burnout. There's always fun different things I like doing. It's more important to take breaks with something productive like going to the gym and eating well. /fit/izen here.
7d2bf3 No.16355419
How come you guys don't like working together?
9b7750 No.16355426
846f0e No.16355436
>>16355419
I was willing to help other moonman anon devs here to help them make more progress but those spriter anons are gone now and the other guy takes forever to upload new shit for me to add in, so for me personally I'm getting tired of working with other people so I would rather go the one man dev route which is the only reliable method for me to develop and I have done so previously more or less. And I have other issues to take care of such as lack of motivation to do fucking anything even playing gaems which obviously means I need to take care of my health issues first before I can dev again which also means I would like to learn C89-C99 + SDL 1.2 so that I can make my own 2D engine instead of using other people shit like that fucking broken piece of shit called Spring RTS or Graf Zahl sourceport.
7d2bf3 No.16355456
>>16355436
That's reasonable. I was thinking of learning UE4 myself, but I don't have any art or programming skills. I wanna start soon, but I'm fucking terrible at math. So I'm conflicted
f9e6c0 No.16355466
4c015d No.16355469
>>16355466
My gut says the official(?) github is better than a 3rd party tutorial site, not sure though.
846f0e No.16355479
>>16355456
My art skill is pretty shit too, previously I tried to do some weird texturing method such as cubical/square only atlas texturing to texturize my tonk models or even generating textures from some stock images I have found from the internet but I did not much care about such details because it will take for me forever to do it properly. And Blender baking tools is pretty shit to begin with so I don't feel even feel like to make low-polygon model of my SU-100 model for baking, I already did something similar with my Ore Mine and certain details was not visibile on the baked mesh. I mostly do a lot of hardsurface modelling such as vehicles, tanks, structures and weapons because they are the easiest to work with.
>or programming skills.
if I could have turned back in time I would slap my old me and force to learn programming instead of wasting months on Wings3D as it has given me a very poor net benefit.
f9e6c0 No.16355484
>>16355469
I think basically everything on the github is community-submitted. I'm asking because the tutorial makes fuck-all sense to me and doesn't seem to handle the case of separate present and graphics queues properly in terms of synchronization.
f6d6d3 No.16355606
>>16355419
Pretty much everyone is too inexperienced.
Working together means you have to actually make a game, not dabble on the project every now and then and get demotivated for weeks straight and then lose interest in the project and start a new one. And that's not even accounting the skill required to make a whole game in the first place. Just because you can make a skeleton of a platformer doesn't mean you can make all the features and handle the entire codebase of a full platformer game. Just because you can draw some decent drawings doesn't mean you can design and execute an entire game's worth of artwork.
And even after that someone needs to have the ability to design a whole game. Neither programming nor art accounts for designing levels, writing the plot/scenarios in detail, and generally the ability to figure out all the small steps and details that goes into everything. Most programmers here are enginedevs or have never made a full game or both, and most artists are drawfag-tier who mostly have experience drawing isolated characters even though a videogame consists of 90% non-character assets that need to be put together into a large whole. The more you go towards arcade-y games where those details aren't as relevant, the harder it will become to find motivation in people to work on it. And even if you have all the abilities to JLMG, you still have to find someone who not only has that ability as well, but also wants to make the same kind of game that you want to make.
4408ed No.16355715
>>16350158
This is really quite impressive. I think the coolest part of it are the surface reflections. Like when he stood in between those hanging metal pipes or stood before the glass window inside the room where the music was playing and you could hear the glass vibrate.
>>16350220
As >>16352602 pointed out, you can expect this to be very resource intensive when compared to traditional "simple" doing it by hand. I didn't look into it in detail, but if it's anything like past systems that went in that direction, you'll have build a separate audio collision level to keep the polycount low and remove small objects that would create interference.
And even then it's questionable how this would scale in real world applications. The level of dynamics involved in actual gameplay is nothing like what that tech demo showed. You'll have a much higher number of surfaces and audio sources, some of which might move all the time. So the obvious question is how that system scales in the face of potentially having to repath and recalculate a ton of audio sources all the time.
Another solution worth looking at is Steam Audio. It doesn't do the craziest shit from that tech demo, but it would still be a substantial upgrade from "normal" game audio.
I'd love to play with something like that, because I'm a sucker for good game audio (even though I hate working on it, because I obviously can't listen to music while working on audio). It doesn't make sense for codename waifus though, because it's a topdown game, so neither reflections nor occlusion are a concern and we'd just be wasting CPU cycles. If it made sense for us, I'd use Steam Audio since it has an fmod Studio integration and we use fmod for audio rather than Unity's standard audio system.
>>16351318
He probably had to meet a minimum length requirement, couldn't do it on content alone and decided to pad the shit out of it. (And it's still quite short). I did the same thing in the thesis I wrote in preparation for my bachelor's thesis. In my proper bachelor's thesis I didn't need to pad anything, because the subject matter was complex enough that I had no problem reaching the requirement on substance alone.
>>16351337
>In other words, he didn't actually do any of the work, he just wrote the paper, which is sadly all too common.
There is nothing sad about that, because original research isn't a requirement for a Bachelor, because it's not the point of it. Not everyone who goes for a Bachelor wants to become a researcher, after all. If you go from Bachelor to Master, the requirement for original research is increased, but it's still not the main focus. The place where you have to do a serious contribution to your field is when you want to get a PhD.
4408ed No.16355783
>>16355419
Well, codename waifus is a team effort, because it originally started as a small gamejam project. At this point I'm doing most of the work myself though, because the other guys are busy with real life problems.
The last thing I finished was ripping out the ballistics simulation I wrote for it. It supports:
>gravity
>wind
>rebounding off walls for low-energy impacts
>ricochets for high-energy impacts
>penetration of walls with trajectory disruption
>the above three depend on the hardness of the surface, which is configurable on a per-material basis
>exerting force on a rigidbody, if one is attached
>built-in system for pooling projectiles for performance
>detection of when the projectile flies past a reference point (can be used to play a supersonic crack whenever the projectile flies past the camera, as seen in Arma 3)
8167d4 No.16356116
YouTube embed. Click thumbnail to play.
fucking with behaviour trees again, not actually willing to put enough effort towards real progress
added a sub-script node, although i currently don't have use for it
added a node to find the closest pawn/item/storage
if the move node changes result and the last one wasn't null, it's considered as a failure in order to reset the sequence
901859 No.16356260
>>16355466
https://software.intel.com/en-us/articles/api-without-secrets-introduction-to-vulkan-part-1
https://vulkan.lunarg.com/doc/sdk/1.0.57.0/windows/tutorial/html/index.html
It doesn't matter. No matter what you follow you're going to make something stupid that you have to rewrite, just follow any one of them, get something on the screen, then figure out what you did wrong.
0f29cf No.16356715
> Implemented rotation and restoration of angle when piece is unselected
> Added black pieces but they need to be reflected and that won't work that easily.
I need a way to know whether black or white piece selected. I guess it's time to implement turns.
d2ce1f No.16356804
Should I use a Character Controller or a Rigidbody for player movement in a Unity FPS game when I am planning to add rocket-jumping or similar forces?
14ea17 No.16356808
>>16356804
Both. Use the CharacterController to steer. Rigidbody is needed for collision detection (on at least one of the bodies colliding).
14ea17 No.16356822
>>16356804
You would probably not use the Character Controller hitbox for the bullets if you're going for body part hits. Sage for double post.
cada32 No.16356850
This question has probably been asked a million times already, but what's the likely hood some average non-coder autismo could grab a program like Unreal Engine 4, watch some tutorials, read some pdfs, etc. and crank out a simple arcade-style game?
And if that likelihood is attainable in any way, what program would be best?
8167d4 No.16356856
>>16356850
UE4 is a wee bit more user friendly
cada32 No.16356874
>>16356856
So do you think if I focus my autism enough, I could make something playable?
8167d4 No.16356915
>>16356874
probably not, but it's easier to learn than unity
cada32 No.16356938
>>16356915
Close enough to give it a shot. Thanks.
6679aa No.16356940
>>16356850
If you mean simple 2d arcade like pacman, it's possible. I think there is a very simple Unity arcade space shooter tutorial. Unreal Engine should have some arcade too, or at least some simple mobile game tutorial. Keep in mind though you'll need to at least understand the very basics of coding even if you're just copy pasting everything from the tutorials, so you have to go through the respective beginner coding tutorials.
cada32 No.16356974
>>16356940
I took a year of C++. Think that would be enough?
6679aa No.16357050
>>16356974
Should be more than enough to follow any simple tutorial as long as you haven't been sleeping for that year. Before i started game dev i did Java for a few months and that was enough for me to replicate space invaders with a tutorial.
a8357c No.16357207
>finished my short platformer game
>brother plays it for roughly an hour
>player character MAY HAVE gotten stuck on a ledge during a jump, but maybe not, and it didn't cause any major issues
>probably played this game for dozens of hours while creating the levels, and this never happened
>try to recreate the issue by jumping on ledges at different heights and angles but it works fine
I think I know where in the code it might cause this, but I don't want to break anything at this point, so I might leave it alone. I just don't want anyone to find this on demo day not that anyone will play this game anyway
135c4f No.16357447
>>16355123
its for our faming sim with deep mechanics
>>16355234
no lewding allowed
4c015d No.16357549
How difficult would it be to make a coherent randomly generated platformer? I know Spelunky and maybe Rogue Legacy did it, any others?
347159 No.16357734
5d0844 No.16357801
>>16357549
>create genetic algorithms and genetic methods to to randomize enemies/enviroment
>change them until they are balanced
>???
>just like make gaem
bce311 No.16357827
>>16357207
It is going to be found by multiple people during Demo Day and those anon's are going to help you fix it right after laughing at you
033f34 No.16357858
>>16356850
Godot's 2d is perfect for that, but all of it is usable.
0f29cf No.16359242
>>16357858
I second that. It was easier for me to create hello world game on Godot than on Unity. If you are trying to make something complex like HL2 or something, Unity/Unreal might be preferable yet. Oh, and if you want monetisation - intergation with other services like admob sucks in Godot, at least for now.
c23d25 No.16359478
>play game made in Unity
>it has noticeable input latency
>it's 2D but requires a real video card
>there's awful frame skipping at random intervals
Unity a shit
daee39 No.16359493
>>16356974
No, 0 is enough but if you are just asking if you can do something rather than doing it then you can't. Get off your ass and start doing something programming is best learned during practice.
135c4f No.16359632
>>16356850
Just start learning to code, will be faster and better.
Javascript is the easiest to learn due to how dumb easy is to run a piece of code on it
C# is probably the most useful, but it's a bit more complicated to start coding on it
(Although downloading Unity and creating a script then attaching to a game object is not all that hard, is way harder than pressing F12 and writing code in your browser, also, most 'learn javascript' websites actually have the whole IDE inside the website, making it super easy to get started)
135c4f No.16359647
>>16357549
Make a lot of pieces of maps that fit together with some rules
create a entry and exit point
Literally A* using the map pieces
Although some variation of A* would be better, otherwise it will make a lot of samey straight line maps
b97aa9 No.16359837
>>16355419
I am completely unreliable and any help I would offer to a group effort would be an empty gesture. Better to just keep to my own devices so nobody else has to be disappointed.
0594ed No.16359968
Begun working on levels today and made touching some tiles take you up or down one or more levels. Ground floor tiles still show and the offset gives it some depth I think.
0594ed No.16359996
>>16359968
fuck forgot to turn to sound off. Ignore the podcast
68d695 No.16360027
>>16359632
That advantage of Javascript may be true, but it has the disadvantage that you're doing Javascript on top of the liquid shit that is the Web stack.
135c4f No.16360075
>>16360027
Just for learning my man, move away from that shit the second you are not doing a web page
f9e6c0 No.16360382
>When the validation layer segfaults
f9e6c0 No.16360394
>>16360382
>When the validation layer doesn't segfault in valgrind, only outside
I fucking love heisenbugs
455a87 No.16360484
f9e6c0 No.16361424
>>16360394
>compiling without any debug checks makes the segfault go away
>compiling with -fsanitize=address also makes the segfault go away
fuck.
f9e6c0 No.16361839
>>16361424
>JIT-compiling the source code and running it in the fucking Terra interpreter causes it to segfault in a different location
>Linking against a newer version of the validation layers literally reverses all the above behaviors
And after all that…
>Turns out it was because I left pResults in my VkPresentInfoKHR uninitialized; all I had to do was assign it to a null pointer
gotta love that undefined behavior
846f0e No.16361894
>mfw it might be theoretically possible to "port" Spellcross game to OpenApoc
>Replace building grafixs with territoral maps or some sort of check points
>there could be at least 3 types of territory fake buildings and several "checkpoint" types
>1 the main base of the player
>2 the regular contesting territory
>3 the 'other side' territory
>4 'other side' territorial main base (the "final" battle for the territorial operational range in order to unlock more territory to fight for)
>1.5 some few buildings that are responsible for the production of equipment
>use some kind of rock or whatever that limits the player transporter vehicle operational range
>since 'other side' is not that passive the dimension gates could represent its local aggression power or something like that
>which means there will be 2 types of "rock" one is a hard one that never changes and the other one does change
>the "local soft rock" can only be destroyed when the player captured the local other side main base
>player mainbase will only hold the necessary stuff such as repairing of vehicles and holding loot, items and other equipment
>Add several "strategy transporter ground vehicles" which serves as the main basis for the transportation of troops
>those strategy transporter variants could represent the rank of the player which means higher rank = more units allowed to deploy
>depending how true the mod should be it could be possible to make it a bit more open-ended/sandboxy whatever i.e the dictation of progress is fixed by week counts
>the map time could be "fixed" i.e never allow to go past 12:00 or if possible change the way this counter works and use it that to represent "turns"
>obviously that also means it the way the strategy layer time passing mechanic needs to be changed for it to work properly
>problem 1: is it even possible to have vehicles with full control over its weapon behavior on tactical mode? As well turreted vehicles support which I am not sure if it works at all
>solution 1: I could just make 2 variants of the same weapon one that is in auto mode and one in manual mode kinda how the grenade weapon works that way (manual)
>problem 2: how the fuck do I make sure that certain equipment types can only be used on certain unit types/class? For example commando unit should be allowed to have better weapons and/or ammo while light infantry and heavy infantry have fewer equipment.
>major problem 3: I would have to figure how the fuck I can output different mesh pieces with different frame animations which I still don't know yet how to do this in blender
>problem 4: will the vehicle units have angled sprites on sloped terrain?
>problem 5: I would need to figure out how lua works which has a different syntax and behavior compared to c and bydłon.
>problem 6: line of sight and range bonuses on terrain height, I am not sure if OG Apocalypse has this feature at all that units get bonuses when they hold a mountain
It is kind of a interesting way trying to port a game that way, considering OpenApoc has a lot of features set that is similar to how Spellcross works (Research of units and topics, Experienced Units, Persistent Units, Upgrading units, Territory system) instead of just gidding gud at C and making my own engine but the amount of effort required to make this mechanic work at least in concept-ish way seems quite huge too, especially the problem 3. Meh I guess it will be just that another ideafagging of mine. I think even Jeb Bush has more energy left than me tbh
7ca61a No.16361975
>It would be fun to write a real NES game
>Start learning 6502's assembly
>Having to memorize these instructions and manually juggle registers is annoying, I should make something that is easier to write and then convert it
And that's how I accidentally started making a compiler.
15b2a4 No.16361981
>>16361975
There's already a C compiler for the NES. Kind of defeats the purpose of programming for it, though.
7ca61a No.16362044
>>16361980
>>16361981
I'm not after something as high level as C just more readable and convenient than assembly by hand.
15b2a4 No.16362060
>>16362044
you're right, we're illiterate
4ed0e7 No.16362075
>>16362044
Is C worth learning for the control and potential efficiency?
286d75 No.16362104
>>16362044
How about C with heavy use of extensions that allow you to use assembly?
I haven't done any of the sort so I don't know how good it is, but I believe there's extensions that give you inline assembly where you're permitted to use variable names, possibly defined in C code. That should give you the automated register allocation/shuffling that you're looking for.
If you decide to build your own compiler anyway, I'd say it's at least worth looking into.
15b2a4 No.16362148
>>16362075
The real benefit from learning languages is figuring out how to structure projects. C is fundamentally very simple, but that simplicity means you have to be careful not to created a convoluted mess of a project. After your first language or two learning the syntax isn't going to teach you much.
8f330e No.16362170
>Turn on -wpedantic
>Get hundreds of lines of whining about m-muh designated initializers m-muh C99 extensions
>I have -std=c++2a
Clang blows. Why is it even whining about designated initializer list ordering? Why does it give a fuck?
005c67 No.16362184
Anyone has any experience with Jaxe? Any opinions?
4408ed No.16362190
>>16362170
>Turn on -wpedantic
>Why is it even whining about designated initializer list ordering? Why does it give a fuck?
Anon, are you for real? It's like you don't know what the word "pedantic" even means.
005c67 No.16362194
>>16362184
Fuck me, I meant Haxe. Sorry for the typo.
4408ed No.16362199
>>16362184
>>16362194
The engine speebot was made in uses Haxe. So he's your guy.
112599 No.16362292
>>16362184
I tried Haxe years ago. It was my first experience with type infering. I thought I could rely on that, but it got in the way when it couldn't handle some int vs. float difference. I think it also demanded Java like file and directory structuring, that was annoying. I was targeting Flash/AS3, it fucked up the exception handling in a way that the browser plugin never accepted and that made debugging impossible when the project got more complex. I abandoned it after that.
286d75 No.16362428
>>16362170
-Wpedantic makes the compiler demand compliance with the programming language standard. That means if you use extensions, it'll complain.
The point of -Wpedantic is to make code universally compile-able. Don't turn it on unless that is your intention.
0f29cf No.16363571
> Implemented turns. I had very little to add because my game wasn't totally hardcoded shit. Feel proud AF.
> Added cathedral which was also relatively simple.
Basic gameplay is ready. But I still need.
< Mirror black pieces
< Prevent opponent building in areas occupied by player
< Removing opponent buildings from occupied zones
< Victory conditions
I have a lot of math ahead.
I should be ready with my shitty tabletop clone by the demo day.
Also, OpenGameArts having Game Jam, might as well take part in, since I stole most of the stuff from them.
f9d8b9 No.16363769
>>16363571
I don't know if i'll have anything I'd like to show for demo day, if I do it'll be my first. Any requirements?
3c9638 No.16363989
>>16361975
>Having to memorize these instructions and manually juggle registers is annoying, I should make something that is easier to write and then convert it
Just print out a card of 6502 instructions.
c3697c No.16364257
>>16359478
There must be some setting configured incorrectly. Because I've played a lot of Unity games that require precise inputs. And I've never heard anything about Unity having any input buffer like Unreal's infamous 6-frame buffer
901859 No.16364693
>>16361975
>>16362044
This is the plebs response to assembly. Just swallow the pill and learn how your macro engine works. You're going to have to learn how the registers work no matter what you do. If you ignore your tools and ignore the language you're trying to write it's going to be even more work when you have to start over so you can do it correctly.
854357 No.16364854
Working on bathing animations. This one is when she wants you to GTFO before undressing.
70cef0 No.16364870
>>16349623
>hardcoded shit
what dat mean
>image
>>>/cancer/
f16fe1 No.16364880
>>16364854
Well, she's not very nice, is she?
15b2a4 No.16364891
>>16364870
Go be a nigger somewhere else.
f9e6c0 No.16364895
>Shitty Vulkan application runs 5x faster using X11 than on Wayland (probably because it's ramming itself through the XWayland compatibility layer or some retarded shit, or maybe Sway's window loop is actually shit)
Speaking of hardcoded shit, are there are any opinions out there at hard-coding some data into shaders?
854357 No.16364902
>>16364895
>hard-coding some data into shaders
Like what? Is this for a custom engine?
f9e6c0 No.16364908
>>16364902
Like vertex data for a model, and yes.
fbc2b3 No.16364926
>>16364854
Cute.
>>16364870
Software can read information from files, or data can be stored within the software itself. Hardcoded means that values are stored in the software and not in the file. The term is often used when talking about data that really shouldn't be stored in the software such as login credentials.
901859 No.16364934
>>16364895
>probably because it's ramming itself through the XWayland compatibility layer or some retarded shit
Aren't you getting a direct graphics path through the VK_KHR_wayland_surface extension?
f9e6c0 No.16364941
>>16364934
I'm using GLFW and it's using VK_KHR_xcb_surface.
901859 No.16364959
>>16364941
Then that's it I suppose, you'll just need to change it, speaking of Vulkan this is my luck with it currently. Something about my triangles being fucked, OpenGL could deal with it somehow, Vulkan doesn't like it. So I have to actually fix my shit. I've been wanting to port my stuff to wayland for a while but I haven't found nearly enough time to fix up the Vulkan renderer. It still has a few bugs left.
537e45 No.16365006
>>16364895
Wayland is the ultimate RedHat/Freedesktop meme. How the fuck is it still lagging behind a freetard implementation of a 30-year-old display server is truly astonishing.
f9e6c0 No.16365015
>>16364959
Thanks. I'm going to deal with it for now.
>>16365006
Sway does get rid of the screen tearing I got on i3 but it comes at the cost of performance and of course having a bunch of input shit broken. Input devices, mouse capture in vidya, etc. Fun shit.
854357 No.16365029
>>16364908
That's not very specific. Why are you being hesitant about it? Do you want to store massive amounts of memory on it or? Arrays? Read only variables?
112599 No.16365067
>>16364959
What's going on there? Did backface culling eat half of your triangles?
Actually, how low level is Vulkan? Do you have to reimplement usual features like backface culling?
7ca61a No.16365093
>>16364693
I'm not talking about avoiding how things work (after all I'd need to know how they work to have non-awful output), just reducing the manual busy work. Compiler probably wasn't the right term, something more like a pre-processor or macros I guess. I should try and get something finished before starting yet another thing anyway.
4408ed No.16365487
>>16363769
>I don't know if i'll have anything I'd like to show for demo day, if I do it'll be my first. Any requirements?
Anything playable is fine and it pretty much doesn't matter how rudimentary.
4408ed No.16365490
Also, before I forget:
>only 19 days till demo day
f6d6d3 No.16365784
>>16365490
Time to speed-dev for the next 3 weeks
901859 No.16366427
>>16365067
Backface culling is described in the spec here:
https://www.khronos.org/registry/vulkan/specs/1.0/html/vkspec.html#primsrast-polygons
This is a winding order issue. I honestly don't know why OpenGL renders it correctly. I guess I just have to fix my compiler until it generates the right thing. Probably will just add a final check that makes sure all of the verticies for every face form a correct triangle strip. This is what one instanced call looks like in RenderDoc.
>>16365093
Your assembler should provide everything you need in it's macro engine already. I guess that's what i'm trying to say. Something like this:
https://www.cc65.org/doc/ca65-12.html#ss12.1
0f29cf No.16367050
>>16363571
Mirroring black pieces was easier than I thought since most pieces are exactly the same.
Here is the rules i'm truing to implement:
> If you completely enclose a part of the city with your buildings alone or with your buildings and the wall, this part of the city becomes your property and your opponent may not place any of his buildings within it. The buildings must meet wall to wall, a corner to corner contact is not acceptable. Your opponent may claim space in the same way. You may not use the Cathedral as part of the boundary to enclose the claimed space. Neither you nor your opponent may claim space on your first move.
> If you enclose and therefore isolate one and only one of your opponent's buildings or the Cathedral you may remove it and claim the space enclosed. The building must be removed immediately after the move during which it was enclosed otherwise it must remain where it is and the space is still available to your opponent. Your opponent's building may be replayed in a later move but the Cathedral once removed is not replaced for the remainder of that game. If you enclose two or more buildings. one of which may be the Cathedral, then none of the buildings may be removed and the space is still available to your opponent
I've spent evening on watching graph theory.
What I've learned is that I need to find circuits that are formed by buildings and city wall.
And to do that I need:
Remove nodes with 0 or 1 edges since they can't contribute to circuits and repeat until none left.
Find disconnected graphs. Should be easy with flood algorithm.
Add wall as artificial node to graphs
Find circuits. Have no idea how to do that yet.
4c015d No.16367740
If you're using a rendering library (say SFML for example), how would you create HTML-based UI elements for your game? I know it's used a bunch elsewhere but I really don't get it
ec1a47 No.16367939
today's progress on babby's first engine:
>map levels and a way to transfer between them. When on higher levels (like up on roofs), visible tiles on lower levels still render. It's a start.
>render and control pixel transparancy
>simple shooting AI
>missle to npc / player collision. Depending on who's doing the shooting.
135c4f No.16368528
>>16364854
I want to bully her
4184c6 No.16369114
>>16364854
How did you do the lip sync?
854357 No.16369122
>>16369114
Put a sound clip in the Video Sequence Editor, then enable sound scrubbing and showing wavelength form.
b34d4a No.16369366
I forgot the name of that zelda-looking pixelart game that made a successful kickstarter. Did the guy abandon /agdg/? His game isn't even in the wiki anymore.
b34d4a No.16369368
>>16369366
Nevermind me, I'm retarded.
e81675 No.16369633
>>16366427
>I honestly don't know why OpenGL renders it correctly
Throw in a light and shading, it might highlight the same issue.
Blender too uses winding order. I wrote a model generator. I have no idea how to determine winding order while generating the vertex/face data, so I made a loop to fix that after Blender has the data. It knows the general direction for the normal vector, so for any face where Blender calculates a normal in the opposite direction, the order gets reversed. It works, but it feels a bit dirty. Blender renders polygons double-sided, so it wasn't missing polygons looking strange, but inverted shading causing dark patches in a bright area. That problem is not visible in shadeless mode.
c3697c No.16370375
>>16370283
Seeing that reminds me of how I've wanted to make a game around Homestuck's early Sburb system. Where you you keep building up a tower to try and get to a certain point with enemies spawning in faster and tougher over time.
You could easily build a whole game around the idea, I'm surprised no one's tried it yet.
901859 No.16370386
>>16369633
The problem was something else entirely- it was supposed to be rendered as a series of triangle fans! There is actually an optimization that requires me to start using triangle strips to merge faces within a leaf but that isn't being done yet. Once I can totally figure this out I need to start applying those optimizations.
The engine doesn't do lighting yet.
4a6131 No.16370434
>>16361975
Congrats anon, you were born too late to properly bank on your own intelligence. Instead you're stuck here with us. Forever.
286d75 No.16370444
>>16370375
I don't really see how the construction would really benefit the gameplay. In Sburb, it was only necessary to build to reach portals up above, serving only as a gating mechanic.
The other issue you run into is that having monsters traverse player-designed environments requires some serious design in mechanics and AI to prevent cheese strategies.
bddc5f No.16371362
>>16370444
>The other issue you run into is that having monsters traverse player-designed environments requires some serious design in mechanics and AI to prevent cheese strategies.
One easy way to solve that is to have monsters teleport beside the player if they can't find a path.
286d75 No.16371495
>>16371362
Solving cheese strategies by making the game cheese the player in response doesn't make the game enjoyable.
bddc5f No.16371507
as the saying goes, an eye for an eye makes the whole world blind. If the players wants to cheese the game I will cheese them back.
11518a No.16371672
>>16370444
It benefits the gameplay because as the larger enemy variants start spawning in. You need to build to get away from the danger because they start taking up physical space and the only way to mitigate that is either to try and kill them faster than they can spawn in (a strategy that won't work for very long) or you can build and make more space to allow you to get around them.
>>16371362
>>16371495
Well as you can see from the image, they spawn everywhere so cheese strats involving pathing abuse won't really work because you can't reliably predict where they'll spawn and they're spawning constantly. So if it's built right, while one enemy spawning right next to you isn't much of a problem, they can potentially draw your attention for long enough for the rest of them to find their way to your position.
d11cdc No.16371678
>>16349623
Can anyone explain how this is done? Basically someone put a 2d image on a plane for a 3d model and made it so no matter where you look at it, it stays the same. Also can you just make parts of textures transparent?
135c4f No.16371688
>>16371678
the technique is called "billboard"
google something like "billboard sprite" + the tools you are using and you will stumble on 300 tutorials
b34d4a No.16371913
>>16371678
It's a sprite that rotates to always look at the camera. Remember HL1? The electricity/ray segments did the same, following their path but rolling to stay perpendicular to the player's view.
d11cdc No.16371973
>>16371913
>>16371688
Ok got it, thanks anons.
b34d4a No.16372123
>>16371973
As for parts of textures being transparent, textures can have alpha channels, like a gif or a png does.
4408ed No.16374402
>Page 13
>only 149 replies
RIP
Where's the progress, guys? Personally I've been looking into various ways to realize the AI for our game.
I settled for GOAP (which is the AI the original FEAR pioneered) which will be used for low level decision making of the individual NPCs and using a finite state machine to guide squads of GOAP agents.
I hope to have something for demo day.
0b6640 No.16374430
>>16374402
I got inspired to finally work on my cross platform game engine which uses native apis instead of SDL. I have no visual progress to post beyond a window. The absolute state of native audio is fucking gay especially on linux. I see why sigmadev used openAL.
af9f5c No.16374439
Need a setting or general theme to get started with my 16 player driving game, because gamenights suck being comprised of FPS games and sourcemods.
286d75 No.16374459
>>16374439
How about *forklift trucks**?
4408ed No.16374463
>>16374459
>forklift trucks
40t forklifts. Absolute carnage.
846f0e No.16374467
>>16374402
>Where's the progress, guys?
Don't have any. Wrote a small bash script helper for OpenApoc fiddling, thats the only "progress" I have.
af9f5c No.16374469
>>16374459
>>16374463
>froklifts
Thats a great idea, i'll go make some forklift and character models
135c4f No.16374482
>>16374402
>Where's the progress, guys?
Doing interfaces, I wanna die
286d75 No.16374565
>>16374402
Since you're researching AI I'll go ahead and talk about the AI implementation I put in place yesterday.
Non-player entities are each given an AI object, and the AI is given the following functions:
>AITick: called by entityTick every tick with a delta, makes the AI actually analyse the situation and make decisions
>AIUpdateAnalog: called by entity behaviors so that the AI updates the analog aim and movement inputs
>AIUpdateAction: called by entity behaviors when checking for inputs so that the AI can push whatever action it wants to perform to the entity
The design requires the AIUpdate functions to be integrated into the behaviors where necessary, but also make it so that all entities can be controlled by either a player or an AI.
I'll later make some changes so that the AI is somewhat similar to the GOAP-model you mentioned. AITick will periodically (not every tick) decide a strategy, and based on that strategy the update functions have different behaviors. Examples of strategies are:
>staying out of melee range of target, possibly strafing
>engaging the target, attempting to enter within <x> range and then performing action <y>, possibly with a slight strafe rather than direct approach
>stand still, stare at target
>move to last location target was seen
I haven't really decided how to compose the world yet into levels (use pre-designed chunks, level generator, slab-chunks together in a tile-esque fashion, ???) so I can't get started on integrating pathfinding just yet. But that wont be necessary for demo day, as my focus is refining the combat into something enjoyable first.
4bf6b9 No.16374593
>>16374430
How are you planing to port the game? Really interested in the structure of your game and how it chooses which API to use. Also what was the reason, is it an exercise or is the performance of SDL shit?
4408ed No.16374634
Does anyone here know what the music in >>16374467 is called?
>>16374565
What I'm trying to figure out is how to handle movement in an elegant way. I need a way to ensure I can comfortably have my agents look in a specified direction while running, so they can engage while they're headed for cover.
What do you mean by update every tick? I hope you don't mean every frame, as that would be a bad solution.
4408ed No.16374636
>>16374634
For clarification: I mean the genre.
286d75 No.16374676
>>16374593
>>16374634
>What I'm trying to figure out is how to handle movement in an elegant way. I need a way to ensure I can comfortably have my agents look in a specified direction while running, so they can engage while they're headed for cover.
Generate a path, as well as a target position/entity to look at, and make your agents follow the path whilst looking at the target position?
>What do you mean by update every tick? I hope you don't mean every frame, as that would be a bad solution.
The update functions are not "smart" functions. They simply look at a few objects and update the entities' inputs accordingly.
For example, AIUpdateAnalog updates the movement input (wasd) and aim input (mouselook) based on the direction to the target entity is. It doesn't re-evaluate the situation, it simply turns the entity. Likewise, AIUpdateAction evaluates the distance to the target entity and pushes actions based on that.
AITick calls every tick, but basically just counts down a timer (which has a random time of 0.25 to 0.5s currently) which when it reaches zero actually does a proper update that involves evaluating the current target or looking for new targets to aggro on.
When implementing strategies, I intend to divide this update into two different types, a frequent tactical (0.25-0.5s) and infrequent strategical (2-5s). The strategical decides the current strategy, whilst the tactical makes it happen by engaging in pathfinding, deciding which action to use next and re-evaluating the path.
286d75 No.16374678
>>16374676
ops didn't mean to quote the first one
0b6640 No.16374685
>>16374593
It's very much an excercise. I use OpenGL, Xlib and WinAPI. I would have used xcb, but iirc it doesn't have a GLX equivalent. So far I have windows, grabbing input, and a timer which is enough for a cube test now that I think about it. I'm planning to do a Wolf3D style game with it so don't get your hopes up for extreme engine dev posts.
c3697c No.16374715
>>16374402
>Where's the progress, guys?
Shelved for academics. My school has strict grade requirements for their game development program (because you actually make games in it, I've seen them and they're not bad for student projects) so I've been hitting the books.
I also showed the degree chairman pics related and he assured me that no one in the major is taught to code like this, so that bodes well.
4bf6b9 No.16374738
>>16374685
Do they function similar to SDL or do you need to know more about the OS to be able to use it?
0b6640 No.16374786
>>16374738
Some things map one to one like swapping the buffers, others require more research like hiding the cursor on X11. It's not much more complex to use native libraries than it is to use SDL or GLFW.
4bf6b9 No.16374975
>>16374786
Thanks for the info, are you the guy on the last thread creating using only API's that had no motivation?
df9a5c No.16374989
>>16374402
I spend months redoing what I already have, and have had for the last several years, and still have nothing meaningful to show for it, that a competent person could program up in a week, tops. I call is Sisyphian Programming for obvious reasons
9b5edc No.16375008
>>16374989
I have the image saved
4408ed No.16375041
>>16374676
>Generate a path, as well as a target position/entity to look at, and make your agents follow the path whilst looking at the target position?
You misunderstood. I meant I'm unsure about how to integrate that procedure with the planning AI.
If you consider FEAR's FSM, you have a dedicated "Goto" state. There is a problem with that, because it separates going into position into its own action. For example, the agent might plan to go into cover, then take aim and then fire. Those actions happen sequentially and there is no builtin way to "layer" them to create an agent that is already aiming and firing while he's headed for cover.
One GOAP implementation for Unity I've looked at, handles that by making the target position a property of the GOAP action class. In it every action has a target position and a radius within which you have to be for the action to actually begin. I don't like that approach, because not every action needs to have a target position and it's also questionable how to handle actions that can be used on different types of NPCs with different movement modes.
One approach I thought of is to separate the movement logic out into its own "locomotion" component. The AI would no longer handle movement directly, but instead tell the locomotion component to go somewhere and use events to listen for success/failure inside the implementation of the other actions.
0b6640 No.16375072
>>16374975
I have no motivation, but I'm not that guy I just sit down and work through it so I can justify shitposting here.
>>16374989
I did this for years and regret it greatly.
4bf6b9 No.16375100
>>16375072
Good luck on your project anon.
>>16374989
I might need to do this later after the results of my SoA vs AoS performance tests.
286d75 No.16375118
>>16375041
>Those actions happen sequentially and there is no builtin way to "layer" them to create an agent that is already aiming and firing while he's headed for cover.
Have you tried adding a builtin way to layer them? There's no reason you can't add a bunch of variables to the "Goto" state that are used to create the desired behavior.
Your problem seems to be a "I want to use this design pattern but it doesn't meet my requirements.". The solution to that is to adjust your design pattern, use a different pattern, and possibly embrace the spaghetti.
Pic related. It's what good game code looks like. No matter how much of a design pattern purist you try to be, you cannot escape pasta.
>One approach I thought of is to separate the movement logic out into its own "locomotion" component. The AI would no longer handle movement directly, but instead tell the locomotion component to go somewhere and use events to listen for success/failure inside the implementation of the other actions.
I intend to do that as well in a certain sense, where the tactical layer would be tasked with finding a path to the destination, reporting back to the strategic layer if no path is found.
I have not considered making this into its own object within the tactical layer, and I don't intend to either until I make multiple AI's with different behaviors. I don't want to waste my time creating abstractions that don't benefit my project.
9b5edc No.16375202
Had a long weekend away for a festival, where I did a bunch of thinking about my Brainfuck thingamajig. The problem I was working on was variable jumping, where you move to the right X amount of times based on the value of a cell value. Originally I was thinking about how to implement two new symbols into the language to do so, but then I realized you can do it with pure brainfuck in a completely different manner: adding a bit of buffer memory to each pixel. Why I didn't think of that right away is a mystery, since I already had it for each scanline purely to be able to jump over the visual memory to the free memory easily.
So I added that, and while at it cleaned up my .ini file reader so you can only define the values that you want to be nonstandard. You can now set the screen width and height, a scale value for the screen, the total memory size, how much memory is before the visual memory, how much padding before each scanline and pixel, and I added a definition for external color files that will define the palette.
Now each "project" will have it's own .ini file, so developers can just hand out two or three files containing the code, settings, and palette if needed.
I've thought about making most of these variables defined in the brainfuck memory, but I'm not sure if it adds value or just needlessly complicates things, plus for screen width and height it's require a bunch of changes to my code because setting that shit up ain't easy. Palettes could also be defined by 256 * 3 memory cells. The only thing that wouldn't work is memory size because it requires values way over 256.
Now to get to a certain pixel, first set up the scanline numbers and pixel numbers. Then you copy two cells containing X and Y to the first or last scanline-padding memory, subtract the linenumber from the Y value in a separate cell, and if it's zero you do the same thing for X; otherwise move X and Y to the next scanline and repeat.
My main concern is if this is going to require so many instructions per pixel that naive interpretation is going to be too slow to write games with, so I might have to add a preprocessor that compiles the brainfuck into more optimized pseudobrainfuck to interpret, or even JIT-compile it completely. Multithreading can't save my ass here. But first I should finish up cleaning the code and add those palette files and then write my snake implementation so I can see if it's a problem in the first place.
Also, I think I'm going to settle on the name BrainFuckBoy for this project. It's a play on words on the Gameboy and the term fuckboy. Thoughts?
8167d4 No.16375204
>>16374402
my dreams are greater that anything i can handle and i don't want to settle for less so progress is basically dead unless i'm not a lazy shit
556384 No.16375403
Finished (sort of) default monster behaviour.
f16fe1 No.16375410
>>16375204
Something similar for me. I've spent weeks working on a model and whenever I have time to think, I realize that I'll need hundreds. Whenever I write a song, I remember that I'll need over 70. Whenever I implement a feature, I remember that there's fifty that I don't know where to even start with that I'll need very soon.
As long as I keep working, though, that's all that matters.
4408ed No.16375472
>>16375118
Trust me, Anon, I'm the last one you need to tell not try to follow patterns at any cost.
What I'm trying to do is figure out how to minimize the amount of redundancy without making things overly complicated. A "goto" action is simple in its implementation. However, if my goto action needs to handle every single combination of let's call them "modes", then it will increase in complexity. I will need to write custom code for every single case and that's quite prone to error.
I think the idea of a separate locomotion class is good, because it allows the AI to work its logic regardless of how the agent moves to the designated position. I'll have to sleep over this, but it seems like the way to go is to simply have more actions. There is no reason to have only one action that moves the agent. I'll just have a separate "attackMove" action and that's it.
>>16375204
>my dreams are greater that anything i can handle and i don't want to settle for less
I can relate to that perfectly and I'll give you one piece of advice on that. Read that post you made again and think carefully about whether or not you have a little bit of a hubris problem. Because by definition the fact that your dreams are greater than anything you can handle means that if you were to achieve your dreams, you wouldn't do them justice.
Personally, I like to be able to be proud of my work. So I decided that the games I'd love to make have to take a backseat while I gain experience elsewhere. As long as I'm not good enough to make those dream games of mine happen, they aren't really dreams but delusions. It may sound weird, but I believe I need to earn the right to have a crack at those dream games first. I don't think I have the right to ruin a great idea just because I happened to have it, if that makes any sense.
f16fe1 No.16375523
>>16375472
In most cases, it's not that one's dreams are unachievable. I highly doubt anon cannot achieve his dreams if he works hard, and I know I can one day complete what I set out to do. It's just that it's pure madness to even try because of the amount of time and effort required. But then again, can't that be said about any game that you feel true passion for?
9b5edc No.16375562
>>16375472
I think you're going about it wrong, goto always only sets an intended position and then calls animate, with the walk animation. The point of FEAR's system is that Animate decides other things like spawning bullets during the fire animation, but this includes the actual moving. You can just have a walkandfire animation, run animation, and walk animation and decide which one to use at the time you call goto, or add interrupts to switch between them based on line of sight and the like. You need to have those anyway, since it's stupid for the enemy to just keep walking to a point when they see you.
>>16375410
>>16375523
Most things become easier as you go on. Music becomes faster when you settle into a style and get one or more leitmotifs for your game going. Art tends to get faster as you get a better grasp on the workflow for your artstyle, plus not every object requires as much time as for instance important NPCs or setpieces. Programming new features can become faster as long as you design your codebase well enough and/or figure out ways to reuse functionality.
In the end, how much time things take is a matter of skill, and presumably your skill level should only go up. Unless you go to Calarts
4408ed No.16375675
>>16375562
Thanks for the clarification, but I'm not married to FEAR's exact GOAP implementation. I don't intend to squish everything into three states, because I don't have to. There also some limitations to FEAR's approach that were later addressed by better AI systems like HTN. I would use HTN, but the only readily available solution for Unity is from literally Tencent, so that's a hard pass.
The way I'm intending to use GOAP kinda meets HTN halfway and gives me some of the benefits of that system over vanilla GOAP.
830d11 No.16375689
It's been four years /v/, where are the games?
901859 No.16375694
8167d4 No.16375696
>>16375689
in the wiki, my portfolio or lost in time
c571e2 No.16375715
>>16375689
I don't understand why anyone would post their actual games here. It's like a guaranteed way to get it flagged by sjws and kikes and blacklisted. It's no small wonder there are so few replies to this thread or quality finished games to come from here.
9b5edc No.16375763
>>16375715
You say that as if being blacklisted by those people matters. Yandev got tons of money while posting here, the dev from LISA came here for an AMA, Doug Tenapel came here, Yoko Taro, American McGee, and all of those did well afterwards and never really got negative exposure over it anyway. Only AAA companies might generate enough asshurt to get a reaction, but we only have a samplesize of one there.
bb00ad No.16375806
>>16375694
>speebot dev also made a godawful shmup
wew
11518a No.16375812
>>16375694
>4 years = 6 games
It's a pretty good ratio
286d75 No.16375821
>>16375472
That's just a weigh-off between redundancy from creating multiple actions that do 90% the same shit and the complexity of creating a single action that covers multiple behaviors. You'll have to decide that for yourself depending on your project requirements.
>>16375689
When it's done.
>>16375715
Given that I want to create a game that's challenging, I don't really see why that would matter. There's zero overlap between my target audience and the people who listen to those idiots.
830d11 No.16375952
>>16375821
>When it's done.
9b5edc No.16375984
>>16375952
At least there's a demoday in less than three weeks, there should be a handful of demos to play. A lot of them may be very early in development, but some of them might already be fun to play around with. I know I've had fun with some of the earlier demos.
15b2a4 No.16376110
>>16375812
I've made some shit that's not going to end up on a wiki, for obvious reasons.
fbc2b3 No.16378804
>>16377746
Neat. Mind explaining what's going on here?
df9a5c No.16379363
>>16375008
Thanks, I hate it
0b6640 No.16379919
>>16379363
just don't obsessively reimplement things that work
df9a5c No.16379925
>>16379919
I think I enjoy learning more than creating
0b6640 No.16379977
>>16379925
Then think about it as learning how to finish something
adf63f No.16380568
>>16378804
Experimental spaceship building.
fbc2b3 No.16380621
>>16380568
Oh hey, you're the gondola dev! I like the space ship so far.
512b87 No.16380627
>>16375008
Is there a name for when you obsess over theory and how things should be implemented without ever actually fucking writing anything.
df9a5c No.16380645
135c4f No.16380647
>>16380627
Autism, I got that too
e99ee6 No.16381481
8167d4 No.16381757
messing with the combat today
finally have a use for the subscript node
43d1b3 No.16381841
>>16381758
Very neat anon.
c3697c No.16382265
>>16379925
You learn something whenever you create something. Especially if you finish it because you'd be amazed how much work is done in the "last 10%" of a project.
014440 No.16382439
>>16380627
I always have this for a long time, until I force myself to sit down and write it, at which point I go into autistic focus mode and don't realize anything else going on around me for at least a few hours.
8167d4 No.16382546
>power to the whole building goes on and for half the day
it's not like this is the one day of the week i feel like deving or anything
b34d4a No.16382636
>>16382546
Life is the hardest game. What are you, a journalist?
8167d4 No.16382640
>>16382636
i gamedev profesionally and as a hobby, hence why weekends are for hobbydev
7df121 No.16382646
>>16380627
the proper term is 'analysis paralysis'
0b6640 No.16382650
>>16382640
Is professional gamedev as bad as people say? If so, why do you do it?
8167d4 No.16382666
>>16382650
the company i'm in is pretty small so it's fine, hours and pay are good, we're making actual games designed to be fun, stuff gets done on time, so no stress and i'm happy with everything
i'd imagine that it's way more stressful for the AAA companies because of the crazy deadlines
f6d6d3 No.16382733
Does this count as vidoegames
836c39 No.16382751
>>16382733
From Dust prequel looking solid anon
6d975a No.16382759
>>16382733
9/10 meditative experience - Polygon
e99ee6 No.16382788
>>16382733
You've got a mechanic, now add a goal and win/lose condition.
f6851f No.16382815
>>16382733
Literally Powder Game.
014440 No.16382866
YouTube embed. Click thumbnail to play.
>>16382733
Not yet, but if you hurry up you can make one out of it before these guys finish theirs.
e81675 No.16382873
>>16382733
I already played this on Newgrounds a decade ago. I think it was literally "Sandbox".
ed3151 No.16382899
>>16350075
kek'd but then died inside
104afd No.16383997
Anyone know a good place to find contract modellers/animators? I know I should just learn myself but I need something to work with until then.
347159 No.16384038
>>16383997
Gumroad, Sketchfab or the Blender forums would be a start I'd say.
8167d4 No.16386180
YouTube embed. Click thumbnail to play.
i think i can successfuly rip off bionicle, i don't think lego would care enough to sue me
6b0424 No.16386372
File: f8174fa725b0a6d⋯.png (Spoiler Image, 660.21 KB, 757x1208, 757:1208, 9_year_old_drag_queen.png)
>>16386155
i hope its not one of those 10yo drags
f26ea6 No.16386405
>>16386180
>Video
Congratulations, you have successfully replicated those fucking annoying dogs from Dark Souls. Now burn this abomination
df9a5c No.16386440
>>16382733
How would you check behaviors on adjacent dots without brute forcing it?
f6d6d3 No.16386568
>>16386440
>brute forcing it
I'm not sure what that means. In any case the way it works is kind of esoteric.
When you add a pixel or an emitter, it uses it's y position to expand a min/max vertical range on the current column. Every tick, all pixels in the min/max range of all columns are checked from bottom up. The alpha value of the pixel determines what to do with it, "emitter" pixel draws a sand pixel below it on regular intervals, while sand pixels just fall down or check left/right if there's already something below, and both will mark their position on the column as well so they get checked on the next tick.
b0897e No.16386576
I've tried to program basic things in languages like C, Python and even BASIC, but I fucking hate it. I'd really like to be able to make my own games from scratch, but coding makes me angry and bored. My code never works and I get pissed off and delete everything.
Is this just not for me or is there a way I can overcome my frustration? I've been trying since I was at least 14 but I've never got even basic things working.
I have some anger-related mental issues, by the way.
df9a5c No.16386595
>>16386568
What would you do in the case where there's two sets of pixels in the column? Eg, air, sand, air, sand, air?
0f29cf No.16386596
Finally forced myself to code.
> Implemented matrix representation of gaming field.
Figured that round in GDScript round actually returns int and not float as I thought.
Now I can start looking for areas occupied by player.
Nothing much, but still progress.
8167d4 No.16386603
>>16386576
stop being buttblasted and git gud
66e18a No.16386613
>>16386576
If you've got an anger disorder so severe that you wipe all your progress over minor frustrations before you get anything done, then I have to sadly say you should give up this hobby entirely and focus your efforts elsewhere. I'll do you no good to waste your time on something you cannot enjoy.
>>16386603
>stop having a mental illness xd
That's not how it works, chief.
014440 No.16386618
>>16386576
Set up version control so when you delete everything, you can get it back anyway :^)
A programmer pretty much has to enjoy butting their head against a problem for the dopamine rush when they figure it out. You're probably better off learning everything else related to making vidya and finding a programmer to work with.
f6d6d3 No.16386627
>>16386595
By min/max I meant that the highest and lowest positions are marked, and then everything in-between is checked. So if there's a pixel at the top and another at the bottom, all the pixels in the column would be checked regardless of what's inbetween.
268146 No.16386729
>>16386568
Do you do the calculations sequentially for each sand pixel? CPU or GPU? This looks like the kind of thing that should be done using a compute shader.
846f0e No.16386803
Well I made my bash scripts for OpenApoc ready for publishing, its not much but it should be enough to get anybody else interrested on this engine a bit more prepared. The big downside is that I only tested it on Linux Mint so I dunno how it works with other system or to make it even 100% POSIX compitable which I will probably do it later as there is another Issue I need to clear before making my never ever mod a tad bigger such as figuring out how to force reloading the values on savegames which doesn't happen so far for me, only at new game. Also I documented the scripts as much as I can its probably a bit too verbose for some other but as long you can understand the intended flow it should be good enough for (you) I suppose.
Package contains (remove the .pdf suffix):
generate_crc.sh, generates a OpenApoc checksum.xml file which uses crc
build_mod.sh, the simplest of them all it just zips ups your folder and its naming depend on the variable as defined on bashalias.txt
pretty_print_xml.sh that one relies on a external program called xml_pp which as it name says it makes OpenApoc xml file less ugly to stare at, it is relying on input and output folder variables as defined on bashalias.txt
f6d6d3 No.16386828
>>16386729
CPU, I just draw a pixel onto the window when one needs to change. The whole thing is around 150 lines of code excluding my framework for input and window and such, and I didn't feel the need to make it more complicated. Here's rough pseudocode:
for (each column = x)
for (column[x].ymax -> column[x].ymin = y)
if (pixel(x,y) == emitter)
if (interval == 0 && pixel(x,y+1) === empty)
drawPixel(x, y+1, sandwithemittercolor)
column[x].ymin = min(column[x].ymin, y)
column[x].ymax = min(column[x].ymin, y)
else if (pixel(x,y) == sand)
if (pixel(x,y+1) == empty)
dest_y = y+1
dest_x = x
else if (pixel(x-1,y+1) == empty)
dest_y = y+1
dest_x = x-1
else if (pixel(x+1,y+1) == empty)
dest_y = y+1
dest_x = x+1
if (dest_y != y && dest_x != x)
drawPixel(x, y, emptypixeldata)
drawPixel(dest_x, dest_y, currentpixeldata)
column[dest_x].ymin = min(column[dest_x].ymin, dest_y)
column[dest_x].ymax = min(column[dest_x].ymin, dest_y)
df9a5c No.16386838
What are some good resources/libraries for freely usable sound effects? How much would a cheap foley artist charge for different sound effects? Is it better to get a decent microphone for my laptop and record shit myself?
ec1a47 No.16386847
>>16386838
what kind of game are you making?
df9a5c No.16386851
>>16386847
Maybe a Roguelike, something simple but growable to keep me busy
ec1a47 No.16386896
>>16386851
Record the samples yourself to make it unique and personal in any case if you're not making something for mass appeal.
df9a5c No.16387700
Any progress is good progress, r-right?
As for today's stuff, I worked on top of some shit that SFML did. SFML current has utility classes that do some work in a friendly, but less than fast way. Eg, there's Transform which is a 3x3 matrix (XY transforms), but offers no methods to directly access the 9 fields inside it. There's a get/set reference class on top of it called Transformable which contains a Transform and friendly things like Position, Scale, etc, but it's cumbersome to use.
Instead, I implemented a DrawHelper class to create the Transform class directly. It's at least at fast as their approach, too.
df9a5c No.16387931
>>16387700
And for some reason, when i=0, SFML thinks it should be (-1024, -924), which I suppose makes sense, but when I used XNA, it only really affected the XY offset from the top left of the actual drawn sprite, at the end, not like a global offset
bd2cd5 No.16388097
What books about game design and development would you recommend?
Something general, not too specific.
df9a5c No.16388184
>>16387931
Nevermind, it was how Scale was applied to position that made it seem weird.
fbc2b3 No.16388212
>>16388097
The Design of Everyday Things
Structure and Interpretation of Computer Programs
Game Programming Patterns by Robert Nystrom
First two aren't about games specifically, but they are still very helpful in the area. I have the first and last as a PDF I can share, but I have the second only as a physical copy so I can't share it.
More importantly than reading books about game design, play games with a critical eye paying close attention to how certain things are accomplished and spend as much time as you can spare making games. Another good source of information is developer commentary and postmortems about games you enjoy, though that most often takes a form that's more specific than what you asked for.
df9a5c No.16388259
No, that's not working either. What I want to do is write commands the way XNA did it:
>SpriteBatch.Draw(x, y, w, h, texture, color, source, rotation, scale, origin, depth);
For some reason, the way SFML is interpreting matrix scaling is being applied to origin too. This means I have to do stupid shit like apply a scaling factor to my origin, whereas with XNA, I'd just have to have a generic origin value for the texture source coordinates, not the drawn ones, this is fucking retarded
f0a17b No.16388281
>>16355123
Stepped away for a while, was upset that the player rotation was only working independent of forward velocity in very specific cases. Now works always. Plus movement axis locking, as well as the player slowing upon turn (inspired by Solatorobo for platforming accuracy). Next I have to build the player pause menu for abilities, inventory, squad commands, hacking terminal, and options menu. To do this, I must rework GUI tilesets write some placeholder script tonight. Skill system and inventory both already work.
Also some older attempts for what the new sprites will look like, I did these in maybe an hour or two so this wont take a whole year this time (drawing without a tablet is dreadful). SPRITES ARE NOT A FOCUS RIGHT NOW, I am working on getting actual game-play at this point. The sprites will be worked on when everything else for the demo is built.
cbeeb7 No.16388303
One of you dev fags need to make a 4 player side scrolling beat em up with the combat depth of Devil May Cry. Essentially CUHRAYZEE Double Dragon.
434497 No.16388311
>>16382733
Interesting. Please continue.
cbeeb7 No.16388333
>>16382866 (checked)
terraria was a mistake
580c33 No.16388335
Uni starts soon so I finally have an excuse to not make any progress towards my project, whooo.
Wish I had proper work ethic.
/blog
b34d4a No.16389363
>>16386596
>var x:int
man, GDScript looks a lot like Actionscript.
da63d2 No.16389384
>>16350792
>reverse google this to get more cute robutts and binary paintings
>booru page is full of centaur & spider porn
https://danbooru.donmai.us/posts?page=6&tags=ze_%28sawakihein%29
014440 No.16389474
>>16388335
>finish uni recently
>internship "fucked up" my sleeping rhythm to the point where I fall asleep around midnight and wake up at 8 AM unless I force myself REALLY hard some would call this fixing your sleep rhythm, but whatever
>started my recent project and get so invested in it that I haven't felt like playing vidya
>days consist of some jobhunting and related shit, programming, and a bit of shitposting
I expected to turn into a lazy faggot the moment I got "infinite free time" yet somehow became more productive than when I was supposed to be working on my education.
/blog
0f29cf No.16391312
> Implemented searching for areas occupied by player
Quality of code has dropped severely, but I just want to finish this project and take new one. Already have ideas. Still need to take into account only valid areas. And for that I need to define what "valid area" is. Obviously, it shouldn't be possible to build an L block and lock whole map. Then I need to actually block opponent from building in occupied zones.
Then I need to implement removal of opponent's pieces from the board.
0f29cf No.16391548
>>16389363
Never programmed on it, so I'll take your word on it. When I came to programming flash was dead already. AFAIK ActionScript is closer to JavaScript and GDScript is trying to look like python. But most of modern languages looks the same. Which is a good thing, I guess.
2c5658 No.16391704
>>16388281
>have only pixel-shit sprites and zero code
Oh no.
On the plus side I'm not an indie dev or a furry; on the minus that game I have in mind's never getting made.
f0a17b No.16391980
>>16388281
Have some tasty blood emitters. Eventually these are planned to leave blood splat decals. Kill barriers also work. I have to rework the sprite manager and get physics for these gibs working.
>>16391704
Elaborate pls. Maybe I'm dumb but what are you talking about?
390fa6 No.16391994
>>16374715
>I also showed the degree chairman pics related and he assured me that no one in the major is taught to code like this, so that bodes well.
That's because Yandev doesn't have any formal programming knowledge
bd2cd5 No.16392164
>>16388303
I actually wanted to make something similar. Sidescrolling game with something akin to fighting game combat. But it is a big investment and Im still not confident enough to start it. Maybe in a few years yes.
1304fb No.16392192
>>16392164
So like Mortal Kombat: Mythologies: Sub-Zero?
bd2cd5 No.16392222
>>16392192
Yeah something like that only good.
f0a17b No.16392670
>>16391980
Didn't think this was getting done today. Gibs have physics, next is some more player stuff and more work for the Terminal (pause menu).
61e569 No.16393630
>>16392670
>>16391980
>>16388281
Lurker here, I'm unironically hyped for this since this isn't a paywalled porn game nor cliche gay vn. You're bound to get a following if you finish this as people are longing for cleaner stuff. It's a step forward.
423dbf No.16393695
>>16393630
Thanks for the support m8, video and an axis-locked platformer demo soon. Have some test walk cycles from earlier this month. they are bad lol
130a22 No.16393797
>>16393630
>>16393695
Ditto. I'm sick of animu shit and this is one of the few projects I like to see updates for
b34d4a No.16393807
>>16393695
I never reply but I get glad every time you post progress. I don't even know what the core gameplay is, but I like it.
854357 No.16393940
Blending IK with some logic to make her point to places dynamically. Mainly to prevent her arms from pointing behind her back in a broken way.
423dbf No.16394098
>>16393695
This is what the debug room is planned to be turned into. I have to make maybe 3 or 4 meshes and knock out what is on this markerboard. Sage for ideaguy post.
0c98d5 No.16394175
Suggestions for tools/programs to use for music and sound effects?
Going for an NES aesthetic but I'm not good at using the actual NES music thing, I prefer something like FL where I can just place notes on a scrolling time sheet thing.
Freeware/etc preferred.
014440 No.16394198
>>16394175
sfxr or bfxr for sound effects, and just use FL or something similar with a chiptune VST for music. LMMS is like FL but free as in freedom, but might be buggy on Wangblows from what I heard.
I don't know shit about making music and sound, so I have just about zero experience using any of the above except making some bleepy jump sounds for gamejams sometimes.
76ea51 No.16394199
Does anyone have any tips for coming up with decent ideas for short games? I kinda want to poke at being an all rounder and doing music, art, etc, so the "make your dream game starting from having no skills" path would be a sisyphean task. At the same time I struggle to think of anything that fits the bill when it comes to being small but not being crap.
0c98d5 No.16394203
>>16394199
Try giving yourself some limitations and see what gameplay you can come up with to fit them.
I did this with the idea of making a phone game with only three buttons (left, right, and jump) and making a very simple to learn, hard to master platformer.
dbec07 No.16394221
130a22 No.16394351
>>16393940
but why does she want you out?
135c4f No.16394470
>>16393940
yare yare da tetekeike
c3697c No.16394669
Has anyone found these demos useful for learning Godot?
c3697c No.16394671
>>16391994
Really? I thought he was a college dropout.
f16fe1 No.16394700
>>16391994
Hey, I have no coding experience either but I can usually gauge if something should be possible to optimize, mostly down to gut feeling and assuming that everyone else who does this is much smarter than I am. It's not always the case but most usually is. At the very least, I know what a god damn switch statement is.
fa7580 No.16394787
>>16389384
Sorry anon, cute robots are hard to find. Here's some extra for you.
64e50a No.16394959
>>16394700
I don't have any formal programming knowledge either, outside of high school, and I learnt about switch statements watching youtube tutorials on game maker.
There's being self-taught, and there's being ignorant. At least read the documentation.
f560a2 No.16395065
>check /cuckquean/ for my bimonthly kek
>they're making a fucking RPG now
014440 No.16395181
>>16394700
>>16394959
You can't use switch statements on strings Aparently you can use switch statements on strings in C#, and I'm fairly sure switch statements compile down to pretty much the same thing as a string of if-else in C# anyway.
The actual problem is the fact that he's using fucking strings to control state
He should have an enum reactioncauses { WEAPON = 1, BLOOD = 2, INSANITY = 4, LEWD = 8 } and make this.Witnessed an int that gets OR'd with those values instead of creating the long strings. And then use a fucking switch statement.
Reason being that comparing a string is simply slower than comparing an int.
I once had a nigger argue that that isn't true for C#, but I scoured the .NET documentation and found that it is, this particular function winds up being used in the case in Yandev's code https://referencesource.microsoft.com/#mscorlib/system/string.cs,8711fff131bc4d0e it does have an early-out which makes it a little better, but that actually doesn't matter for his overly similar strings, and fuck you it's like a hundred times slower than an int compare
01260c No.16395219
>>16395065
The fuck is this shit even about playing as a dumb cunt and your boyfriend fucks everyone and everything?
64e50a No.16395221
>>16395181
>the fact that he's using fucking strings to control state
AHAHAHAHAHAHAHAHA
f560a2 No.16395255
>>16395219
>boyfriend
They're her childhood friends and yes, they fuck everyone but her.
01260c No.16395263
>>16395255
kek what the fuck why though?
Why not just play as one of those dudes fucking all the grills AND her?
b34d4a No.16395301
>>16394700
>>16394959
>>16395181
>>16395221
The guy might be a fool, a lolcow and whatever, but I can never hate him. For every person who gets into gamedev, there must be thousands who forever dream about it and never dare to try.
959f29 No.16395312
>>16395181
>strings for state control
Fucking lol
6ea48a No.16395321
>>16393940
Shinobu is thicc
286d75 No.16395407
>>16395392
The bump limit isn't 300.
1a0d3d No.16395410
11518a No.16395746
df9a5c No.16395809
>>16395181
Why wouldn't you be able to use switch statements on strings? They compile down to a chain of if/else statements that check exact values, that's all.
0b6640 No.16395842
>>16393940
>sound effect on the wrong side
What's up with that, did you mirror the video?
11518a No.16395855
>>16395809
You can, but that anon is right that using enums would be a far superior method to what cuckdev is doing.
286d75 No.16395858
>>16395809
Because a switch statement is an odd thing that is supposed to mimic the jump table behavior from assembly. In C, you can only toss integers into it. This is also why switch cases "spill over" in many languages if you don't put a "break;" at the end.
Also I'd assume it would compile it to some more efficient data structure to search through. Very little reason not to.
130a22 No.16395862
>>16395842
don't you see the vr hands near the vr menu?
014440 No.16395864
>>16395809
I'm mainly a C++ programmer. Switch statements there only accept, and I quote, "any expression of integral or enumeration type, or of a class type contextually implicitly convertible to an integral or enumeration type, or a declaration of a single non-array variable of such type with a brace-or-equals initializer"
Which tl;dr means you can't use std::string or a char array.
For instance when writing file parsing, if I want to have a switch statement for string identifiers, I have to set up a enum and a map<string, that_enum> then use switch(that_map[the_string])
>>16395858
Hell, in C/C++ it doesn't merely mimic jump table behavior, the compiler actually just compiles switch statements down to jump tables.
ffde50 No.16395878
I come back years later and there's still circlejerking over one fag's failure to use a switch statement. Is it really that amusing?
014440 No.16395887
>>16395878
First of all, yes it is that amusing.
Secondly, I already explained that there's a lot more wrong with his code than just the lack of switch statements.
286d75 No.16395897
>>16395864
>Hell, in C/C++ it doesn't merely mimic jump table behavior, the compiler actually just compiles switch statements down to jump tables.
Not necessarily. Some caveats apply, like safety checks and optimizations.
If you want a raw jump table, you can use the GCC extension that allows you to fetch the address of a label. https://eli.thegreenplace.net/2012/07/12/computed-goto-for-efficient-dispatch-tables
df9a5c No.16395902
>>16395897
>>16395864
But the compiler in C# will unroll the switch statement literally into a series of if/else statements, which will resolve into jump statements on the CIL(?) anyways, so it's not like the language implementation is not faithful to how C/++ does it anyways.
Yandev's just a retard.
713114 No.16395954
>>16395263
I think you're missing the point of cuckqeaun.
Also why is it not spelled "cuckqueen"?
f560a2 No.16395967
>>16395954
That's what the fetish is called. Why? I have no idea.
ffde50 No.16396558
>>16395887
It's pitiable that you find it worthwhile multiple years of repeatedly commenting "use a switch statement", "use an enum".
db9f53 No.16396592
>>16395954
>>16395967
Why would it be queen? That's silly. A Quean is a lowly woman.
b34d4a No.16396844
>>16396558
Some archetypes stay forever. The memes run deep.
014440 No.16398401
>>16396558
And it's pitiable that schools keep warning kids not to do meth, yet you didn't listen. I merely explain to people new at programming why the code posted by someone else is bad, so they can learn.
That it happens to be fun to laugh at the guy who wrote said code, because he's a faggot who whined at us for not licking his ass enough, and then ran off to reddit, only to then lose his pretty much golden publisher deal where he had someone take his nigh-unusable code that takes 15 minutes to compile and turn it into something that compiles in seconds, runs infinitely better, etc., lost over half his patreon gibs, and keeps whining at his fans for daring to have expectations of him, is just an added bonus.
cd95a1 No.16398696
>>16394098
Started work on menus for the console (basically pause menu stuff) tonight, slept all day yesterday since I was up from 6am-9am the next day. I have more work to do than I thought. I want to do something like UWindows for UT99 for the main menu but I have to tear apart some code and rework some sprites. Off work soon, shall post progress later.
cd95a1 No.16398721
>>16398696
whoops forgot example image
846f0e No.16398781
>>16398721
>>16398696
>UT99 GUI
You are mein neger. When are you going to imitate UT99 BSP shader/lightning? I should play a tad more UT99 to make more screenshot of comfy looking maps
cd95a1 No.16398869
>>16398781
I will attempt to get something like UT's lighting but Godot lighting is really bad. I have to use light layers just to make light not pass through walls and light a floor on the other side. That said, fun fact: Godot has CSG so I will test it at some point.
779776 No.16399265
>>16398869
Have you tried baked lighting?
d0fb6e No.16399296
>>16399265
Let me get some tutorials on that. If it involves probes, wont happen. Trying to make this game on a toaster for toasters.
f32c3e No.16400720
Shmup anon here.
Slowly getting back into the project after some weird cold/flu that keep on going and coming back.
I did some basic declaration for weapon projectiles, it should scale up without much problem.
Then checking on the hit spark/particles generation part, I will need to add a model resolution since right now stack '0' will render the first model loaded. So it's declaring stack 'n' is associated to model 'm' and adjusting the rendering loop.
Later on I will need to improve the memory management so I can more easily add stack inside a memory slice (from the main malloc)
135c4f No.16400726
Finally something worth posting
Now you can actually start and finish the cooking process
Items get spent, results goes into inventory
Now all that remains for 5/5 is a whole lot of bug fix and UI improvement, like searching items by tags and whatnot, making items not vanish when you scroll inventory too fast, etc
(and maybe testing if all of the item effects are actually working)
df9a5c No.16400816
Any resources about writing a rules-based puzzle game?
df9a5c No.16401059
Why do axis-aligned rotations fuck up when my texture coordinates are integer, but I set a 0.5f offset to the TARGET origin, it bleeds the texture? That doesn't make sense.
SFML on right, mine on left. Both transform are being created equally
803a40 No.16401181
>>16350792
It really depends on the game, what do you plan on making?
033f34 No.16401300
>>16398869
You can add filters to your camera, experiment with them.
7e958c No.16401775
>>16398721
Have an input issue I have to work around for this to have a shortcut tied to it, but have an image test for the main menu window. I plan to have scrolling text on the bottom bar and options on the top bar. More tomorrow, running out of time and woke up late again.
8167d4 No.16401832
>>16400726
>literally nothing on screen
>20-40 fps
b34d4a No.16402046
>>16401832
functionality first, optimization second, i guess.
<also unity.
cbeeb7 No.16402198
>>16392222 (checked)
How about instead of something like that. You make what I said, a side scrolling beat-em up with a DMC/Bayonetta combo system with a score for clearing the level that ranks how stylish you played. Streets of Rage but with way more depth to the combat and to the enemy verity. Essentially just Bayonetta but at a traditional Beat-em up vantage point instead of the 3rd Person point of view. Some games came close to breaking out of the mold like Shadow over Mystara and Dragon's Crown but none of them hold a candle to DMC. Just imagine what Kamiya would make if he was limited to late 80's early 90's tech and make that.
9b5edc No.16402309
>>16402198
How about if you want a game to be made, you do so instead of begging others to do so. These are threads for people making games, not fags who want to have games made for them.
286d75 No.16402495
>>16401059
I'm not sure what exactly you're doing, but I'll go ahead and make a guess.
You're rendering a bunch of quads which have that checkerboard texture of 16 different colors.
For this, you create a tiny 4x4 texture, and to not make it an ugly blurry mess, you turned on nearest texture filtering.
The problem you're running into is that nearest texture filtering is a lot less reliable than what you think it is.
Let's say you draw a quad (4x4), and perfectly align the vertices with the grid where 4 pixels meet. If you do this, the actual fragments you draw have coordinates that are half a pixel offset from the actual vertices you draw. This is ideal, since when sampling textures, the actual texels you sample also are half a texel offset. In a 4x4 texture, texel 0,0 is situated at UV 0.125,0.125, not at UV 0,0.
But if you move the grid half a pixel, so that instead of vertices lining up where 4 pixels meet, they're situated at the center of the pixels, you start to see artifacts.
In this scenario, the texture is sampled with UV coordinates that perfectly aligns the corner of 4 texels. (e.g. 0.5,0.5) If the math you do is not 100% accurate which FPU math, especially on GPU's, never is, then the tiny inaccuracies will make pose a predicament for nearest texture filtering. Pixels that are horizontally or vertically aligned might have extremely small differences in UV coordinates, and if that UV coordinate is on the boundary of two different texels in a nearest-filtered texture, they will sample different texels. This is what you are perceiving as artifacts.
Explaining this is kinda a pain in the ass, so I got pictures for you too. Each square is a pixel, and I'm drawing a beautiful 2x2 quad on top. The purple points are the positions of fragments, and also the positions where the quad texture is being sampled.
Pic1 is the ideal case. I'm sampling nicely in the center of quad texture texels. It grabs the right texel as expected.
Pic2 is the shit case. My sampling is exactly on the border of quad texture textels. At a glance, and thanks to my terrible art skills, you can't even tell which texels I'm sampling! Is the bottom right fragment sampling the blue, red, yellow or green pixel? If texture wrapping is on, would the top left fragment even be blue? Since I'm on the border, even a change as small as 1.0/4,194,304 UV can cause me to fragment a different sample entirely! God forbid I apply transformations like rotation, scaling, shears, or any of the other stuff that isn't perfectly accurate.
My recommendation is to look at this statement you made:
>my texture coordinates are integer
And complete erase the idea of "integers" in the context of GPU FPU math. It will never work the way you want to. Everything is float.
If you want the artifacts in your demo to go away, then simply don't draw the quad at a subpixel position. Calculate the pixel coordinates, and make sure your vertices are always at the point where 4 pixels align. And also, if your viewport is of uneven size, keep in mind that 0,0 is not placed at such a position.
135c4f No.16402734
>>16401832
60~% of it is Unity's editor, inspector window, console window, project window, etc
If I set the game window to fullscreen it goes to 60fps.
cad239 No.16403204
>>16402507
Does she curl her toes when you kiss her cheek?
cc4d36 No.16403802
>can't get excited about making a game that's potentially profitable and doable in a reasonable time frame
>get somewhat excited about making a game that'd probably be a flop, would take years to make and would require skills I do not posses
>get so excited about making some feature or a tool completely unrelated to any of my game ideas that I stay up late and wake up early to finish it and later discard it because it's useless to me
Kill me. **Can you make a game from a patchwork of completely unrelated features that wouldn't be shit?
I though I might have a change of making it and earning enough cash to provide for my parents when they get old, but it seems gamedev won't help me in this situation. Sage for blogpost.**
b34d4a No.16403806
>>16402530
Is this a star trek boarding/killing game?
cc4d36 No.16403807
>>16403802
And I fucked up a spoiler.
cbeeb7 No.16404012
>>16402309
>make it yourself
Because I am smart enough to realize there are better things I could do with my time and it would be better to delegate my excellent game idea to someone more competent. They would be better suited to carry out this vision in terms of comparative advantage. After all time is the most valuable commodity and I ain't got time for learning a new trade. It's called division of labor you big commie dumb dumb. Also there is already enough trash indie shitz from unqualified developers who think they can DYI games, do you really want more of that trash?
901859 No.16404035
>>16404012
>division of labor
The whole premise of this post is that you are making some kind of contribution with this "idea". You aren't.
>make a game that is like X other game but with Y mechanic/twist
>make a game at a level of work that X successful developer could have made
Do you realize how dumb and vague this is? Literally everyone who has thought of making a game has gone through this line of thinking before.
0f29cf No.16404092
Now I forbid opponent from building in occupied territories. Need to add ability to remove building from occupied territories. And actually determine win condition. But I'm fucking drained for some reason.
>>16404012
Fuck off, nigger. And never ever in your life tell anyone what to do without giving good reason to.
135c4f No.16404095
286d75 No.16404108
>>16404012
>smart enough to realize there's better things to do with his time
>not smart enough to realize his idea is worthless "X but with Y"-tier garbage
>wants other people to make his idea into reality for him
>calls them commies when they refuse
6ea48a No.16404119
Why are you fags taking >>16404012 so seriously? It's obvious he's joking.
f16fe1 No.16404128
>>>/v/16404012
So the anons who do invest the significant amount of time required to learn not just one craft, but several (consider that you need programming, music composition and either 2D pixel art or 3D modeling) shouldn't use that to make the games they want, but rather the game you want because your time is better spent elsewhere? What the hell are you doing that's so much more important than what any other anon is doing? If an anon was to learn how to make a delicious cake, should he not make the cake for himself and those close to it? Do you deserve that cake because you have better things to do than learn how to make a cake? You say that it's "division of labor" but a general idea is labor in the barest sense. I've had more elaborate game ideas while taking a shit, while showering and dreamed up in my sleep. Five minutes of effort on your behalf, five years on another anon, how is that even close to "division of labor"? That's the most jewish thing I've heard all week. And you call other people commies?
As a sidenote, your game idea is nothing special and I almost guarantee it already exists. Especially considering the fact that it's just "A game combined with B game". Skyrim with guns was laughed at for a reason.
0f29cf No.16404720
I have next to 0 knowledge of C++. But I really want to revive abandoned project. It had CVS as version control. I've menaged to find the client and download it. But it won't compile, because I have Mac. I know it's gay. How to most efficiently remove platform dependent code and revive the monster?
846f0e No.16404747
>>16398869
I see. Well good luck I hope you can find a solution to your problem, can't wait to see your new rendering stuff.
Are any of you modeller anons experienced with making a template for isometric rendering? Because OpenApoc is in need of a template. I wrote with a few of those developers via IRC and they are not capable at all of getting the OG template that was used back then to develop the game either the information is lost in time or possible held hostage due to some licensing bullshit. Having a template at least will help with further modding of this game and allows other anons and anybody else really to make more stuff for this engine. Obviously it won't pop up a freedoom like project for OpenApoc overnight but it is at least one step closer to it which is better than nothing. Such a template exist but its for C&C series rendering only forums.revora.net/topic/97398-blender-templates-tdra-ts-ra2/ . Also I refined the bash scripts a tad more but its not much and there is some few things I have to do before I push the commits at my forked repository, such as cleaning up the variables a bit, improving the formatting and making it less prone for spontaneous behavior. Making new weapons is pretty easy minus the weapon sprites (held by agents) because I don't know yet how. Before I bother making even more items I want also to try making a intermediate format and writing a bydlon script for it to make it easier to edit, perhaps a support for Libreoffice calc can be done too. The point for a intermediate format is because I don't want all those shitty </block> clutter which gets on my nerves. Having to type out <!– → for commenting a block or line is annoying too instead of just /* */ which is faster to type.
b34d4a No.16404834
>>/v/16404128
> >>/v/…….
Does this stop the (you)?
>>16404747
If it's orthographic it shouldn't be difficult, you just need to match the camera height compared to the object until you get it right. Distance from object wouldn't matter. A sample of a squarish building in-game would enough to match it against. For different angles, you'd just make however many cameras there are angles (8, I'd imagine) surrounding the object. Lightning can be done to taste.
b34d4a No.16404838
>>>/v/16404834
Damn, I fucked it up, but now I can try for myself.
b34d4a No.16404849
>>16404838
And I forgot to sage, shit. It works though.
846f0e No.16405028
>>16404834
I understand this part and I know how to set up myself, this is the least problem. The biggest problem is that I basically need to fit the weapon sprites somehow with those type of sprites. I am digging through the sprites files now from the dump which is over 16mb big so I can't upload it directly. So far I only found individual single sprites from agents with armor they are chopped around these parts I think (hands, left arm, right arm, torso, head, left leg, right leg) which for the arm alone there is 509 sprites + some blanks inbetween them. I think its a bit easier if I could find the damn weapon sprites somewhere. The gui "sprite" is easy as pissing adding one in, here is some examples with the shoddily done montage command I used. And that's also where anuda problem kicks in they are not in same resolution either but can be as low from 1x1 upto 64x64, non square which is quite a bitch. I can upload the game files if you want but you also need to convert them with a tool which is named OpenApoc_ImageDump and uses this as a pattern `{FORMAT}:{path_to_pck}:{path_to_tab}:{index of image in pck file}:{palette file to use}`. Or I'll just link you the bash for that which describes the pattern briefly. pastebin.com/yZ5H5NsY
135c4f No.16405307
>>16405028
dude, make an atlas
This will explode in your memory
https://www.spriteuv.com/ for instance (first google result)
fe398f No.16405533
ad4d01 No.16405554
fbc2b3 No.16405672
>>16405554
It could be a double action. There are some double action 1911's out there with visible hammers.
But it's more likely that they didn't put any thought into it.
b34d4a No.16406219
>>16405028
Must you use the same exact values as the source? I mean, if rifle_in_animation_frame_03 is W and H in width & height (pixels) and placed at X/Y positions in sheet Z, must your custom content match that exactly? Are you limited to simple replacement?
846f0e No.16406343
>>16405307
Its a game made in 1997 so I can't imagine that they way they did with the sprites is that bad. But it does suck modding wise because of all the different resolution and the lack of templates that can produce this kind of output.
>>16406219
>must your custom content match that exactly?
I don't know, I'm still trying to find the weapon sprites because its called different and catfish is too stupid to find the files that are named like this. so I guess its easier if I just find those sprites instead and try to make a few test renders or something. For the position I think it depends how the sprites are placed or something because the weapon itself doesn't define how the weapons are layered, its just called like this: <held_image_pack>BATTLEUNITIMAGEPACK_item8</held_image_pack>
> Are you limited to simple replacement?
No, I can add as much weapon as I want, so far I have Mosin-Nagant, PPSH-41, AKM-47 and Tokarev 33 none of them replaces anything.
fbc2b3 No.16406466
I coming up with names for a fake SWAT like group and fake special forces group that have a presence in a fictional city. So far I have Special Weapons, Operations, and Law Enforcement for the SWAT group, and 8th Urban Task Force for the special forces group. Are those names any good?
286d75 No.16406493
>>16406466
Don't forget about ICE.
9b5edc No.16406508
>>16406493
Immigrant Capture, Enforcement, Criminal Rehabilitation, Education And Management
286d75 No.16406525
>>16406508
>right wing rehabilitation squads
I love it.
9b5edc No.16406537
>>16406525
Of course you do, everyone loves ICECREAM
fbc2b3 No.16406582
>>16406493
You're right, though it would be more lore than actually come into game-play. SWAT and Special Forces is for threat escalation. Firstly there are guards, then if the player gives them much trouble they call the cops, then the cops call in SWAT if the situation demands it, and SWAT calls the Special Forces if things get really hairy.
It's a stealth game, and the player wants to prevent escalation. Most of the time it stops at SWAT and you either die to them or run away. You have to be really good to mess up bad enough to get Special Forces called in. If that happens you are as fucked in the game as any random 2-bit thief would be in real life. At least that's the idea. You see SWAT retreat out of the building, then moments later you hear a boom and see a bright white light, and then you die.
>>16406508
This is great. I just may steal it.
If you want credit other than "Anon" speak now or forever hold your peace.
9b5edc No.16406591
>>16406582
No need for credit m8, puns are free. Just put in the right logo somewhere hidden and all attributions by /agdg/ anons are taken care of.
9b5edc No.16406716
>use ifstream to read brainfuck code from file because it's easy
>] has to search backwards through the file
>try a rather long program
>breaks on ], claims there's an uneven number of [ and ] symbols
>check it manually, not true
>finagle debugger I wrote to scroll through the code while searching
>it returns beginning of file (character -1) on character 8191 despite knowing it's character 8192 right before
>turns out ifstream loads files in chunks of 8192 characters, and doesn't reload when moving backwards through the file
>no documentation about this anywhere
Fucking cuntflaps. I was going to have to rewrite it at some point anyway to make it faster, but I'd prefer to do that during the optimization stage at the end instead of being forced to do it after tearing my hair out over a seemingly impossible bug for two days. Especially since I'll have to re-rewrite it at the end anyway.
b34d4a No.16407581
>>16406591
>put agdg logo in credits
>make sure it's the curvy 8 one
>game is successful
>fans trace the origin
>8/v/ is swarmed by kids
>profited by shitting up 8/v/
You know, If an /agdg/ game with a logo made it big, anons here would end up hating the guy.
9b5edc No.16407598
>>16407581
First of all, most people aren't going to backtrace an image to here. Secondly, I explicitly said hide, not put in the credits. Third, I posted the problematic one, so normalfags would be scared away.
And frankly, after all the exposure we've had in the past, those kids are already here anyway.
846f0e No.16407720
>>16407598
>And frankly, after all the exposure we've had in the past, those kids are already here anyway.
I think it is even more of a reason to upheld the guard even more, the previous exodus which happened after the gamergate event if I recall it correctly where endchan got involved didn't went so well for better or worse. That was also around the time when hotwheels was still in charge. This agdg thread did very well to survive the redditor faggots.
>so normalfags would be scared away
That is reasonable, this place is actually one of the very few left it is still possible to lurk without getting driven crazy because its not so fucking shit and survived several newfag onslaughts. Despite the meme that /v/ is shit its not that bad because having discussion is still possible. So if anything the aggression against newfags needs to be increased a little bit more often, that might drive 1 or 2 newfags like that relative recent rust fag who tried to shit up sigma anon thread.
>>16404720
>How to most efficiently remove platform dependent code and revive the monster?
I am not that much experienced with programming so I cannot provide you any more tips except that you need to practice as often as you can which is the best method to learn programming, which is also one of the tips given by the ANSI C book. For platform dependent code you would need to look on standardized specification such as POSIX and work from there, but don't overwhelm yourself with all the tiny details focus first on learning how C++ works.
f16fe1 No.16407747
>>16407581
>>16407598
I don't particularly see any reason to advertise this place like that. If a dev actually posts progress on /agdg/, anons here will recognize the game. I don't think Speebot or Shinobudev's marriage simulator had any mentions of this place, though I might be wrong. The only real result of including a logo like that anywhere in the game is to advertise this place to normalfags outside of here.
fbc2b3 No.16408369
>>16406591
I've planned to make the game open source, so hiding anything in it would be more trouble than it's worth.
>>16407581
While I would love for my game to be so popular that people spend time tracing it's origin I really doubt people will care enough for that to happen.
>>16406716
Are you going to write an actual parser and tokenizer now? I remember when I was working on my text adventure I got carried away and started inventing a programming language and writing an interpreter. Wish I knew about GNU Bison then, but it sounds like you're trying to roll your own so you wouldn't want that.
Also I learned about this which is helpful.
https://compilers.iecc.com/crenshaw/
The section on lexical scanning in particular would probably be something you could use.
9b5edc No.16408479
>>16408369
I've made a scripting language using Flex and Bison before, but brainfuck is so dead simple that I don't think I'll need it. Everything is already a token, pretty much. Right now I'm going to just load the program code into memory instead of reading it from a file at runtime, and later on I plan to at least preparse the brainfuck to combine repeated statements. Changing it from an interpreter to a JIT-compiled thing is on my maybe-list, largely because I think it's already plenty fast unless I use the debuggers I added, and those are slow because of console printing.
But first I just want to ensure I can run any brainfuck program I find, so I can actually test the ones that run slowly to see how long they take to complete. I encountered this ifstream bug while trying to run a mandelbrot generator, for instance.
fbc2b3 No.16408595
26e40e No.16410546
This is the theme that plays when you finally corner, encounter, and fight against a certain somebody. He has a surprisingly debilitating spell list, though I am still working out some kinks with the fight.
https://clyp.it/ug205bq0?token=c90604a4fe7e25b87abbdedabc7846f8
9b5edc No.16410548
Got it working again, still need to update my debuggers to work with the new code handler. It takes about a minute to run this mandelbrot generator though, it slows down a lot on the blank spaces. Supposedly a JIT compiler could do it in seconds, so it might be more necessary to JIT this than I initially thought.
>>16410546
>starts of calm
>huh, this doesn't feel very bossfighty
>ohshitnigger.wav when it hits
Do you have partial song looping, so it stays in the hype bit once the intro finished once?
000000 No.16412399
>try to make a snake game on godot to ease into gamedev
>make one tilemap for displaying the snake/food and another for displaying the walls
>for now using a hardcoded array to define the level, stored in the script of the snake tilemap
>set other tilemap as child so it can inherit the array
>script the wall tilemap to automatically fill the playing field with walls taken from the array
>can't access array
>after some research manage to fix it by changing the inheritance of the wall script from the wall tilemap to the script file belonging to the other tilemap
<the set_cell() method in the wall script works with the correct tilemap without needing any prefixes
<the tilemap isn't mentioned anywhere in the code
What the hell? What exactly is the difference between scripts belonging to different nodes when the contents identical?
000000 No.16412409
26e40e No.16412636
>>16410548
>Do you have partial song looping
Yeah. The intro only plays once, and the outro only plays at the conclusion of the fight. Only the exciting part should loop during the actual combat. The majority of tracks are just simple loops, but I use a few with intros. This other track is still being worked on, but you can hear exactly how it's intended to loop. I hope the tracks are passable, because I've never actually dealt with music in my life.
https://clyp.it/1nwjxmem?token=df3864b36a71b96e8fb8216388393623
>>16402530
These posts have given me flashbacks to an old project I wanted to make. It was a shmup where you upgraded and designed your ship interior, layout, and systems, and missions would frequently have you send a landing party. But I figured I couldn't make it as good as Tyrian, so what would even be the point?
Your ships look nice and sterile. Is that actually going to have complex management, or is it just a vessel to serve another kind of gameplay up, like FTL?
5890af No.16415782
Where can I learn programming and C#?
135c4f No.16416388
bc93b6 No.16416797
>>16415782
I do recommend you to learn algorithms first. You will learn to think like a programmer.
There's plenty of information and resources on the web. Visit >>/tech/ and read the pinned thread. Also: https://wiki.installgentoo.com/.
Good luck.
Read SICP
penis
9b5edc No.16418020