/agdg/ + /vm/ Thread

I had to make this thread at 8 AM edition

Post your progress, ask any doubts you have about development, or ask for feedback for your game or mod.

Remember, there's #8/agdg/ in Rizon for discussion, we also have resources and FAQs in >>>/agdg/ and >>>/vm/

Other urls found in this thread:

mega.nz/#!hVJlxYII!HtVLhdGV7DWt3Bj_dl_pGW0hSR4-yb1XilosGLJncjs
commons.wikimedia.org/w/index.php?title=Special:Search&limit=500&offset=40&profile=images&search=clouds
commons.wikimedia.org/wiki/Commons:FAQ#What_licenses_do_the_files_I_upload_have_to_use.3F
docs.yoyogames.com
mega.nz/#!gNogFQ6R!yhoQRZTagmzgE9HDfht5BJ3-9TVOBV9zZRmDJnRdH1o
textures.com/
my.mixtape.moe/dtinwi.webm
mega.nz/#!BU5kFLYC!9vnBas5QftOycsXp-_PKXWm8UzNNF1qDp8pkqMNdapo
learncpp.com/
github.com/nashmuhandes/GZDoom-GPL
github.com/freedoom/freedoom
dotnetperls.com/unsafe
youtu.be/HUxqfVSwSz4
youtube.com/watch?v=xjUHMsAZti4
html5xonix.appspot.com/
webm.land/media/TBeC.webm
tutorialspoint.com/cprogramming/c_pointer_to_an_array.htm
wiki.unity3d.com/index.php?title=Triangulator
p.trumpshare.com/38tg/mauo.zip
docs.unity3d.com/ScriptReference/Bounds.Contains.html
docs.unity3d.com/ScriptReference/Mesh-bounds.html
docs.unity3d.com/ScriptReference/Renderer-bounds.html
stackoverflow.com/questions/22709094/c-application-built-with-120-xp-toolset-on-windows-xp-initializecriticalsecti
dropbox.com/s/cnk3zr891i2q6te/PlatformingTest.gmz?dl=0
twitter.com/NSFWRedditImage

Does anyone actually use the irc?

Anyone knows any FOSS turn based rpgs?

Get in line nerd.

Make one yourself fag

I want to look at the code to know how to make one in a good way fgt

So, I have been working on a Super Metroid and A Link To The Past romhack for almost a year now. Where do I go to upload it without getting DMCA'd or whatever by Nintendo?

Also, I stopped on my own game and found out that I enjoy Writing more. Good luck to you dudes.

Everywhere you don't need an account with your real data to get in.
Upload that shit somewhere, if it gets taken down, upload it again.
They can't DMCA you if they don't even know your e-mail

How do the share threads work here? I could try there. My play testers say it is better then Hyper Metroid, whatever that means.

It looks fucking ugly though.

Just upload it on vola/mega/whatever and post a link

Okay okay okay.

Thanks dude, care to show your amazing work?

pwnt.

>Ship redesign for no reason than to be edgy (but at least the spritework on that is decent)

Fuck off if you can't take criticism.

ALSO


Seriously, go away

Show ur shit faggot.

HAHAHAHAHAHAHA

Reposting from last thread because no replies:

I have a new build of my game, with a new level that I made. Here is a video of me playing through it. Its supposed to take place on a sky-scraper.
If you're interested in trying it yourself, you can download it here:
mega.nz/#!hVJlxYII!HtVLhdGV7DWt3Bj_dl_pGW0hSR4-yb1XilosGLJncjs
What do you think of it? What should be improved?
I had friends play the game, but they couldn't beat any of the finished levels. I'm not really sure how a more serious player except me would play my game, since I dont know any.
It would be helpful to see what you think of the level design. I think I might be too stingy on health-packs, because two of the levels that I made did not have any health packs, this one included. It took me a lot of tries to beat this level because I would purposefully engaged several enemies at once while recording, so that I could show the combat better, although it made the level harder.
If anyone has the time for it and is interested, I would like to see a recording of someone playing the game, so I can see how another player will attempt the level.

Why? I'll just change my IP or wait for the next thread.

Eh, just curious. If you have no content then I dont give a shit about ur "criticizm"

Way to show how much of a fucking newfag you are. This is goddamn Holla Forums, we don't hand out asspats, and we give straight honest criticism. Maybe you should go to >>>/reddit/ or whichever hugbox will infinitely circlejerk your hard work.

Sage because this is really not /agdg/ material now.

Threadly reminder to stick to your guns. Don't sell your soul to tumblr and reddit once your project gets popular. Use only free software to make your vidya.

the horror

Could be worse.
Could've been an else if tree 500 lines long.

Oh no. Oh no. Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.Oh no. Oh no.

Reminder that "it's shit" isn't criticism and doesn't contribute anything of value at all.

Reminder that and he still sperged out about my criticism

Your post is shit

Guys relax and start working

ur shit

Why not base your game off the doom engine? It's pretty cool you made that, but it's going to be a lot of work for you to make something comparable.

Please tell me that code was made as a joke. Please.

I'll leave imageboards as soon as I can find a community elsewhere. /agdg/ including these threads aren't necessarily bad, but the surrounding community in this website is fucking ass and brings nothing but negativity to your life.

You can't handle the bants? How thin-skinned are you? Yes please do leave when you can

You know, sometimes a little negativity is useful.

That's not just a community, it's humanity. A darker shade of humanity than you might find elsewhere, but humanity none the less. Best you can hope for is a situation where people act polite and professional without throwing drama and emotion into things, which is rare as fuck on the internet and never guaranteed in reality.

There's literally nothing BUT negativity here.

There's a difference between criticism and negative attitude. The former is you pointing out what's wrong with something and trying to steer it for the better. The latter is just whining and complaining and provoking people and trying to make everyone feel agitated and bad in general, it has no practical purpose at all except distracting the agitator from their own miserable life.

I just want to make cool stuff with cool people, but people in imageboards can only focus on the negative things in everything.

I'd rather the negativity here than the constant positivity in places like reddit where they eat shit and say it's gold because there's one piece in it that's not shit. So long as you have friends in real life haha yeah right then you should be fine, unless they're all negative too.

But I like getting my dick sucked, user.


Well I was expecting this type of community at the very least to be all over my nazi loli game idea, but the response was kinda a downer.

Devving is already depressing because of constant self-criticism and flagellation.

I was ready to abandon my known forums because of growing infestation of SJWs, but one thing is pretty clear: they actually finish and release more projects than alt-righters here.

I kinda get it, and I used to think that way too, but over time I've come to disagree.

Constant shit flinging isn't worth any more than constant cock sucking, and you're much more likely to find skilled and productive people among the cock sucking community than in the shit flinging community. Plus the cock suckers won't make you feel negative and depressed all the time for no reason.

Alright see ya later.

I don't feel negative or depressed all the time by being here. Maybe you browse too much since I only visit this thread. If I find anything depressing / frustrating it's how fucking retarded I am and how I can't work out how to do something I want to because there's no documentation and reading through the source code trying to figure it out (there aren't actually any samples of how it's meant to be done) is a pain in the ass.

I'd rather the honesty and criticism you get here than the dick sucking from reddit or other communities that doesn't help you improve your game.

That looks like it was decompiled after a compiler worked its magic, not something someone actually wrote

here's one
or two
or twenty

Here's what I said in my original post:

I'd rather skilled people give me advice than a bunch of imageboard whiners.

I think this is just another excuse that people have to feel superior. Just as those communities have a whole lot of cock sucking, these communities have a whole lot of whiners and agitators. It's actually not very common to receive genuinely good criticism or advice here, most of the "advice" I see tend to be opinions and subjective things, and even if there's no objective reasoning behind the "criticism", you're an outsider for disagreeing.

The idea that criticism doesn't exist outside the "honesty" of imageboards is a myth created by the people who want to justify their shitty behavior.

this is a pretty nice community, all things considered

You sound like a bitch, good riddance.

If you're talking about /agdg/ specifically, then I agree. Too bad it's attached to a steaming pile of shit.

Discussing anything that isn't completely generic erodes anonymity

I think that's why people here are so friendly

I think the main reason this community feels bitchy to you is because you're a bitch yourself.

Just bought the Gamemaker Studio stuff off humble bundle.

Now what? i wanna make a chapter master clone and sell it.

Now you make game

The trick to that is that you have to present yourself in such a way that you clearly only want actual constructive criticism with some thought put into it.

If you ask for criticism, then people will look at your product, trying to find the first best thing to complain about and will give you that. That's the quickest and easiest way to give criticism; and if you come to Holla Forums asking for criticism, that's exactly what you'll be getting.

Although this criticism can be useful at times (sometimes you need to be told that your UI design is fucking retarded, which then immediately becomes evident to you because you've just not spent thought into it, being too busy with other stuff), it is generally pretty shallow criticism. And yes, that's all you're going to get. Usually.

But if you want to actually get in-depth criticism, then you need to setup your thread as such. Tell people that you're looking for criticism; but not in the form of nitpicking, but in the form of actual fucking discussion. Here's the hard part: You have to actually present your game design and its inner workings to other people, describe why you took what decision, so that other people can read through it and then present criticism that is actually meaningful.
You want people to give you the in-depth and well thought out criticism you desire? Then fucking put in some effort; show them that you're serious about making the game better. Show that you put actual thought into your product and show that you're not afraid of changing your game's fundamental design if necessary. If people provide criticism, instead of replying "because x y z", respond "I could do that, but then I'd run into issues with x y z because a b c. How would I solve those then?" If possible, provide some of your own suggestions, and actually get a discussion rolling.
Just because you discuss these things doesn't mean you have to actually DO it. You can simply take the parts you like afterwards and integrate those into the game's design, but of course you have to be aware of the marketing aspect: if people got an idea different than the one you gave off in the discussion, then chances are they'll complain.


The main risk I see from trying to get a serious discussion rolling is that a lot of people will just ignore you completely because they feel too intimidated by the "actual putting in effort" aspect. But in Holla Forums there seem to be plenty of folks that want to give it a go just to prove that they're better than leddit&halfchan.

But that's

HAAAAAAAAAAAARRRRDDDDDDD

I give up.


You know what you're doing, can you make game for me and I'll cut the profits with you. It was my idea.

I know that, i own gamemaker and Clickteam fusion 2.5 thanks to the wonders of specials, anything in particular i should look at before getting started? already looking at the boards up top.

What should I use to make a pretty simple top-down shooter?

3D or 2D?

2d

gamemaker

it was owned by a jew, got sold to a catholic jew, and is now owned by a gambling company so it's absolved of all sins

Might check it out, but doesn't game maker have pretty bad reputation

absolved of all sins

Holy shit, shut the fuck up you tryhard, cringey little spastic.

It's always faggots like you that shit up threads and then cry victim with "u cannot handle the bantz!!". Fuck off back to Nando's you polo wearing gimp.

So I can make a game using it and not be considered a gigantic faggot?

If people know you're using GM, you'll be known as a huge faggot, so you'll have to put extra effort into making your game presentable

Okay thanks

What are your skills? Gamemaker as the other guy said is good if you've got very little experience. If you do have experience with C / C++ the typical advice around here is to make your own engine using SDL or SFML.

I have very little skills so I'll go with Gamemaker

I have nearly no skills but want to learn so i can git gud, will that help me a lot?

Yes but you're better off getting a book on programming first and reading through that at the very least.

Gotcha, anything in particular or just start reading anything i find thats recommended?

Remember to not just read, but also DO it.

There are always a lot of answers to that. Thinking in C++ seems alright from the first 100 or 200 pages I read of it. Otherwise there's always "The c++ programming language".

When it comes to programming, learning by doing is better than anything else. Experience trumps the rest. You can always just jump into programming whatever you want and learning the 'hard' way like that, but you should at least have a basic understanding of data types, data structures, memory if you're working in c / c++ and OOP for c++.

We're like 10 people online all the time. We talk plenty.

After several days of trying all kinds of shit
I finally got my game to compile and run using clang
Thank fuck for clang GCC fuck their shit up and i thought i would never be able to dev on linux again.

Okay, so right now I'm pondering how to implement health feedback. No numbers or anything, that's gay shit. I'm going to base this on real life™. I've been told that, when exposed to excruciating pain, the shock caused makes people see white, orange and then red. Now, I need some of your feedback here, what kind of white would be seen when being in great pain from, say, being shot?

Pic related is me toying around with postprocess settings, trying to get around.

I want to learn about writing game engines. The goal of the project isn't so much to write a game, as it is to learn about making game engines.

Sounds like you are contributing to the negativity tbh fam.

Be the faggot change you want to see.

I managed to gain about 10% performance boost just by rewriting one function in assembly. This particular function takes my 8bit frame buffer and converts it into SDL_Texture which is later displayed. I Believe this is the fastest way of achieving 256 colors software mode ala VGA and if you're making pixelshit then I think there's no excuse to not do it this way. Here's code in C++ and asm and a small webm


void blitScreen16_2xScale(){ WORD* screen_data2 = (WORD*)theSurface->pixels; for (int y = 0; y< SCREEN_RESOLUTION_Y; y++) { for (int x = 0; x< SCREEN_RESOLUTION_X_half; x++) { *MAIN_TEXTURE_PIXELS++ = pallete16_2xScale[*screen_data2++]; } }}void __cdecl blitScreen16_2xScale_asm(){ asm(".intel_syntax noprefix\n" ".globl pallete16_2xScale\n" ".globl theSurface\n" ".globl MAIN_TEXTURE_PIXELS\n" "mov eax, theSurface\n"//0x806cda8\n" "add eax, 8\n" "mov eax, [eax]\n" "mov ebx, MAIN_TEXTURE_PIXELS\n" "xor esi, esi\n" "x_loop:\n" "movzx ecx,WORD PTR [eax]\n" "shl ecx, 2\n" "mov edi, 0x806d580\n"//pallete16_2xScale\n" "add edi, ecx\n" "mov ecx, [edi]\n" "mov [ebx], ecx\n" "add ebx, 4\n" "add eax, 2\n" "inc esi\n" "cmp esi, 115200\n" "jnz x_loop\n" "xor eax, eax\n" ".att_syntax\n" );}

The code calls this function 1000 times and displays the number of miliseconds it took to do. As you can see, the asm version is about 100-150 ms faster.

If someone is internested in this "mode" (for SDL2) I can post a bit more.

It would be interesting to see the disassembly of the C++ version side by side with your handwritten one.

jesus man, just use shaders.

.text:08049617 var_C = dword ptr -0Ch.text:08049617 var_8 = dword ptr -8.text:08049617 var_4 = dword ptr -4.text:08049617.text:08049617 push ebp.text:08049618 mov ebp, esp.text:0804961A sub esp, 10h.text:0804961D mov eax, ds:theSurface.text:08049622 mov eax, [eax+8].text:08049625 mov [ebp+var_C], eax.text:08049628 mov [ebp+var_8], 0.text:0804962F.text:0804962F loc_804962F: ; CODE XREF: blitScreen16_2xScale(void)+61j.text:0804962F cmp [ebp+var_8], 167h.text:08049636 jg short loc_804967A.text:08049638 mov [ebp+var_4], 0.text:0804963F.text:0804963F loc_804963F: ; CODE XREF: blitScreen16_2xScale(void)+5Bj.text:0804963F cmp [ebp+var_4], 13Fh.text:08049646 jg short loc_8049674.text:08049648 mov eax, ds:MAIN_TEXTURE_PIXELS.text:0804964D lea edx, [eax+4].text:08049650 mov ds:MAIN_TEXTURE_PIXELS, edx.text:08049656 mov edx, [ebp+var_C].text:08049659 lea ecx, [edx+2].text:0804965C mov [ebp+var_C], ecx.text:0804965F movzx edx, word ptr [edx].text:08049662 movzx edx, dx.text:08049665 mov edx, dword ptr ds:pallete16_2xScale[edx*4].text:0804966C mov [eax], edx.text:0804966E add [ebp+var_4], 1.text:08049672 jmp short loc_804963F.text:08049674 ; ---------------------------------------------------------------------------.text:08049674.text:08049674 loc_8049674: ; CODE XREF: blitScreen16_2xScale(void)+2Fj.text:08049674 add [ebp+var_8], 1.text:08049678 jmp short loc_804962F.text:0804967A ; ---------------------------------------------------------------------------.text:0804967A.text:0804967A loc_804967A: ; CODE XREF: blitScreen16_2xScale(void)+1Fj.text:0804967A nop.text:0804967B leave.text:0804967C retn

I forgot about this addressing mode, I should have read the dissasembly before coding. Now the code looks like this :

void __cdecl blitScreen16_2xScale_asm()//unsigned int* dst){ asm(".intel_syntax noprefix\n" ".globl pallete16_2xScale\n" ".globl theSurface\n" ".globl MAIN_TEXTURE_PIXELS\n" "mov eax, theSurface\n"//0x806cda8\n" "add eax, 8\n" "mov eax, [eax]\n" "mov ebx, MAIN_TEXTURE_PIXELS\n" "xor esi, esi\n" "x_loop:\n" "movzx ecx,WORD PTR [eax]\n" "mov ecx, dword ptr ds:pallete16_2xScale[ecx*4]\n" "mov [ebx], ecx\n" "add ebx, 4\n" "add eax, 2\n" "inc esi\n" "cmp esi, 115200\n" "jnz x_loop\n" "xor eax, eax\n" ".att_syntax\n" );} And it's even faster. Now it's down to 150 ms

Was going to try this yesterday but I had to go somewhere and forgot about it when I came back, will try it today and tell you if it runs well in Wine or not.

He smiles upon thee.

thanks. I know that it runs on wine on a friends computer, but if it doesnt work tell me the error and i'll try and fix it

What kind of libs are you using where you'd have trouble compiling shit with GCC on linux?

Making some progress on those footstep sounds.

No all that is left is altering the volume based on the player's speed. Quieter when sneaking/crouching, louder when sprinting.

I'm aiming for 60 FPS on pentium 1 so it gotta be fast.

What's a good, retard-tier engine that I can use as an entry level introduction? I suck at rote learning things and learn better from putting things into practice. I don't mind doing things that are difficult but coding individual things with no application never sticks in my head.

Your friends aren't necessarily casuals for not beating any of the levels: the player movement is fucking awful and feels like a walrus sliding across the arctic. Play some Doom or Quake to get a feel for things and take a look at their source code to see how they implement it, and maybe study their level architecture too while you're at it.

I went to plebbit to shill and all I got was whining or cockstroking, neither of which are helpful. fuck plebbit and fuck you.


this tbh fam


please do us a favor and join him on his way out you cocksmoking faggot


multimedia dickteam fusion. so easy that even retards and children can use it. I'm not being sarcastic either, I used to make shitty imitations of flash games in the older version as a kid.

Just use game maker or something

SFML + C++

You'll never stop being a retard unless you know what's going on under the hood. Cpp forces you to take on some low level syntax and better understand the mechanics of how a game engine works so you don't make the typical retard mistake of using too many resources for a tiny ass game.

Theres an ebook on SFML game development. After you learn some retard basics of C++ you can Google a free copy of that.

I started with making a shitty engine in SFML then I started using Unity and it was much easier now that I understood how shit actually works.

Recommended subjects to study on:


After that make a simple, REALLY simple, 1970's arcade machine tier simple game, and finish it. Then use everything you learned so far to make that slightly more advanced. Then use everything you learned making the game more advanced to make it slightly more polished. Repeat the process as many times as you'd like.
You are now ready to make a game after that.

Overall, to learn anything, you need to push your comfort zone. Think of a mechanic, and implement it EXACTLY how you want it. No compromises, none of that "Oh, well, I'll just let it be buggy" shit, make it fucking work.

MoM user here; Maybe one of you guys can help me out with something pretty simple. I do not have a digital camera, but I would like an image of blue sky with cloud clouds so I can use it as a texture.

Anyone can help with that?

Might want to rephrase that?
Also, I don't think using real life textures would look good with your overal cartoony graphics, but that might just be me.

Just pixelate that shit, fam.

You could always look for free stuff.

commons.wikimedia.org/w/index.php?title=Special:Search&limit=500&offset=40&profile=images&search=clouds


I think he's obviously going to crop it.

Only the characters are cartoonie; environments look like they are from an N64 Zelda game (its always a good rule to not have the environments and characters look cartoonie, as its always important to make the characters pop out!).


Wikipedia won't sue my ass or anything, will it?… while I dont plan on selling my game, the recent "We will DMCA your ass if you use any copyrighted material what so ever" has made me pretty cautious with whats assets I use. I usually try and make sure the stuff I use is not lifted from another website anyway (unless of course its COMPLETELY and UTTERLY free use)

I'm pretty sure it's completely free, bro.

commons.wikimedia.org/wiki/Commons:FAQ#What_licenses_do_the_files_I_upload_have_to_use.3F

Awwh yes. Thanks a lot for that, then!

Now I just need to find a musician who is a complete neet. The last 2 I had got busy and disappeared, and I am contemplating making MIDIs myself despite no experience

So how many of you let your games/codes tested by other people with at least basic game knowledge?

i do.

That's good. I do it all the time when visiting small game design cons. I'm the asshole that'll litterally try anything or push the game engine to see when it breaks (yes, when, not if). It's hard in the beginning to have something you worked on for so long be tested thoroughly. But in the end it's a very valueble process during your development.

Thanks for the tip of your cock, faggot. Can you still make bank on shitty mobile games or has it become oversaturated and consumed by bald, thick rimmed glasses wearing, buzzword spouting, marketing fuckboys?


I'd rather have a steep learning curve and a higher payoff at the end, isn't gamemaker the casual tier of development?


I probably worded what I meant poorly, I mean that I want to learn what's under the hood but repeating code out of any practical context has never stuck in my head so I think I need to learn it while applying it to something to get a grip.

Which one has the most potential/documentation out of SFML/Dickcream? I've also heard Construct thrown around a bit, is that worth a look?

A script is a function you can use on any object, and which you can pass arguments to.
It's for common functionality so you don't have to copy paste shit into different objects, and don't have to change all of them manually when you want to change that functionality.
docs.yoyogames.com

scripts are reusable
and you can pretend you're using an object oriented language, instead of some bastardization of it

Make them yourself. It's worth the experience you gain from it. Because next time you know exactly how to address audio-nerds and make the idea of doing stuff for you appealing.

Making lots of levels, here's a sneak peek into one of them.

this makes me sad
you've gone from a stupid little robot to something resembling a game and I'm still wage slaving away

you're shit at compressing webms, though

You never could, the "massive revenue" the mobile market makes has always been gained by a very small selection of apps.


Would you consider releasing a level editor along with your game?

I'm a little disappointed that it's an actiony twitch platformer instead of a cozy exploration platformer, but that's just personal preference. It looks really good.

How did you get that sort of cartoony aesthetic? Is it just a matter of using mostly simple, solid colors for textures?

At least I can pretend to be fast at 170 fov and jumping past everything is sort of fun. Using opus is pretty neat too, don't see many games doing that these days.

Game Maker Pro and it's extra modules are available for $15. Alternatively you can just get GM pro for like a cent or a buck.

Just thought you guys should know, I wish someone told me faster when it was $6 almost a year ago, I missed it then. Grabbed the $15 option now.

I dig it.

Meant to post pic related instead of image #1

I do have a full-time day job though. Good luck user, it's tough. 8ch has a small filesize limit, this is as good as it gets for a 60fps webm. Might play around with the ffmpeg settings later for future webms though.


It's a possibility, but don't hold your breath. The current editor I have works well but only as a part of a development workflow for me as the developer, since I also have access to the source code. A user-friendly editor with some sort of shareable level hub would take a lot of time to make and is not currently a priority. Although if I make a sequel then it'll probably have a polished editor then (possibly integrated with steam workshop).


It's not usually twitchy, I'm almost speedrunning in this video since I know the route. The gameplay also slightly varies in different levels. This particular stage is actiony and has some tricky jumps, but the next one has wider areas and more exploration. Some friendly robots can be met in calmer areas and they talk fun stuff.

I tweaked the lighting and textures a lot to achieve this look.

Right but what did you do in general?

sfml
but that wasn't the problem
the problem was i upgraded my ubuntu version
which apparently introduced some bugs to gcc which I wasn't able to fix
I know I'm an idiot

There's a single 8/agdg/ game in the bundle too, in case any of you remember Uncanny Valley.

Yep, I remember it. I have no idea how it got that popular though, but good for the dev. There's also a Ross gaming dungeon episode on it.

I don't understand why you think you would be doing something like this.
If you wanna know whats going on under the hood you should use something like sfml
Engines are a form of abstraction and don't allow you to see what's going on clickteam fusion is an engine
sfml is an opensource graphical library

You should also ask yourself how low you wanna go and how much you want to understand
The lower you go the more time it will take and the more understanding you gain
In a way it's balancing act and where exactly you fit in you need to figure out for yourself

sfml is a good starting point
if you wanna be less autistic you can try an engine
if you wanna be more autistic you can try SDL

It's also worth noting that these options have variance in how much you learn from them
You can more or less get sfml working without really gaining any understanding if you want to
and you can also learn a lot using an engine in some cases like EU4 has open source code

This is not
how you write a paragraph
faggot

I'm pretty happy with it actually

I guess my strategy is to make the textures flat colored with crisp details where appropriate, have the colors be vivid but not too saturated, and the light bright enough but not too bright. Sorry that I can't explain any better, it's really just a matter of tweaking and recompiling to see what looks good.

I mean previously I tried learning some coding by doing tutorials, but they were aimed more towards computer science students who knew what the applications for the code would be. I could do the tutorials, but I wasn't really sure what the purpose of the code I was creating had outside of the tutorial, and because of that it didn't really stick well.

I can't see myself ever bothering to create an engine. It would be good if I could become skilled at an engine that allowed tweaking though should I ever be onto a winner and want to get some more talented people in on it with me.

I'd ideally like to get to a stage where I can create concept prototypes and debug in UE4 and Unity. Then I can go early access and not complete any more of it like every other indie developer :^)

Could you post a texture file, as well as a screenshot with and without the lighting applied?

If you're up to it, of course. I'm mostly just curious.

Is this with maximum compiler optimization?

anybody here uses fuzzy logic?

That's not really twitch platforming. Speebot user was merely speedrunning the level so it seems that way. If you really think about it that level layout is not unlike something you'd see in Super Mario 3D Land/World(In the case of World, it's earlier levels that are easy).

I think I did once for AI

I know, maybe a bad choice of words. But I get the impression it's just "go from A to B across this linear path"… not really any choice in how you get there.

I was planning on mixing diferent techniques, maybe learn to do a finite machine state with hardcoded basic shit like A* and decision making with fuzzy logic.

Maybe I'll learn to use neural networks if they're not that complex to develop.

Sure I'll take some screenshots tomorrow (as long as I don't forget)


The levels are generally linear, but there are optional collectible crystals which are placed in locations that would require the player to go out of their way to grab them (if they're up for the challenge). As the game progresses the levels get more complex (I've made and polished 9 levels so far), and they will eventually be bigger, with multiple paths and multiple finish flags in different locations (the engine already supports that).

Never used neural networks, but I think Quake 3 does something like that. AFAIK the bots store their memory in files and improve their skills over time.

they dont play first person shooters like doom, at all. that what I meant. I'll fix the ice movement.

The game slows down because the path-finding function will call printf a couple hundred times. I was doing that for debugging, so it makes the game seem slower. I dont need to debug that anymore, so once I remove the printfs it will be faster.

I used webp before I heard about .flif, so its my mistake..

My code for openAL is kind of buggy, sorry about that.

Can you tell me where enemies were firing through walls? I havent encountered this yet.

The only enemy that can shoot a homing projectile is the "captian".

Which levels did you try and what did you think of them?

I know you can clip into the walls sometimes, but its hard to clip through the walls. The engine is basically like a car built out of scrap metal and duct tape.

There are ammo pickups, the first level "tower" is unfinished, but in the finished levels "facility", "radio tower", and "temple" there are ammo pickups.

You should try playing those three levels, since the rest are incomplete.

Thanks a lot for your feedback, i'll work on your complaints and try posting a version where the problems are hopefully fixed.

There's even a creepypasta on that.

Ice cream? I love ice cream!

>doesn't work on React
chocolate doom works, you have no excuse


reeeeeeeeeeeeeeeeeee
open pak0.pk3 and shut the fuck up

Some random shit before I delete it from my hard drive

It's true, my uncle's wife son works at John Carmack.

I'm sorry, this is an issue i'm having where visual studio wants to compile it to use a bunch of modern windows features, so it gets all messed up because im not sure whats calling that function. I'll try and fix it, but its hard to track that issue down.The issue came up with someone running XP trying to play it, but I didnt take the time to figure out what was causing it.

I agree. It's not being thin-skinned (necessarily). I get tired of the autistic fits of rage that people have frequently here. I get tired of the retards/pol who derail every thread with unrelated shitty memes. I get tired of hipster faggots who hate every game unless it's from 20 years ago or completely obscure (and feel the need to opine on something they've never played anyway).

Also known as a coucksin.

Something is wrong with the mouse controls running in wine. I lose camera control if I do the equivalent of alt-tab, but if I do it again I sometimes re-gain control, but sometimes control gets really jaggy, like I'll turn in increments of 5 degrees, the amount I turn depending on mouse acceleration. Have you looked into making a native linux build?

alt-tabbing is broken right now… open the console with "~" to unlock the mouse cursor and then tab out. The problem is that I forgot to stop locking the cursor when the window is out-of-focus

ReactOS is still far from finished, though. Most Game Maker vidya doesn't work at all. However, I do see as a potential platform for retro gaming and homebrew somewhere in the future.

Just in case my client is around here, I hope this looks good, I still have to add lots of detail but the base is nearly complete, the plan is to make the magazine detachable and maybe make a bullet model as well, what will give me nightmares will be the whole detailing thing.

...

Alright, ive attempted to fix the all the bugs you're encountering, that can be immediately worked on at least…

The ice movement is hopefully gone now, you have much more friction and a faster acceleration, but the maximum speed of the player is the same.

The gunfire sounds dont freeze or loop anymore

The path-finding slowdown is gone

You can alt-tab out of the game without it locking up the cursor

If the user with the dll problem is still here, I would very much like to know if this works with your setup on reactOS. I think I may have been able to build it to be compatible with your system, but I cant check from my end.

Here is the download link:

mega.nz/#!gNogFQ6R!yhoQRZTagmzgE9HDfht5BJ3-9TVOBV9zZRmDJnRdH1o

Let me know about any other bugs you're encountering, or anything else that needs to be changed. I'll try and fix it.

Thanks again for the feedback.

My only complaint is that it seems very samey. Everything about it looks great EXCEPT the actual mechanics at work. I know you're nowhere near finished, but I'd really like to see the number of platforms, gimmicks, mechanics, etc expanded upon farther into development.

Godspeed, though, man. You've got me jealous.

How are you doing your footsteps? What I've got is that on the animation when the player foot strikes down on the ground, the animbp does a trace to determine what type of material it is, and them plays a footstep sound with varying noise levels (for stealth mechanics) based on the mat. Works pretty nicely.

Coming along nicely then. I look forward to details being added. If I can nitpick I'd say that the barrel is possibly a bit too long and the stock being wider than the rest of the gun looks a little funny and would probably make it annoying to handle.

Making a bullet model is all up to you, I've already got a few that would suit just fine.

Complete ideas guy trying to understand coding here.

Let me try to at least get one thing straight.
Let's say I wanted to implement a weapon that kills enemies in anywhere from three to four hits.

Am I wrong in considering that I have to set the weapon to do a specific amount of damage from a range of anywhere from 30-50, per hit, and then adjust enemy health properties as needed?

yes I need to fuck off and read more tutorials, and get the basics down first, but I'm just trying to see how someone else would approach trying to assign multiple random variables to one specific action

No, you need to give the enemy a health value, then make being hit by the weapon deplete the health so they die in 3-4 hits. There are no hard-set numbers involved here. You can give the enemy an internal hp of 3, then make the weapon do 1 damage.

ffs bro do a python tutorial or something before you try to dive into gamedev, and if you're using GM:S don't use the drag n drop, you'll learn nothing and end up with a giant mess

Yup, I'm retarded.
Figures. And for some reason I tried looking at C++ first. I understand the drag and drop bullshit in game maker will fuck everything up, but not entirely why, so I'll take your word for it.
I know I'm getting ahead of myself, I was making it way more complicated than it had to.

This isnt even a programming question, its basic math- so do it like this:

enemy health = avg weapon damage * 3.5

then adjust the damage range for how random you want that to be.

with your 30-50 example, enemy health should be 120.

It'll fuck everything up because it quickly turns into a giant mess, organizing actions is a giant pain, and it has probably like 10% of the total ability of GML. It's really only there because the original version of gamemaker was an educational tool and had it.

For programming, start with C or Python. C is more powerful and a lot more like languages that are used in gamedev, Python is a lot more flexible and easy to learn.

Anyone have more game dev-pertinent infographics?

I heard somebody paid 2000 dollars for a revolution 60 tier model, anybody have it so i can gawk at it?

I urge Unity Devs to use Source SDK Hammer + Object Viewer Export to Blender for making levels, this shit is pixel perfect mesh snapping, I made full city block in under two days, this would have taken me atleast a week to do in Blender alone and I get a better result at a lower poly count and more accurate collision meshes.

I absolutely URGE you to try this out if you use Unity.

You that guy doing the Neotokyo esque stuff? Looks decent. Was it hard to learn to use hammer?

Ehh I am making a dystopian future type game but I think I know who you're talking about but I'm not him.

It's not hard to pick up, however there is a high learning curve considering you are REQUIRED to use 4 view ports minimum, but you should get the hang of things quick if you're only using it to make geometry which is what I'm advocating.

Check out 3klikphilip and TopHATTwaffle if you need Tutorials.

Most important buttons are:

'Z' for viewport freelook camera
Ctrl + Click for Multiple object selection
[ and ] to increase and decrease grid size
Mouse wheel to zoom into out of grid based on where cursor is, I solely use this to navigate the grid view.

More Info*

When you export to obj then blender, merge all objects into one, go into edit mode, select all, then remove doubles.

Also one more important button for hammer is X when in the view port to allow you to resize stuff in the viewport.

Thanks m8. Will look into it. Did you have any experience modeling before working with hammer?

I really don't want to be "that cunt" in every thread but can someone point me to an engine for LITERAL RETARDS that can help me make:

I just want to make something, jesus.

Also, if anyone needs a really basic writer, I can probably help with that.

Actually hammer was my first 'modeling' program I ever used, back when I was 13 making surf maps for CSS, I have about a years worth of experience with blander, my strong suits rely mostly in drawing and programmin', however I'm competent in 3dcg no master though.

Also final tips for hammer hopefully

Most important concepts to learn (not complex)
-Carving
-Wedge Tool
-Vertex Manipulation Tool
-Geometry Scaling

(noice)


MUH DICK
Also, I think GameMaker, LOVE and even Godot could be useful for that. There was also a project for remaking that game in the Shadowrun D


Source engine sounds like your thing.
If you want to go lower, Darkplaces and GZDoom could be nice options for you

I kow that feel, learning some /loomis/ and Python now.

Sorry, I messed up that part

Alright, I'll give it a shot. Is there some sort of tutorial or royalty free sprite place where I can get assets from? I can't draw for jack shit.

As much as I'd want to play a remake, a turn based one would be the last kind of remake I'd ever want to play.

Go with anything made by id Software especially the original Doom engine and ioquake3 or Cube 2: Sauerbraten.

I tried making a mod of Demonsteele at one point for a spiritual remake of Shadow Tower but looking through the code and how to set specific shit made me feel like a mexican immigrant in china.

clickteam fusion / construct 2

Are those only for side-scrollers though?

Not necessarily, though
The next Penelope is made with construct 2

I used to be a wannabe Source modder. Still love the engine, it's just that I don't have time to change 2000 lines of code and make my maps painfully small so I stopped using it.

How do you export to Blender in Hammer? Also, does it save the UV mapping? Because that's the main reason I hate using blender.

I think I can do this without too much feature creep


Since space doesn't really have aerodynamics, I've been trying to figure out a way to make "interesting" shaped ships viable, compared to gigantic cubeoids.

Thruster placement doesn't really make a difference between top/bottom graphics, because they'd still be both just as good going forwards, the only difference would be the turning speed. Right now the cube is optimal because it reduces surface area, therefore chance of getting hit, and reducing material waste. Maybe some kind of heat transfer system?

I'm going to ruin your dream. That game has already been released and had relative success.

I do like your heat transfer idea though, it's a good way to solve the problem you identified.

Really, what game is that? Although to be fair, most game ideas have been done to death already. I just figure it's a simple enough project to keep me busy before I lose interest.

Reassembly and Captain Forever remix come to mind.

I made it a little thinner, the barrel can be shortened with no problem.

Captain Forever Remix looks juvenile, but Reassembly looks more like what I was going for. I don't really into vector or non-square shapes and rotations and shit… I just want a nice arcadey sprite based thing. I hope

Well, play both of them and get ideas if you're gonna maek gaem.

...

That code can be improved by turning ESI into a downward counter and thus removing CMP instruction.
void __cdecl blitScreen16_2xScale_asm()//unsigned int* dst){ asm(".intel_syntax noprefix\n" ".globl pallete16_2xScale\n" ".globl theSurface\n" ".globl MAIN_TEXTURE_PIXELS\n" "mov eax, theSurface\n"//0x806cda8\n" "add eax, 8\n" "mov eax, [eax]\n" "mov ebx, MAIN_TEXTURE_PIXELS\n" "mov esi, 115200\n" "x_loop:\n" "movzx ecx,WORD PTR [eax]\n" "mov ecx, dword ptr ds:pallete16_2xScale[ecx*4]\n" "mov [ebx], ecx\n" "add ebx, 4\n" "add eax, 2\n" "dec esi\n" "jnz x_loop\n" "xor eax, eax\n" ".att_syntax\n" );}
Now assuming that my x86 skill aren't so rusty I've messed up the memory addressing, this can be further improved by using esi as a pointer offset, which allows you yo eliminate the add instructions. A side effect of this change is that will now loop backwards over your buffers.
void __cdecl blitScreen16_2xScale_asm()//unsigned int* dst){ asm(".intel_syntax noprefix\n" ".globl pallete16_2xScale\n" ".globl theSurface\n" ".globl MAIN_TEXTURE_PIXELS\n" "mov eax, theSurface\n"//0x806cda8\n" "add eax, 8\n" "mov eax, [eax]\n" "mov ebx, MAIN_TEXTURE_PIXELS\n" "mov esi, 115200\n" "x_loop:\n" "movzx ecx,WORD PTR [eax + esi*2 - 2]\n" "mov ecx, dword ptr ds:pallete16_2xScale[ecx*4]\n" "mov [ebx + esi*4 - 4], ecx\n" "dec esi\n" "jnz x_loop\n" "xor eax, eax\n" ".att_syntax\n" );}

Find and download 'Object Viewer' from nem's tools, I have not tested the UV mapping so I am not entirely' sure, the program converts VMF to OBJ, so there might be native UV support but I cannot say for sure.

Looks like you made the whole thing thinner, not just the stock. Maybe the angle is confusing me. Only the stock needs to be thinned down a bit IMO.

OK, will do.

An optimised version of the C code, which should involve less second guessing by the compiler.
void blitScreen16_2xScale(){ WORD* screen_data2 = (WORD*)theSurface->pixels; size_t count = SCREEN_RESOLUTION_Y * SCREEN_RESOLUTION_X_half; do { *MAIN_TEXTURE_PIXELS++ = pallete16_2xScale[*screen_data2++]; } while(--count);}


Looking at your function, constants, and table names, I do question what you're actually trying to achieve here. Are you trying to double the pixels horizontally? Or do a look up on a 16 bit palette? In the either case using a 256kb table for loop ups will probably do poor things for your memory cache.

I don't have any but one good trick I learned years ago (in UDK) was that you can actually scale and layer textures for making patterns less discernible at a distance. For example, let's say you have a repeatable tile that looks good up close but when you step away (for instance, looking at a hill in the distance) the texture looks like it lacks depth and has this "video-gamey" aura about it. What you can do is blow up another texture (that has more defined, macroscopic features) and then interpolate it with your main texture based on a normalized distance factor. Up close you won't really notice it but at a distance it really makes your terrain have character and "pop".

As long as you pick a good normalization function and have a sensible upper limit on your distance factor, you won't really need a hi-resolution texture and can get away with 512^2 or even 256^2. You could even use another texture that would already exist in memory provided you shift the color correctly.

This effect can be chained for different granularities (up close texture vs medium distance texture vs far-away texture).

Another thing you can do is procedurally texture a terrain based on heightmap data (or even just your terrain mesh) and the relative slope between individual vertices in the terrain. For a basic example, let's say you clamp an underwater sand texture to between z:0 and z:200 because you have a water plane that extends across you map at z:200 and then have a grass texture between z:200 and z:300 and finally polish off with a snow texture (snow cap) for z values above z:300. This will automatically texture your terrain, albeit very poorly, so that you don't have to store everything in a monolithic texture or use manual vertex based painting (except for touchups). This basic principle can then be made as convoluted as you want: maybe have a texture that serves as a map for defining and interpolating between different texture sets, giving you the ability to tweak and define different regions; perhaps define a case for generating cracked mud at the bottom of a desert basin or rock crags on the side of a mountain or canyon; interpolate between a mud texture and a dirt texture depending on an input factor determined by how long it's been since it's rained in your game. The possibilities are endless.

textures.com/ for free texture resources. Most have to be touched up or added into a bigger composite. This is where photoshop skills come in handy. Use GLSLfor your shaders or a graphical shader design program that outputs GLSL code. Most larger games on-the-fly compile GLSL into an actual shader program. If you've ever played a game where you got a bunch of visual artifacts and had to clear its "shader cache", this is what you are deleting as it forces the engine to recompile all the shaders using the faculties provided by your graphics card's driver suite (though they use glCompileShader to abstract this).

IIRC Deus Ex did this, where if you got close to some (not all) wall surfaces, it'd add a secondary texture over it that gave it finer details, making it look a lot more detailed instead of a low res blurry mess.

This was an aspect built into the Unreal Engine.

Yeah I know, was just stating a game that I knew used it.

It would be useful to have textures so we could see colors other than white


Yo that's cute as fuck nigga

...

Okay so here's a better detailed version of how I picture ship building to work at this point


As a consequence of how pieces overlap graphically, the game would actually use (1) or (2) as the hitboxes for projectiles, so it it might look like you hit the sprite, but the bullet would fly past. A scanner system could reveal the actual underlying nodes so you know what to aim at, for example.

What's that even supposed to be, it kind of reminds me of a Steyr but it's definitely not one.

The foregrip is too thin it needs to be wider, make a round to scale and see if it fits in the magazine.

A custom AUG, still needs some detail though, you mean the one that folds? still need to check some pictures, is harder to have some sense of scale when I only find side views and sometimes videos don't tell me much, in any case, if you know more please tell me.

now this faggot troll encouraging people to make their own engine ahahahahhaha

When was this, was this prior to GG?


Um… the accolades by Kotaku and RPS don't clue you in?

First prerelease build was on October 26th 2014, last one on January 29th 2015. So not really.
And if I remember correctly, the dev just sent it to youtubers, youtubers love horror games because kids like it when they act scared, and thus it got popular.

Just look up some GM tutorials on youtube or some shit

Since I don't have a player model but a capsule, I use a different approach.
I determine when the next sound should play based on the distance the player moved. This threshold varies depending on whether the player crouches, sneaks, walks or sprints. When the footstep sound is played, the velocity of the player (relative to the max. velocity he can achieve, which would be WalkSpeed * SprintSpeedModifier) is used to alter the volume and the pitch of the sound. This wasn't in the video from yesterday though, as I implemented it afterward.

Are you using tris for collision?

Off to a good start.

Probably time to grind out more code, anyway.

Names are not easy, I tend to have like 5 completely different options, plus some variations of those

I guess you could say you're having an identity crisis

Well the first thing you should decide on is gameplay and then the story, world and name should come from the gameplay and then you'll find something that fits.
See what direction your game takes and then attach a name to it. If you called your game "Fire World" and ended up making a game set underwater that would be pretty embarassing.

He's got a game, gameplay, and story, he's got a few levels planned out and everything, he's the one making the not-samus space pirate FPS game.

Iv gotta name right here for ya: The legend of faganon dildo of time.

The game I'm making has reached its minimum viable product. Working on adding more complex things like making the combat actually interesting and filling out the world.
I can't stop thinking about the visual polish though. Even though it's basically just a text based game I can't help but want to switch around what the interface looks like. Every time I test a feature it fucks me off with how bad it looks.
Worst part is I've already spent two days getting it how it is and now I want to change it entirely.

oh ok then
I would break out the old thesuras and look for simlar words to "Future" "Pirate" "Space" "Action" and "Adventure" etc, then put them together in different order and such, find 5 or so titles from this process. That or take some names from the story that are significant like the name of the locations or worlds or important items in the world and then mesh them together with the phrases earlier.
That or just give it a one word title like the planet of the game or something.
I guess it depends on what tone you're trying to establish, if its a comedy game it can be light hearted or if its more serious a one word title would work too

...

I honestly thought it was called Space Pirate in Space, and loved that name.

Is is the right time to start practicing killing myself?

It's the right time to add ASCII representations to your game.

I haven't posted an actual progress on MoM awhile; I been working on the actual trailer for this project, and I wanted to keep a few things a secret. That being said, I have been working on this project non stop for the last few months, and I can't wait to show off the trailer when its done, as they is a lot of surprises and changes I have done that I want to show!

Heres the Battle Stance of Leo the Wolfman when you attack him. You can kill every NPC in Memoirs of Magic, but some of them will fight back, and may even be Minibosses; Leo, the guy in pic related, is one of them!

I am hoping to have my trailer out by next month; they is only a few more major things I need done before I can start filming it!

Yes, perfect.


Like the other user said (who knows me, scandalous!), I've already got the gameplay in mind, though story is WIP. The problem is just that I'm extremely picky and don't like settling for stuff unless I'm absolutely 100% sure on them.
Too picky for my own good, really. I'll probably spend a few more days doing the thesaurus method and then throw up a public poll or something, so as to stop progress being blockaded by my own autism.

In actual progress news, apparently timing the heavy shot of the blaster at the exact point of the upwards arc of the drinking sway causes it to multiply exponentially. Oops.
A glitch that will rarely happen, but has hilarious results so here's a webm.

Haven't posted this here yet.

no bully pls

Oh hey you finally got that Lightgun Toy sprited?

Just kill yourself while you're ahead user. You're almost at the bottom of the bell curve.

my.mixtape.moe/dtinwi.webm
Apoligies for the resolution and quality, it was tough getting it even this good. I'm the user from earlier with the mouse problems in wine. The problems I had are now fixed and it runs smoothly. However, as you may have noticed my sensitivity is crazy, though I'm gonna blame that on linux. xset doesn't seem to work for changing mouse sensitivity, so I can't test lower sensitivities. I can compensate with very small hand movements and the game does seem to register fractions of milimeters correctly, so that's a plus. I tried setting the sensitivity to 0.1 in the games config, but it seemed to round it to 0. Consider fixing this. Additionally, when I die I can't get back to the main menu. Spacebar doesn't seem to work and the game crashed with a segfault once after I hit space a few times after game over. Also, sound doesn't work at all. I don't have a playback device in pulseaudio, so it's failing to even create an audio device. Could be a wine thing. Again, consider getting a native linux build going. Also, the arrow keys seem unresponsive in the main menu. Either it's not reacting instantly or it's dropping key strokes.

Is the kebab removable?

Its no problem, im just glad that you took the time to help me with it.

The problem with mouse sensitivity might be related to how i'm getting mouse input- i'm using winAPI's procedures to poll raw mouse input, so it might just be skipping all the overhead and settings and going right to the hardware. Also, the config file only takes ints.. I know this is a bad system, but I never fixed it. I'll work on getting that changed.

I need to re-write my sensitivity system because its old code from about January or February… it can be done much better.

The issues with the menu are because I hacked it together in an afternoon, since its mainly just for letting you pick which level to load. I'll work on making it better. It delays responses on purpose, but it does it too much, and I never got around to changing it.

The game basically frees all the structures and becomes unresponsive when you die, so you cant get back to the main menu. I didnt get around to fixing that, but I will now.

The sound issue is because the included openAL32.dll is non-functional and I havent taken the time to find a working one yet. You can fix the sound by downloading and installing openAL, but if you dont mind I do need to figure out which DLL will work. Unfortunately you cant static-link openAL.

Once I fix all these things, i'll get back to you with a new build that hopefully solves the problems.

i finally managed to pass my exams and now i have time to make progress
behold, the best boss fight of 2016, with groundbreaking jumping mechanics, since i have no idea how to do interesting combat for a sidescroller

Something about that rotation seems very off. Detailed enough, but the tail, while being offset, just looks fucky in some angles. Two of the diagonal sprites don't read well, away-left and toward-right. Especially the former.

I still like the general style, though. The kind of saturated and crisp that you don't often see anymore.


That bottle is giving me Grezzo vibes.
I'm not sure if that is a good thing.


There is something very pleasing about the phrase "removable kebab".

I mean WHAT
Even if you are developing for mobile, you could use the arrow keys for testing user. It looks nice.

it's supposed to be for mobile
i was initially testing by using unity remote, so i could still use the phone to control shit, but for some reason it stopped syncing so i'm just using the mouse to click shit instead

Seamus O'Dash and the Booty of Space?
Doesn't work if the main character is female, of course.
Wanted to work alcohol in the title but couldn't find something with a nice ring to it. Also, there better be rum in that game.

...

You have Bulgarian ancestry?

kebab and doners taste great though
t. filthy tatar

I saved these a few threads back. Fairly basic stuff but whatever.

...

...

What are you gonna use that effect for?

With time, you'll find out that the shittier the culture, the better the food. Arab and Indian cuisine is delicious.

Is this a roguelike or a RPG?

clearly low health

you should do some color change for the screen as well, like in that segment of MGR where raiden gets his arm chopped off.
speaking of which, you could hide the UI too

He might be as a sidequest.


RPG. It'll have a time limit and limited saves, but that's as close as it gets to a roguelike.

I'd be completely distracted by the higher intensities of that if my health went low. At least there's no chromatic aberration

Yo, can I look at the code?

I'll release the source code when it's done. All I have so far is collision and rendering.

Alright, I thought you were some other guy who had a combat system semi-ready already, he used placeholder sprites just like yours.
Anyways, thanks

t. ideaguy
>>>/reddit/


shooting your blaster in the air is a necessary part of drinking

The Direction fuckery is an issue that occurs with 2D sprites when you try and give them complex "Poses" like this idle animation, unfortunately. It is rotated correctly, but theres a reason you see the more "Action" kind of poses in more limited angle games such as Fighter Games.

Were the placeholder sprites @ signs and D's?

Because if so that was me.

And I scapped that game because I was rushing too hard and the code became spaghetti.

What's the best doom port for building a game nowadays. I know the one MoM dev uses is unfortunately no longer being updated.

pls rate.

Why is codecademy bad for learning and what's a good book to start with C?

Is C good to start with?

You need more spinach because you got flabby legs.

If you want to go full gentooman then yeah. If not, go with some horrible mix of Python, Java, PHP and GO and make sure it ends with .js.

I read K&R C and then went to the NeHe tutorials to learn openGL. If you're using winAPI, theforger's tutorials are a good way to get started, although you only need the bare minimum explained on NeHe's first tutorial if you aren't planning on using any other UI features.

Its rather easy to get something working with C, but you have to know what you're doing, so it takes a little while.

Here is a short example of pong for windows, with the source code: mega.nz/#!BU5kFLYC!9vnBas5QftOycsXp-_PKXWm8UzNNF1qDp8pkqMNdapo

The game itself is really terrible, but the code shows how you can get a window that lets you draw graphics on the screen and how to control your framerate and make the game speed relative to your framerate.

I haven't tried codeacademy so I couldn't comment on it.

After you understand the C language, go on nehe.gamedev.net to learn basic graphics programming.

Is this resource good?
learncpp.com/

I'm flipping through it and it seems pretty solid. I would go for it.

I dont have any experience in C++ at all, I only know how to write C code, but I would definitely follow those tutorials, they look pretty well put together.

Speed. Make it where having a large front decreases your max speed.

Wait, wait, are you telling me that C++ and C are two completely different things? Isn't C just a basic C++?

Not I. It's a photorip of a friend's PSX Konami Justifier. Needs a LOT of polishing.
There was some coloring, tweaking, and touch-ups done by the MoM guy so that it actually looks like a thing usable in-game instead of a geek's hand, but there's still a loooong looooong way to go.


Likely because it's the exact same sprites from Action Doom 2.
Grezzo nicked graphics from everywhere, and I'm a placeholder whore.


GZDGPL with Freedoom as a base.
github.com/nashmuhandes/GZDoom-GPL
github.com/freedoom/freedoom

But that wouldn't happen in space because there's no friction/drag because it's a vacuum

Oh wait, you are the MoM guy.
Well now I look like a dumbass.

No, That's Terminus. I'm the MoM guy!

C came first. C++ is an extension of the language's philosophies and principles; the biggest difference is that C is imperative while C++ is object oriented.

They are two completely different languages, but have a lot of overlap.

Learning C should be a must, it forces you to:
-learn how to properly deal with memory since it has no auto garbage collector,
-use the low level language itself instead of a standard library function to solve most problems, since it has no auto garbage collector and it has a much smaller standard library compared to C++.

By default C++ does not have a garbage collector though I think they may be adding one to the standard library in a tentative future standard (or maybe just a standard interface). You can implement a smart pointer and do basic reference counting in a memory pool though.

i'll try and explain-

Basically C++ does a lot of advanced things for you that you will have to do yourself in C. C++ has a lot of features while C is more streamlined and only has the bare minimum, but its not more advanced in that everything that C++ implements for you, an experienced programmer can implement themselves in C.

You will learn more about how the machine works when using C because it doesn't let you use short-cuts (Like C++'s giant standard library). You will become more skilled, however it will be more difficult. Using C++ you can get away with a shallower understanding of problems.

Managed to find binaries for it because compiling is a pain in the ass on windows and there's no dependencies listed. The only tools I need are gzdoom builder and slade3, right?

...

Even if you don't use classes or the standard library, namespaces and template metaprogramming are extremely useful on their own.

Looks like google glass is making a comeback in the next 20 or so years.

They are, no doubt, but your post about the possibility of adding garbage collection reminded me of the comic. A lot of bloat has been added to the c++ standard and continues to be added, obviously you can use whatever set of features you want and ignore the rest, but that can become a problem once you have to work with someone else's code. I'm not denying that the features added are individually useful, but I think there comes a point where the idea of a particular language having a particular design paradigm is lost in the mass.
I just posted it for chuckles though.


kek

...

Story of my life. Took me forever to actually try C and it fit like a glove. That doesn't mean I think it's "the best" or would recommend it for everyone, I'm not really the language war type.
I even respect webdev, I don't know how anyone handles a full stack themselves, I couldn't keep track of all that shit.

Ahahaha I recently got the C++ reference book and this comic just about sums it up

For those who use the gzdoom engine, can you randomly generate maps with things like cellular automata or drunkard's walk ? Because I would like to create a roguelike.

Not exactly, but Oblige is an external program that generates levels, and ObHack is a lot more boxy and roguelike-y. It's possible you might have some luck there.

...

disgusting.
I hate that the standard is int blah[]
instead of the more clear int[] blah

The array should be part of the type, not part of the name. Also, if you really, really want to, you can grab pointers to shit in C# if you use a keyword to mark the block the code is in.
dotnetperls.com/unsafe

No, just the implication that not all kebab is removable, but this one very much is. Like a cheap selling point on advertisement in one of those spiky circles in a corner. New & Improved Formula, 7 Action Features, Removable Kebab and the likes.


Maybe it's just the specific angle used as the origin, where the feet look like they're swapping positions with some rotations.
Then again, if I recall correctly, that's sort of the look you were going for with all the other animations, so maybe it'd be inconspicuous enough. It does look a little better when zoomed in.

is that
apple kid: the game?

I prefer the consistency of declaring the array in the same format I use it in,

int[] blah = new int[2];
blah[0] = 1;

bothers me.
To me int[] blah implies an array of types of int all named blah where int blah[] implies an array of blahs of type int.

I like C# in general, switching gears just throws me off sometimes.

Is that article a joke? Using pointers makes your program slower? Directly accessing memory makes the program slower?

That kind of attitude is for retards who have no idea how the underlying system works and need everything spoon-fed to them through a layer of abstraction. Languages that try and hide features because bad programmers will screw up trying to use them are always going to limit you from getting good at programming. Its like refusing to learn how to use an important tool because you wont get it right the first time. By refusing to use it you wont get it right at all.

An array type is less clear than a pointer- With a separate type for pointers to arrays it makes it look like pointers to arrays are somehow different from other pointers, when they aren't… the form: "int blah[]" tells you that there are more ints next to the int pointed to by blah, but blah is just a pointer to an int, that happens to have several other allocated ints right next to it in memory, but not a special type. but "int[] blah" suggests that a list of ints is its own generic type, when its really just a pointer.

Is that post a joke? Using arrays is less clear than a pointer? That kind of attitude is for retards who have no idea how the underlying system works and need everything spoon-fed to them

I suspect in the context of C# directly accessing memory can make the program slower by befuddling the compiler. I'm mostly guessing though.

Not the user that asked, but thats really awesome. I didn't know such a thing was possible.

Just imagine if you hacked together random level generation, plus whatever crazy open world shit they have going on here: youtu.be/HUxqfVSwSz4

you could get an open world doom roguelike. that could be too fucking cool.

Well nevermind then, I'm just going to try to code some raycasting engine myself. Does anyone knows how can I render multiple layers like pics related with raycasting ?

I've been working on replacement coin sprites for Space Pirate.
A friend made and sent me the (second row) left-most coin but couldn't do any more, so I tried making a full rotation for it using the top-row as a reference.
And then making a variety of different coins for different varying amounts.

Now to actually implement them.

Boy this was a day of abusing Photoshop in various unique and interesting ways.
On an unrelated note, FUUUUUUUUUUUUUUUUUCK

Does this work for you?

I plead the fifth

youtube.com/watch?v=xjUHMsAZti4

Looks pretty good, famicom.

user who wrote them here (made more, those pics are outdated and have mistakes). I had to deal with exams so I didn't get to do anything fun in the past 5 weeks really. Now I'm slowly chipping away at a list of things that has piled up over time. The /agdg/ guide, my own game's code, the GDD for my game for pitching it to a potential partner, practicing my writing skills, issues with one of my open source Unity assets some people reported, potentially serious health issues.
Yeah, it'll take some time before I finish it.


The reason for that is that C/C++ arrays are not a type, but merely a set of elements of a certain type. Your C#/Java/whatever array-class is just a wrapper for such an array. So if I chose to describe these ones instead, it would have gone something like this:

I agree with you when it comes to readability though. I like my languages to be unambiguous. The fact that int* may be a pointer to a single integer, as well as a pointer to some element within an array pisses me off in a similar way that the reference vs. value type shenanigans of C# do. (Not nearly as much though.)
C++ has a billion features. A specialized syntax for arrays that makes it crystal clear that a certain variable is one would be great. It's less about understanding my own code, and more about understanding someone else's.


That's weird yeah.
My best guess would be that a) you can lose more than you gain by messing with the compiler that tries to optimize your code like said and b) that there still is some overhead.
Those runtime environments are huge and complex beasts. The bytecode/intermediate language code you put into it, is quite far removed from the actual operations your CPU does in the end.

Would any of you be interested in messing around with a MUD server on a casual level. Nakedmud is what I'm looking at.

Should I into gamedev /agdg/? I have some cursory knowledge of programming, and mess around with scripts and Loonix but that's about it. Want something more (((productive))) to do on my computer when I'm not playing games than shitpost and generally waste time.

No sorry, if you're not experienced by now then you can't do it anymore. Better luck next life.

anyone got a clue how i would implement this mechanic in unity?
html5xonix.appspot.com/
i need to clear away parts of the screen
the way i understand it, there would be black plane on the front, and whenever i connect some lines, i'd be cutting part of that plane out to modify the mesh

good luck

serious answer: why dont you try a few tutorials and try to see if you can see yourself getting some enjoyment out of developing something yourself?

Like 95% of the projectiles don't show because of the compression from the converter to webm I used. There's roughly 4,000-10,000 projectiles on screen at any given time there, but they don't show. Can anyone recommend a way to convert from mp4 to webm that won't lose so much detail?

Anyway, this is a prototype I just finished that I thought I'd share.
I wanted to see if I could get thousands of units and tens of thousands of projectiles working very efficiently.
The tracers suck visually, and they just disappear without any explosion effect when they run out of health, and the behavior of the units isn't very human, but the things I wanted to see if I could efficiently make work are there: Allies avoid colliding with each other, but can end up colliding with and bouncing off enemies, and the AI has some high fidelity to it such as how each unit will shoot at targets it's gotten a lead on instead of just selecting one closest target to try and shoot, and units avoid shooting allies.
On desktop hardware, I think I can get more around 50,000-100,000+ ships and hundreds of thousands of projectiles. Not positive yet.

I used OpenGL ES 2.0 here, which was a fucking mistake. This would have been a lot easier with GL ES 3.0 and that's what the hardware I'm targeting uses anyway.

particle storms ruin video quality… maybe you should upload to another site without the limit

What is the limit on Holla Forums?

My big problem is converting. My source video is 80mb for 16 seconds, and it still doesn't look nearly as good as the prototype rendering.
I used an online webm converter and it pooped on the quality while making it 7mb.

Read the post limits, famalam. It's 8 megs.

8 MB. Try using something like webm for retards with two-pass encoding and all that if you cant link to the source video.

You'll probably get better results with a very basic ffmpeg command.

Use this if you're on winderp
Takes a long time to render but you can spend that time making more progress anyway

Converted with ffmpeg. I dropped it from 60fps to 30fps but it's still 55mb.
webm.land/media/TBeC.webm
But it looks decent this way. Not nearly like it looks when it's rendering from the prototype, but a lot better than the previous 7mb video.

I could upload the 60fps but it's 90mb, so, meh. The 30fps doesn't look much different. It's the compression fucking it up, not the framerate. And that my computer couldn't handle anything higher quality than "veryfast" on OBS, so my source video is pretty shit to start with.

Probably if I did raw video to webm, it'd be much better, but you guys get the idea now anyway.

being able to index any type that has indexers defined the same way is nice. Seeing

var crap = new Dictionary();
crap["fuck"] = "shit";

makes sense, even though it's not an array, the syntactic sugar for indexing is always the same. C# even allows for multidimensional indexing, and there's a difference between

int[][] blah1; //array of arrays of ints (jagged array)
int[,] blah2; //actual 2d array of ints

That, and pretty much every language that uses the C/C++ style arrays forces them to be fixed size unless you're using a pointer.

int size = 5;
int blah[size];

won't work in C/C++, and it drives me flipping bonkers, and it's a part of the way the compiler works. When creating arrays, the compile _NEEDS_ to know the size of the array ahead of time, because it will create a blank area in the program memory that many with a certain size, and it can't decide that at runtime.


C# is compiled into bytecode, and JIT compiled into assembly, and as much as everyone hates Micro$oft, they have a killer R&D department, and there's enough optimization baked into the C# compilers compared to the C++ compilers that the speed difference is rarely significant.

How the hell are arrays less clear than pointers? You know exactly what you're getting, either an int or an int[]. With a pointer, an int* could be either one int, or many ints, and you have no way of knowing in cases that you don't have control over that value, which is why there's the ass-backwards Hungarian Notation for variable names thats so prevalant in C/C++.

You might as well just code directly in assembly if you don't want to hide how the underlying systems work and need anything spoonfed to you through languages that are more similar to spoken/written words.

in c++ you could just do int size=5;int *arr=new int[size];
although i'm not sure how it messes with the memory

Dynamic arrays are generally extremely expensive.

Take JS for example that uses them, changing an array size and using that dynamism is very costly. Your code will be faster if you avoid manipulating their size.

Supposedly they're fairly fast in D, but I'd bet avoiding using them is still faster.

competing with pajeets destroys my spirit

anyone know where I can peddle my ability to program the back-ends of websites, or unity extensions?

It seems all anyone wants these days is a webs.com template done for $5

that eye tutorial is pretty fucking horrendous, but everything else is okay

but then it's just a pointer, not strictly an array.


It's not about dynamic, resizing arrays, but about creating arrays that have a size determined at run time, rather than compile time.

Just use vectors and resize() them

No, it allocated space for 5 ints in sequence. New just returns a pointer to the first element. If you do arr++ it's a pointer to the second element.
tutorialspoint.com/cprogramming/c_pointer_to_an_array.htm

poor pajeet, angry chinks don't want poo in the loo

googling 'never accept indian' gives interesting results

here's the character that causes google to freak the fuck out

anything we can do with it?

(ด)(ด)

Show portfolio

never accept indian

It's not one character, each of those things above it are separate characters. And Holla Forums automatically removes said characters, so posting it here is useless.

what am i doing wrong here
i'm making a mesh by connecting the starting vectors of a few line renderers
but when i do that, sometimes the plane is flipped
i guess the issue is with the triangles in it, but i thought i should be able to fix it by setting all the normals to (0,0,-1)
but that changed absolutely nothing, it still randomly chooses whether to flip the plane or not
this is what SHOULD have fixed it but didn't
Vector3[] norms = PlaneMesh.mesh.normals; Vector3 norm = new Vector3(0, 0, -1); for (int i = 0; i < norms.Length; i++) { norms[i] = norm; Debug.LogError(norms[i]); } PlaneMesh.mesh.normals = norms;
and this is what actually makes the mesh
Vector3[] vertices; //check if it's just one line, or a shape if (lines.Count > 1) { vertices = new Vector3[lines.Count + 2]; for (int i = 0; i < lines.Count; i++) { vertices[i] = lines[i].transform.position; } Vector3 lastPosition = Player.position; if (lastDir == Vector3.up || lastDir == Vector3.up * -1) { lastPosition = new Vector3(vertices[0].x, lastPosition.y); } else { lastPosition = new Vector3(lastPosition.x, vertices[0].y); } vertices[lines.Count] = Player.position; vertices[lines.Count + 1] = lastPosition; } else { vertices = new Vector3[4]; if (lastDir == Vector3.up || lastDir == Vector3.up * -1) { vertices[0] = new Vector3(Player.position.x + 20, Player.position.y, 0); vertices[1] = new Vector3(Player.position.x - 20, Player.position.y, 0); vertices[2] = new Vector3(lines[0].transform.position.x - 20, lines[0].transform.position.y, 0); vertices[3] = new Vector3(lines[0].transform.position.x + 20, lines[0].transform.position.y, 0); } else { vertices[0] = new Vector3(Player.position.x, Player.position.y + 20, 0); vertices[1] = new Vector3(Player.position.x, Player.position.y - 20, 0); vertices[2] = new Vector3(lines[0].transform.position.x, lines[0].transform.position.y - 20, 0); vertices[3] = new Vector3(lines[0].transform.position.x, lines[0].transform.position.y + 20, 0); } } Triangulator tr = new Triangulator(vertices); int[] indices = tr.Triangulate(); // Create the mesh Mesh msh = new Mesh(); msh.vertices = vertices; msh.triangles = indices; msh.RecalculateNormals(); msh.RecalculateBounds(); if (PlaneMesh.mesh != null) { CombineInstance[] combine = new CombineInstance[2]; combine[0].mesh = PlaneMesh.sharedMesh; combine[0].transform = PlaneMesh.transform.localToWorldMatrix; combine[1].mesh = msh; combine[1].transform = PlaneMesh.transform.localToWorldMatrix; Destroy(PlaneMesh.mesh); PlaneMesh.mesh = new Mesh(); PlaneMesh.mesh.CombineMeshes(combine); } else { PlaneMesh.mesh = msh; }

forgot video

and if normals aren't the issue, i'm certain that the issue would be in the triangles, since i'm using a script from the wiki to calculate them
wiki.unity3d.com/index.php?title=Triangulator
but i'm not sure what i'm doing wrong with it

nevermind, somehow this fixes it
Mesh msh = new Mesh(); msh.vertices = vertices; msh.triangles = indices; msh.RecalculateNormals(); msh.RecalculateBounds(); Vector3[] norms = msh.normals; Vector3 norm = new Vector3(0, 0, -1); for(int i=0;i

That's the main issue I always come across when coding in C++
It's a great language, don't get me wrong, but as u said "I like my languages to be unambiguous".


It's a fun hobby, and imo if you want to get into it think of it like that; i.e. not as some chore but a fun passtime.
When your working on your vidya it should be the fun of it, in addition to what you want to do, and not because "I want to make 18 quintillion dolalrs xDD"; else you end up with yandev (wants to please everyone), or most AAA dev out there (muh 60million shekels goyim).

It's easy enough to identify which dev is which when an user posts progress.
Go to reddit if you want validation of your identity, pseudo-user

direction of the winding order of vertices, user
Unity uses clockwise winding order, or else the face gets culled; i.e. it's not the normals.

well, i somehow got it working
now my issue is how to detect whether the line has entered the existing mesh. currently i'm just checking if it's near the border of the screen
i'd imagine that using a mesh collider would be costly
is there any other way to check if a vector is within a mesh?

Still broken on React. Try this


No, Unity does triangles counter clockwise.
Culling issues are normal issues, making a face the wrong way will flip the normal.

Actually scratch that part about static linking, it shouldn't be necessary.

Platform toolset is set to XP specifically, the CRT is static linked… Its not being built in debug mode either.

I didnt know about the debug setting in the linker, I turned it all off in the linker settings, so here is a build for you to test, see if it works: p.trumpshare.com/38tg/mauo.zip

Its got some other features in it that I was working on for but I haven't finished them.

Wrong.
Front facing triangles are achieved via clockwise ordering of vertices.
Counter clockwise achieves back facing triangles.

Wrong, again, it doesn't contribute to the issue of improper culling (winding order); it's the result of improper winding order.
Not having normals, or even having normals in the wrong direction results in the triangle having the default "pink" display of RGB values via the fragment shader.
It does not have to do with culling, but it does have to do with the fragment shader applying the correct colors per pixel (for a frag); in the right orientation for the camera's view projection matrix.

I just finished the intro of programming without skimming/BSing things. took me 1 day and a half.
my ass and writing finger hurts

but I'm really worried about what the hell I'm going to do for a living. I dropped out of college due to depression issues and I don't know how to fix my mental issues. I've been in therapy for a few years and still I feel like shit
/blog

gj

check if the point is within the mesh's bounds

also, if you run into that issue again, in the shader just use:
Cull off

To make sure your script is working, and then you can determine if it's a culling issue; or something else entirely.

how?
i also have to mention that the mesh is 2d, so it doesn't have a volume

look at docs

the closest thing i can find is docs.unity3d.com/ScriptReference/Bounds.Contains.html
and in that case i'm checking if my dot is within the bounding box, not the mesh itself

>look at docs more thoroughly

gee user, that's very informative advice
you deserve an award for your academical achievements

learn how the mesh class works, and then you can figure it out yourself; is all I'm saying.

What does this do: mesh.RecalculateBounds()?
What does this imply?
Look at the classes that the mesh always uses… then you'll know exactly what I'm saying.

recalculatebounds figures out the bounding box of the mesh
in other words the green shit in this picture
which is not what i'm asking

bool pointInPolygon(Vector3 point, Mesh polygon){ int i, j=polygon.vertices-1; bool oddNodes=false; for (i=0; i=point.y || polygon.vertices[j].y< point.y && polygon.vertices[i].y>=point.y) && (polygon.vertices[i].x

actually i got it working with this
bool SameSide(Vector2 p1, Vector2 p2, Vector2 a, Vector2 b) { Vector3 cp1 = Vector3.Cross(b - a, p1 - a); Vector3 cp2 = Vector3.Cross(b - a, p2 - a); return Vector3.Dot(cp1, cp2) >= 0; } bool PointInTriangle(Vector2 p, Vector2 a, Vector2 b, Vector2 c) { return SameSide(p, a, b, c) && SameSide(p, b, a, c) && SameSide(p, c, a, b); } bool WithinBounds(Vector3 pos) { Mesh m = PlaneMesh.mesh; int triangleCount = m.triangles.Length / 3; for (int i = 0; i < triangleCount; i++) { Vector3 V1 = m.vertices[m.triangles[i * 3]]; Vector3 V2 = m.vertices[m.triangles[i * 3 + 1]]; Vector3 V3 = m.vertices[m.triangles[i * 3 + 2]]; if (PointInTriangle(pos, V1, V2, V3)) return true; } return false; }
checks every triangle of the mesh, does some magic with dot/cross products that i don't understand because i failed algebra and geometry, and determines if the dot is within any of the triangles

although now i think i need to also check if it intersects with previous lines, because otherwise when the shape is finished i get triangular stuff which i sort of don't want
and i also need to get it to fill some spots that appear when you make a few colliding shapes, that wasn't an issue at first since i was just colliding with the borders of the screen

I think you're supposed to die if you hit your own line, so self intersection should be impossible.

that's what I do if a mechanic is too hard to program

you don't need a trip for us to identify you
you have a very particular brand of autism

in that case this is what i have left to do:
making a single line should clear half of the screen, depending on which sides are closer
making a ball bounce off of the walls, and kill the player if it touches his lines. that part should be easy, i check if it collides with the mesh like the player, and then i need to make something to check if it collides with the player lines

Removing the debug (by removing _DEBUG from the cl) would only help if it refused to compile because of static linking. Obviously, it's still broken.

You don't need to post builds to test this by the way, just open the exe with a hex editor and look for the InitializeCriticalSectionEx string. If it's present, that means it still expects the Vista kernel.

I was referring to using the mesh's bounding box, which is recalculated via mesh.RecalculateBounds()
So, each mesh has their own aabb (axis aligned bounding box), and doesn't require a collider or something else to provide this functionality (intrinsically instantiates an instance of this by the mesh class).
You could use this to see if your new mesh's aabb intersects with the bounding box of your stored mesh's bounding boxes via a quick conditional:
bool insideOfMesh = false;if (mesh.bounds.Intersects(othermesh.bounds)) insideOfMesh = true;

Although, it is a box, so you'd run into issues with having disproportionate shapes w/boxes that overlap your other mesh's bounding boxes thus giving u inaccurate results (as it can extend beyond the actual mesh size, if the mesh is disproportionate, due to the point that's there's one aabb w/center + extends w/min/max… and that it's one box over the extent of that whole mesh… oekaki included).
Which could be solved via making those disproportionate meshes into sections of individual meshes instead of one big mesh, and implemented as a pre-processing step when creating said disproportionate meshes.
I.e. create meshes that are always square in their shape as to conform to the constrains of the bounding box.
So, two triangles that share edge verts would meet this conditional for only making squares, and are then made into its own mesh w/its own bounding box.
Rinse repeat for other verts.

You could also just use the bounds.Contains to determine, for each of your new mesh's individual verts, if they're within your old mesh's aabb.

Although, just using the aabb comparison is more efficient as it's less granular for the amount of steps/comparisons necessary.
Also, note that aabb are purpose built to accomplish what you're trying to achieve.

I had no idea you could do this-

I found it, its near a bunch of other windows functions. Thanks for the help, i'll track it down a lot faster now.

i don't get it, both renderer and mesh use Bounds, which is just one cube that encompasses the entire mesh
docs.unity3d.com/ScriptReference/Mesh-bounds.html
docs.unity3d.com/ScriptReference/Renderer-bounds.html
and the mesh begins as the borders near the screen, with planes added later on in the interior of that border
so to begin with, the bounding box would already be overlapping with the player, which makes it pointless to even bother checking

oh wait, you want me to have multiple meshes so that i can just check their bounds?
because the way i'm doing it right now, i add the new mesh to the existing one (although i'm not sure if it's necessary, i just prefer not having a trillion gameobjects for every time i make a new plane)

By the way, your dependencies could also be responsible.
stackoverflow.com/questions/22709094/c-application-built-with-120-xp-toolset-on-windows-xp-initializecriticalsecti

However, I checked OpenAL and it seems to be NT5 compatible. Which VS version are you on exactly?

NOTE:
—-

That's because (as u know) you're using one huge mesh, and adding to that; not individual meshes w/their own aabb.
You could later combine these, or you can get around the main issue of having a draw call per mesh/tex via via using a shared material.


True, as it's a bit of pia to have that many GOs at once, but u can just make their transform a child of a "holder" parent GO in your scene hierarchy (you can individually debug each thing you draw too, which is a nice side effect).
Although, imo, this is the easiest solution, and later u can do that same combine mesh method for all the little meshes once you're done drawing them with your mouse.

2015 community. I'm building with the 140_XP toolset. These are all the .lib files that I am using:
opengl32.lib
glu32.lib
Gdi32.lib
user32.lib
libwebp.lib
OpenAL32.lib
libogg_static.lib
opus.lib
opusfile.lib

I haven't had a chance to go through these yet.

I also have a copy of vs 2003, which may work, but I havent tried.

yes, exactly

2003 will definitely work
140_XP toolset should work
Has to be one or several of the libs then. I'd suspect libwebp if it is what I think it is.

its the library from google for decoding .webp images. I'll check the code and see… I do need to go to dinner so I cant fix it right away, but I will be back to fix it.

oh, never-mind, its definitely libwebp doing it, I found it in there. None of my other libs have it.

I will either re-compile it for XP or if that doesnt work change my texture format to .flif.

So, when I can get the next build out, your problem will hopefully be fixed.

You could just rebuild the library.

well fuck me I can't read

re-compile means the same as rebuild… right?

Working on it…

new bread when?

Is it possible to make a strictly 2-D game using Unity 3D, with all the framerate-saving benefits that go with it? I'm looking for a sort of NSMB-like game style with 2-D backgrounds and 3-D characters.

No

What would be better? Gamemaker Pro? I am not smart enough to learn C+ from scratch and would like a program that offers assistance.

In my game, timing is extremely important so I need it to run smooth even on toasters. Is it possible to write code that will skip frames in animation in order to sync with other things exactly?

GM:S is pretty alright (using it right now for a 'droid game) but 3D in it is a bitch unless you're gonna try to Donkey Kong it. I hear Godot is good but I cannot for the life of me figure out how to get lower-level control over stuff in it, like pixel-perfect movement.

Thanks for the suggestion.
I would like to Donkey Kong it up, I always liked that aesthetic.
If we're going for a retro style for indie games, they might as well look gorgeous.

Also, is there a way to automatically set evenly-spaced frame skips in animations if I wanted a character to speed up or slow down?

Keep in mind I wouldn't like to increase the framerate because then the jumping and running would be way too fast, and I'm having it set at 60 anyway.

image_speed in GM:S does that IIRC.

the speed of the game should not be framerate dependent. multiply the velocity of the thing that moves by the amount of time it took to draw a frame and it will not be dependent on framerate.

Thanks!

GM:S is kinda shit like that, but if he isn't much of a programmer then trying to get delta-t working might be a bit hard.

If I can just fix this final little bug with ceiling collisions I'll give you my kinda-bad-but-still-educational platforming experiment if you want it.

dropbox.com/s/cnk3zr891i2q6te/PlatformingTest.gmz?dl=0

Ok so after a lot of code cleanup, if you want to see how some basic platforming physics work in GML, try this. Webm related, it's the thing.

new bread

Yeah, it's just a pointer, like any other int pointer, and can be used in all the ways an int pointer can be used, not strictly like an array, which remembers its length.

Might wanna be careful with that mag back there, looks like it's clipping into the side. Starting to get hype now. Hope it has been a good learning experience.

Woops