976829 No.14713738
Probably the last thread before demo day edition
Resources
>>>/agdg/
>>>/vm/
>#8/agdg/ via irc.rizon.net
Links
>Wiki: http://8agdg.wikidot.com/
>Beginner's guide: >>>/agdg/29080
>Previous thread: >>14692875
QUARTERLY DEMO DAY SCHEDULED FOR MAY 5TH
Polite reminder that the wiki exists, you are encouraged to contribute to it if you can (even if it's just your game page)
ac42f8 No.14713752
I started learning HLSL, made a shader that mixes textures with vertex color and height maps.
9af522 No.14713766
Paired programming was a mistake.
976829 No.14713781
>>14713766
>paired programming is an actual thing that exists
>as in permanent teams
Jesus fuck. I sometimes pair up with one of my co workers to help them solve a bug, but why the fuck would you waste all those manhours having two people work on every single little thing?
ef948c No.14713792
>>14713766
Not really, it's more accurate to say that it's two mistakes :^)))))
9af522 No.14713841
>>14713781
It's not the same 2 people every day. It's different pairs every day. Essentially it's so that everyone will know what's going on, so one person doesn't write a gigantic block of code and no one else understands it. Even though using smart comments and not writing spaghetti code would solve it way better. However, when you outsource more than half of your coders to the 3rd world this doesn't work.
e681bb No.14713891
>>14713752
get yourself a moss texture instead of grass and that will look pretty nice
81d1d4 No.14713973
>>14713841
>>14713781
Because you fucking code monkeys don't know how to comment or document your shit. I've worked with more pajeets with better skills than the pathetic basement dwellers this country shits out into IT. At least they know how to document and explain their code rather than shrugging their shoulders when someone asks them about the abortion sitting in the release that was supposed to be "good to go".
2adb21 No.14714106
>>14713268
Not crabdev, I'm just combing through the archive:
https://web.archive.org/web/20151215074517/http://8ch.net/agdg/res/18336.html
I was the lost post in the thread ;-;
2adb21 No.14714109
8c0cb1 No.14714187
Implemented death and disengaging from battles to return to the field. Mostly need to set up some rough AI and start bugfixing and making skills for the next Demo Day. Third species in coming in, so that'll be nice for demonstrating some stuff. I'm not really sure how practical, for lack of a better word, this demo will be, as it's pretty bare. Mostly there to give an idea of how the systems work and let people click buttons, but it's very small in scope. Still teetering between that or just a tutorial in a video.
cf2611 No.14714204
>>14713752
What >>14713891
said, but it looks good nonetheless. Can you add noise generation or something to the terrain so it doesn't look so flat?
9af522 No.14714215
>>14713973
>I've worked with more pajeets with better skills than the pathetic basement dwellers this country shits out into IT.
I find it hilarious you bring up documentation because the people pushing this paired programming shit also say you should have 0 comments. And 3rd world contractors are the worst coders by far. Even so far that the dipshit college intern who barely knows Java can spot contractor code from employee code. Only the best they have to offer is flown over to the states, the ones you're probably describing, and even then they're mediocre.
8949f9 No.14714813
>>14713973
>he woke up to defend the honor of subhuman street shitters
f7a57c No.14714946
>>14714106
doing a great job, thank you. be sure to embed.webms, like ArcheoDev (RIP, didn't see him edit anything in a while) did for Waifu Souls http://8agdg.wikidot.com/waifu-souls and the images you added
f7a57c No.14714974
>>14714187
bro, could you send us the original YT vid of your original concept for history's sake?
1563cf No.14715158
>>14713973
GOOD CODE IS SELF DOCUMENTING CODE
d11af5 No.14715181
>>14715158
That's a nice way to say "I can't be arsed to write documentation", anon.
But let's be real here. This is gamedev. The standards for software development are pathetically low here. There's tons of developers out there crashing down because the fuckers didn't even use fucking version control, let alone regular backups.
8949f9 No.14715192
>>14715181
// create an array of strings
scores = "Mark is a fucking nigger".Split(" "); // create a string and split it by a space
int score = 13000; // set the score high
// loop through all of the scores
foreach (var c in scores) {
//sleep for 10 seconds
System.Threading.Thread.Sleep(10000);
}
ef948c No.14715198
>>14715181
>The standards for software development are pathetically low here.
That's because they're pathetically low everywhere.
97ba77 No.14715211
>>14715181
If your code needs a comment to explain it, you've probably written shit code.
4ed8db No.14715268
>>14715211
This tbh.
Documenting your declarations is one thing, but the idea that you should have comments all over your definitions is just bad. If a programmer is unable to express himself clearly with his programming language of choice, why should I assume he'll be able to express himself in English? They come from the same confused mind. However, the more important problem that some people like to forget is that comments aren't compiler/interpreter checked. They are not guaranteed to reflect what's truly happening.
1563cf No.14715278
>>14715211
>>14715268
Don't bother commenting function logic, either. That's an implementation detail that the end user doesn't need to worry about :)
8ee087 No.14715283
>>14715211
>>14715278
Unless you're writing in your own engine that you will personally open source, commenting if for your sake only.
d11af5 No.14715291
>>14715211
>>14715268
The reason why "self documenting code" is a fucking meme is because people use it as an excuse to fore-go writing actual documentation of how the program as a whole functions at a higher level.
How an algorithm works, or possibly deviates from the normal implementation? What the rough control flow of the program is? How it's structured?
Nah mate, our code is so good, it's self-documenting! But don't worry, we're working on adding doxygen documentation! That's sure to help you out!
4ed8db No.14715332
>>14715291
Okay, that's a perfectly valid point of criticism right there. As I pointed out, I have comments that document my definitions. They explain the input and output data as well as what's happening. That's restricted to the headers though. My definitions are mostly comment free. With some exceptions (like magic numbers for example). If something really needs clarification, I add a comment.
On top of that I maintain a wiki that contains the classes that will be reused later and more general stuff like the rough control flow you mentioned.
Still, the same way that bad programmers use "my code is self-documenting" as an excuse to be lazy and not write any documentation whatsoever, other bad programmers use "I'll explain it in the documentation" to excuse their inexpressive code. It's little different from "I'll refactor it later" or eternal "TODO" comments.
8c0cb1 No.14715625
>>14714974
I un-privatized the video. Didn't think anyone would've been interested. https://hooktube.com/QwIzqB-Q8HY
>>14715291
>>14715332
I do the self-commenting code stuff and usually just comment TODO to make a note of changes later. Most my documentation is high level stuff like how the systems interact. Things you can't really gauge easily by just reading code. Since I work independent, my notes and docs are minimal and serve as reminders though. Group work, I go ahead and document more detail or delegate the work to someone else and check for approving it.
0869de No.14715637
Does anyone have those pixel art guide pdfs?
976829 No.14715648
0869de No.14715654
>>14715648
I don't really want to have to sign up to read them.
679787 No.14715698
>>14715654
do you have to?
0869de No.14715798
>>14715703
You have to sign in to download it i mean.
5b7784 No.14716049
>>14715654
Have some, fag.
b4f518 No.14716072
I would like to 3d model but the scaling of size of items really throws me off.
Do you guys have any decent guides?
0869de No.14716077
9af522 No.14716255
>>14714813
I think that dude was a bait or something. I work at a fairly progressive diversity touting place and it's still well known that the third world contractors are fucking awful spaghetti coders, even to the non-coders. They're slowly being phased out now.
>>14715158
>>14715211
>>14715268
>>14715283
>Good code self-documents
It's "Don't prematurely optimize" all over again.
b6c9d8 No.14716524
I peek in these threads every so often and you lot truly impress me with your dedication to learning your craft. Here's a video I said I'd make for you all last thread.
8c0cb1 No.14716578
Today I'd been focusing mostly on some under-the-hood stuff and putting some skill scripts in for demo day. Major things otherwise were the new species and putting in a rough display for skill descriptions and costs in battle for reference. None of this UI is permanent, so I'm not too attached to cleanliness right now.
ec76a9 No.14716590
holy fuck I'm so burnt out and dead and my brain don't work but I'm so motivated to work on my game that I wanna get back to it after like an hour worth of break. Please spare me, motivation-sama
>>14715291
I add a single comment for every logical code snippet, no matter what it is. No matter what folks say, even in high-level languages code is not made to be read, it's made to be written easily. You shouldn't subject yourself or anyone else to the torture of reading broken gobbledeegook that is code. Even if it's "self-documenting" or whatever. Writing comments saves you time when revisiting old code. You will be reading it like a stranger even a week from the time you've written it.
There's zero fucking reason to be lazy when writing code. If you're doing something, commit to it with full discipline, focus and care. Even if you're outlining code when trying to figure out how to do something, you can write temporary sloppy shit, but if you do that - be sure to commit and refactor/rename/restructure/encapsulate things that need to be. It's about self-respect and personal standard to which you keep your work. Don't be a fucking pajeet.
>>14716524
Thanks, anon. I think gamedev is the one thing I could never give up - I just keep gravitating back to it, not matter how hard it gets. Feels like I found something I can do until the day i die!
8ee087 No.14716635
Any good resources on enginedev? Is there any documentation for Godot's GDnative? What are some good 3D lightweight frameworks for C/C++?
c15ef0 No.14716639
5f315c No.14716670
>>14713959
It works, once again, but try to keep the head consistent, or you're just going to keep on getting that "shrinking Popeye" problem.
Also, as said before, try to adjust the frames and timing. You only need like, 3 "original" frames for that animation
9af522 No.14716675
>>14716635
>Any good resources on enginedev?
Watching conferences by devs who actually do high-performance coding like Mike Acton. The vast majority of shit you will hear about C++ is slow garbage. If someone even hints at using boost disregard their opinion immediately.
1563cf No.14716678
>C# has a dynamic keyword
>It behaves just like Object, except that it defers type checking to run time, rather than compile time
>object o = 10;
>o += 10; // invalid
>dynamic d = 10;
>d += 10;
>d = "niggers"; //valid
I feel like it would be fairly slow to use, but maybe that might be useful for a shitty script system
8ee087 No.14716683
>>14716675
Any books or online tutorials?
253087 No.14716709
>>14716635
>>14716683
Enginedeving is not the kind of thing you learn by reading books or tutorials about it. It's all about figuring out how to tie a bunch of otherwise disconnected systems and algorithms together into a single whole, and it's best learned through a mixture of experience and studying others' code. Definitely look into the id Tech engines and the assorted stuff on http://bitsquid.blogspot.com .
1563cf No.14716741
What's a better design?
>Gamestate has all these variables but just sits there
>Controllers attach to it and update the variables, depending on the system but they themselves probably don't store any meaningful state
>Gamestate has no values of its own, but a collection of references to each system/controller
>Controllers contain their own pertinent data, and a reference back to the gamestate so they can communicate with each other
ec76a9 No.14716744
>>14716741
data is always better to keep as consolidated as possible.
8ee087 No.14716766
>>14716709
I see. Well, is there any material on OpenGL rendering?
1563cf No.14716768
>>14716744
So the first one, then?
Basically I'll have several arrays (two or three 128x128 jagged arrays) plus a handful of misc state, such as timers, etc. I kinda like the controller approach because that means I can just have one large state object and have things operate on it without coupling
ec76a9 No.14716780
>>14716768
Well to expand on that you need to keep it consolidated but in logical clusters. Since you have multiple controllers with their own related data, I'd say it's better to keep track of individual "pawn" objects of some sort that holds the data for the controller while the controller itself remains there purely for input and similar low-level stuff.
2adb21 No.14716956
>>14714946
I'm glad someone appreciates it. I'll probably do the crab game next.
96c193 No.14717323
Added physics for her hair strands and fiddled with colliders. Still yet to use anything from the asset store. I will try to not use anything from it.
ac42f8 No.14717677
>>14713891
>>14714204
It's all just placeholders to test the shader out.
47c66c No.14717693
>>14717677
>posts work
>receive criticism
>ignores criticism
you are headed to Dobsonhood
976829 No.14717718
>>14717693
>posts work
>receives criticism about something that isn't the work he's doing
>ignores criticism
Wow!
9af522 No.14717748
>>14716683
Read about data-oriented design. It's the only guaranteed way you'll dig up performance based code and learn about CPU-cache. Most books are shit, all online tutorials are shit. IDK why, but all opengl tutorials start you off with glBegin glEnd, which is really fucking outdated since people use shaders now.
1b1a7d No.14717982
>>14717748
>glBegin glEnd
I usually point people towards Anton Gerdelans OpenGL tutorials. His book is good too but the tutorials on his site is all you need to get started.
http://antongerdelan.net/opengl/
96021c No.14718613
getting ready for demo day, you guys get a special door
1563cf No.14718811
>14717748
My code is simple and scale is small enough that performant code isn't a necessity. Or rather, I won't need to worry about things like cache misses and data density, but I do try to consider it and avoid huge awful mistakes. Basically as long as my CPU use is low and memory isn't leaking, I'm happy.
Also I'm with C#, so I assume the compiler/JITter looks after most of that shit for me.
1563cf No.14718820
>>14718811
Meant for >>14717748
Please be patient, I have JS disabled because the site chokes itself on dicks, and pages won't load otherwise.
40e122 No.14718843
>>14718613
Glad to see you're still working on this, we can always use more spooky games.
8949f9 No.14719263
>>14717693
>make placeholders
>geniuses come up with ideas on how to fix the placeholders
d1437c No.14719334
>>14715192
>obvious comments are useless
>therefore all comments are useless
Retarded/10
8949f9 No.14719347
>>14719334
Most comments are useless unless you're a pajeet. Even then, pajeet English is probably harder to read than whatever code they shit out.
4b8cc5 No.14719424
>>14716766
There's http://learnopengl.com/ which is a very useful resource but it only really teaches you what commands do what. There are a few game engine youtube tutorials which is easy to find if you know what to look for. Personally, I recommend using one of these video tutorials to start off with if you're not too sure how it all works and later on split from that and look into doing your own shit it's what I ended up doing. Also, check the fucking wiki, you troglodyte.
>>14718613
Good shit. Can't wait to see what kind of gay shit is behind the door.
f761e4 No.14720046
>>14716635
Don't do it.
I watched two projects die because of engine troubles. The second was thanks to my own incompetence. Then I started using Unity and it was done in 3 weeks, where it took 3 years going nowhere before.
I'm not saying you should use Unity, but enginedev is a shitty, pointless task. You can't outperform commercial-grade engines with teams of god knows how many code monkeys and millions of dollars of budget. I don't know why it took me so long to realize this. Don't try to develop your own engine unless you are a godlike coder, it will only end in misery and failure.
ddb742 No.14720278
>>14720046
> Don't try to develop your own engine unless you are a godlike coder, it will only end in misery and failure.
Your advice is sound, but you shouldn't have included this line. It will only cause anons to think: "So I will become a godlike coder, it's not like I have anything else to do".
d1437c No.14720299
>>14719347
Comments are for explaining why you're doing something the way you are and what gotchas or concessions are involved. No shit obvious comments are useless.
330623 No.14720375
>>14720046
>>14720278
making an engine is something you do when one of the existing engines cannot possibly do what you want it to do
learning to blit text on the screen and get polygons to not vomit themselves inside out isn't how you want to spend the first year of your passion project
don't do something technical unless you hit a reason, that way you always have a "because"
8949f9 No.14720508
>>14720299
You're very close to understanding, yet too stupid to let yourself
52a350 No.14720556
>>14720548
Faggot / 10. Load your resources by list. Never load everything in the directory. You can at most scan the directory for files, when you pick up new mods.
1563cf No.14720572
>>14720556
Actually, since my stuff is defined by name and index number, I'll pretty much have to. This is just to get shit loaded in and working, I'm trying to crunch for demo day
17aaf5 No.14720577
You guys want a sticky for the demo day?
1563cf No.14720594
>>14720577
http://poal.me/a63qxw
Not sure if it needs a sticky or not. I think we had one the last 2 times, though. Here's the turnout so far. I'm predicting anywhere between 4-10 demos. Still a few days to go.
2004c2 No.14720601
>>14720577
Really depends on the amount of participants I'd say.
7b2e39 No.14720611
>>14720594
I-I'll have a demo in a few years, I promise.
Y-you'll see!
2/2/2020, I-I will have something!
17aaf5 No.14720749
>>14720601
Well you know where you contact me if you guys want a sticky.
2004c2 No.14720781
>>14720749
The cheese cake factory?
d24cc2 No.14720799
>>14720375
what this guy said
just pick the best engine and tools for the job and get on with making the damn game, don't stick your head up your arse and think you have to write EVERY SINGLE LINE of your games code yourself
8c0cb1 No.14720803
>>14720749
>>14720594
Traditionally, /agdg/ gets a sticky for a Demo Day specific topic, separate from the typical /agdg/ topic. It'll probably get flagged for sticky suggestion when the day comes.
http://8agdg.wikidot.com/demo-days
7b2e39 No.14720815
>>14720375
Is "there are three main engines that exist, one of them requires me to fund communists, one is bloated to hell and one is at least two years from being viable and asks me to do most the work anyway but it's accepted by everyone because it's FOSS" as good reason to make your own engine?
0b20e3 No.14720827
>>14720815
If you think you know enough to make an engine then you can strip Unity of all it's "bloat", even if that's a huge meme.
Tetradev was doing it back in the day.
330623 No.14721004
>>14720815
I agree with what >>14720827 says.
Although it's hard to tell who's the communists and who's the bloated one between Unity and UE4.
I think that bloat is in your head, UE4 is technically overwhelming but is more than capable of doing what you want, Unity is underwhelming at times but also can do whatever you want.
If your bloat actually impacts performance is when you need to figure out what goes on under the hood, but you couldn't build better internals unless you know what's choking up the top of the line engines out there.
To think about it another way, it's not about how much time it takes you to start – it's about having everything you need to finish. A nodev is probably very good at starting but when you're near the finish line you are very much going to want all the final challenges of crossing it to be problems someone else has had before.
b397ec No.14721013
>>14720815
>cannot create opengl 3.3 rendering context
>cannot create opengl 3.3 rendering context
>cannot create opengl 3.3 rendering context
>cannot create opengl 3.3 rendering context
>cannot create opengl 3.3 rendering context
Dammit, I'm on a thinkpad t430 and I can't get Godot 3.0.2 to load at all.
9af522 No.14721059
>>14721013
Your graphics card probably doesn't support past 3.3
0b20e3 No.14721070
>>14721013
>thinkpad t430
Who's got the meme laptop now HUH?
1620e1 No.14721128
>>14721013
The T430 ships with intel HD 4000 or a dedicated nvidia card, and both of those are compatible with opengl3. Mind posting your specs?
1563cf No.14721136
>>14721128
I actually bought a T420 specifically for gamedev, because a good friend said it was the last solid model before they were acquired by Lenovo and turned to shit, apparently.
1620e1 No.14721202
>>14721136
Yeah, the T430 introduces a few changes that makes the separation between old and new thinkpads. The biggest changes are the chiclet keyboard and the stock display panel (1366x768, fucking gross). Subsequent models do away with even more, like taking away certain special keys, LEDs, use weaker chassis materials, and even do away with the metal roll cage that gives thinkpads their heft and durability. I wouldn't say newer thinkpads are bad, but they definitely cater to a different audience and alienate those who want a business laptop that ibm was known for.
40e122 No.14721358
>>14720594
My 'maybe' can be move to a yes, so we have at least four demos.
732a82 No.14721432
>>14720815
If you think you'll enjoy making an engine, then that's the only reason you need, as long as you fully understand that you are just adding a year or two to your development cycle depending on how complex the engine is. (Maybe less if its some simple shit) A lot of people get burned out on enginedev in the same way that they get burned out on gamedev, and you definitely need a lot more time to get through enginedev. But it's not like you cant do it if you try for a long enough time.
>>14720827
Most of what Tetradev did was heavily integrating Unity into his own pipeline that uses q3radiant for map editing. It imports all the entities and everything, it was pretty cool.
0b20e3 No.14721447
>>14721432
>Most of what Tetradev did was heavily integrating Unity into his own pipeline that uses q3radiant for map editing. It imports all the entities and everything, it was pretty cool.
That's a pretty strange choice honestly.
I wonder why q3radiant in particular
fdd50f No.14721462
>>14721447
I'm guessing he got really good and comfortable using that tool and didn't want to move to something different.
1563cf No.14721513
>>14721462
I know that feel.
I had a co-op during college back in 2009, 2010 or so, helped out a local studio using XNA, didn't know a thing, didn't learn a thing. Basically a bunch of guys in his mom's basement with one or two prior small iPhone games, and I didn't contribute very much (volunteer position so w/e).
Anyways, they used XNA and I stuck with it years later. Only in the last 6 months have I moved instead to SFML, but I still try to use similar programming patterns with it.
732a82 No.14721538
>>14721447
He's very good at using it and prefers it over all the other map editors. The guy is a serious machine at mapping.
He also has that terrain system that he could paint onto his environments, as well but I lost my screenshots of that.
8f60cf No.14721631
File: f7270b93134ca5e⋯.jpg (Spoiler Image, 84.23 KB, 1100x390, 110:39, sfg BO is rotten redditor ….JPG)
77e7e0 No.14721642
>>14721631
>visit /v/
>people here know about rotten redditer too
goddamn that subhuman piece of shit gets around
0b20e3 No.14721644
>>14721642
>>14721631
This has literally nothing to do with agdg
ec76a9 No.14721699
>>14721631
Why do you even care about this faggot? Who keeps digging up more of his shit? For what purpose? What board are you guys hanging out on?
a18aa0 No.14721700
>>14721644
Except the part where he derailed numerous agdg threads over the past few months.
fdd50f No.14721706
>>14721700
Well let's hope he doesn't derail the thread even when he's not here.
a18aa0 No.14721774
>>14721706
You're right. So instead I'll talk about how Pixiv is the perfect resource for anime stalkan game. Unfortunately my main machine is down and I'm stuck on a craptop that randomly restarts because it think it has 0% battery despite being plugged in so the battery doesn't fucking matter.
I'm basically fucked for any actual development until I can build myself a new PC (Or god forbid finance a new laptop). Right now all I can do development wise is what I can do with my sketchbook and notebook. Once I get a functional computer again I'm gonna digitize the sketches and color them for modeling purposes. Then figure out how to translate my gameplay concepts from words in my notebook to functioning code.
1563cf No.14721779
>>14721774
Can't you get a decent workbook laptop for under $300 ?
ec76a9 No.14721787
>>14721774
Pixiv is an excellent resource for all sorts of anime stuff in general.
0b20e3 No.14721794
>>14721774
>craptop that randomly restarts because it think it has 0% battery despite being plugged in
I know why that's happening.
That's usually one of the shit Dell pulls with their laptops. Even if the charger has the right adapter and the right power, unless it's an official Dell charger then the laptop won't charge up.
Go buy yoself a new charger.
a18aa0 No.14721842
>>14721779
How's this? The only part that worries me is using an Intel HD 4000 for developing a 3D game. I feel like I at least need a dedicated GPU.
>>14721794
Actually it's an Asus I replaced the battery thinking it was that and it worked for all of two days normally then gave me the same old "plugged-in, not charging" bs. I currently don't have the money to buy a new charger though. I get my paycheck in two days so maybe I'll try that then.
8949f9 No.14721857
>>14721842
>shit doesn't work the way it should
>buy more shit
>doesn't work
>buy more shit to fix it when you get paid
you're the best of goys
0b20e3 No.14721874
>>14721842
And there's people that will tell you to buy this shit instead of apple or a chinese manufacturer's
Fuck dealing with modern western tech companies
a18aa0 No.14721886
>>14721857
>Repair it myself
>Don't pay some tech school dropout too much money to fix it for me
>It's a long discontinued model so the parts/accessories are dirt cheap and don't go to the chink manufacturers
>The best of goys
Sure, buddy. If a new charger didn't work I'd give up on it and continue to use it as the underpowered desktop it essentially is.
>>14721874
HP is probably top of my shit list for computer companies that can fuck off. But I'm really tired of sitting idle and not being able to do actual work on my game. I'll continue exploring my options as I save money but I'll honestly probably just be stuck with this craptop for like 6 months while I save up money for a new machine.
40e122 No.14721888
>>14721699
>What board are you guys hanging out on?
/strek/ is my second most-visited board after /v/
>>14721842
180 GB SSD will kill you, especially since the OS will take up a bare minimum of 20 GB, and that's if you scrape out all the shit and continuously monitor it for bloat. My (un-optimized) Windows 7 Pro installation takes 27 GB, I'd imagine Windows 10 is much worse. Why bother with a laptop at all? Use www.logicalincrements.com and pcpartpicker to find the cheapest parts available that will fit with the parts in your computer that still work.
0b20e3 No.14721904
>>14721888
This is a laptop to dev on the go. You're not going to have games or big shit saved to it. Maybe 256 would be slightly better, but 180 is just fine.
>pcpartpicker
Because working from a laptop legitimately makes you more productive. There's days where I would rather kill myself rather than sitting at my desk and work.
Changing scenery or working with friends is great for motivation and keeping productive.
8949f9 No.14721906
>>14721886
>shit doesn't work the way it should
>buy more shit
>doesn't work
>buy more shit to fix it when you get paid
>h-haha at least I'm not paying someone to do this
ec76a9 No.14721913
>>14721888
How do I get into Star Trek? It's too corny for me I just can't get hooked.
a18aa0 No.14721920
>>14721888
The problem with my desktop is that all the parts are old (and well used) minus the GPU and PSU so even though I'm pretty sure the problem is that the HDD died. It's probably not gonna be long before the RAM, or CPU, or god forbid the mobo, dies on me too.
>>14721904
I have an external 160gb HDD so storage isn't too big an issue if the laptop's drive is only 180.
>>14721906
>Being such a poorfag that someone spending less than $20 for replacement parts offends you
You complain about street shitters the whole thread, yet you can't stop yourself from shitting up this thread. Really activates my almonds.
ec76a9 No.14721922
>>14721918
i dont understand
8949f9 No.14721923
>>14721920
You're not going to get a battery and replacement charger for $20. If you have to lie to support your point, you might be a kike.
ec76a9 No.14721926
>>14721924
what are you trying to get me to understand, demon?
a13406 No.14721929
0b20e3 No.14721931
I think I'm starting to get a hang of how to level design.
I tried for a week to make a big open space interesting trying to fill it with shit to do without being able to get anything interesting out of it.
However as soon as I switched to a set linear path it became way easier. Now I'm just going back and adding alternate routes and shit to explore to the linear path and it's actually starting to work out, believe it or not.
8949f9 No.14721932
9526f9 No.14721934
>>14721913
Watch deep space nine.
40e122 No.14721938
>>14721913
>too corny
It requires suspension of disbelief. TNG and DS9 have the best characters, stick with it for a few episodes and you'll know if you'll like the rest of the show. Shamefully, I haven't watch TOS.
>>14721913
wew
ec76a9 No.14721939
a18aa0 No.14721942
>>14721923
$19.05, no tax and free shipping on each. Read it and weep, bitch.
9526f9 No.14721944
>>14721938
Me neither. TOS looks so jarring.
0b20e3 No.14722809
Might as well post progress with the level
085fa2 No.14723192
>make hair
>test a few different methods
>all look like shit
>forgot to apply shading like I'm supposed to
>apply
>woah, nice
pretty happy with this hair.
I'm going for an anime look… it looks pretty anime to me, but what u anons think?
Note that I'll definitely do less symmetry in the future, and I still need to add outlines to the blender shader.
0b20e3 No.14723299
What the fuck is Clustering? And why the fuck does Unity take 2 hours on a scene made mostly of really simple shapes to finish that process?
I can't fucking imagine how long that's going to take once I have actual models with complex textures in there holy christ
>>14723192
I think it looks great, but I don't think it works for an "anime style" game honestly. It's got way too many single strands of hair.
085fa2 No.14723334
>>14723299
>What the fuck is Clustering?
pretty sure it's a step in baking the lightmaps using clustered rendering
also what version you on, I've heard there's one that's bugged for lightmap baking so it can take a huge amt of time (2017.1 or something).
>It's got way too many single strands of hair.
humm, that makes sense, anime tends to oversimplify that aspect most of the time.
I'll try to add a "hair helmet" for the top to make it more anime conforming, and clumping more together so they look less individualized
0b20e3 No.14723350
>>14723334
>what version you on
I'm on 2018.2, didn't seem to be among the known issues though
a18aa0 No.14723359
>>14723299
Unless Takeshi Obata is drawing. Guy loves his hair strands.
40e122 No.14723366
>>14723299
>When generating the Scene’s lighting solution using Precomputed Realtime GI, Unity simplifies the calculations required by working on a voxelized ‘proxy’ of the Static Scene. These voxels are called Clusters. Clusters are effectively surface patches (small tiles) mapped onto Static geometry which we use for lighting. Clusters are stored in a hierarchical relationships and are used for the complex irradiance calculations needed when precomputing Unity’s diffuse global illumination solution.
https://unity3d.com/learn/tutorials/topics/graphics/understanding-clusters?playlist=17102
That page is part of a tutorial to help you optimize your light baking times, see the contents at the bottom of the page.
085fa2 No.14723387
>>14723359
true, there's so many variants for anime-style hair it's hard to choose.
suppose I'll experiment a bit more and go with what I like the most.
0b20e3 No.14723407
>>14723387
>>14723359
I guess it depends with your general art style.
>>14723366
This for some reason didn't pop up when I searched, thanks anon.
b42679 No.14723503
I'm drawing a static vertex array in SFML and performance is ass. How can this be improved?
1563cf No.14723512
>>14723503
Post code. It should be very fast.
Are you changing textures often?
b42679 No.14723538
>>14723512
It's extremely simple. Literally:
sf::VertexArray v;
v.setPrimitiveType(sf::Points);
// randomly add white points to the array
v[random number].color = sf::Color::white;
// later
window.draw(v);
The draw call is the only thing in the main loop. Nothing gets changed it just draws the array. Performance spikes up to 16% on my i7.
1563cf No.14723589
>>14723538
>>14723538
Thats weird and I cant think of why that is. If it were being modified then I'd check things like if you were clearing it or texture swapping it. I did a similar thing to you to test out random functions and i dont think it was that bad.
Is it just as slow as when you draw other objects? Are you running updates with an fps cap?
b42679 No.14723605
>>14723589
It's just that one call to draw the array. FPS is capped via window.setFramerateLimit(60);
085fa2 No.14723748
Tweaked it a bit via editing normals, and adding a solidify mod to make the hair thicker.
I think it's a bit more consistent now, and less jarring for all those individual strands.
I like it a bit more than the old one, thanks for the suggestion anon.
still need to re-fix the hairband, don't mind it; too lazy to retake screenies
085fa2 No.14723802
>>14723748
fixed hairband, buffed up the ponytail a bit
1563cf No.14723817
>Supposed to work 11-6 today
>Woke up at 7, called in, wake up at 1 pm
>Just started devving and need to do errands
>Only 1 day off before demo day
Christ
8c0cb1 No.14723910
>>14722809
>>14721931
That's a pretty decent starting point, yeah. Start with a critical path, add branches. I'd suggest adding loops and shortcuts, both two- and one-way. I'll probably have to look at more level design stuff some when I get to that point in my game.
c15ef0 No.14724605
Finally decided on the ramp design to reach the central platform
c004ea No.14724614
Is maya any easier than Blender
ec76a9 No.14724657
253087 No.14724668
>>14724614
2.8 has some UI improvements in the pipeline, although you'll need a graphics card supporting OpenGL 3.3 (basically DX10 level) to run it.
55c924 No.14724673
>>14724614
Long Answer:
It has a bunch of retarded design decisions, some better features, and crashes 99% more often. Maya is still considered the industry standard but is incredibly bloated and broken software that requires a subscription fee to use.
> not pirating your software
Short Answer:
No
9526f9 No.14724698
>>14724673
Unnecessary software subscriptions piss me off. Worst is I applauded Adobe when they started it.
c004ea No.14724702
okay i'll stick to blender
but how do i learn topology, there's no real guide to it, all the guides i've seen just guess it out and say "oh the topology has to be good btw"
9526f9 No.14724723
>>14724702
There's tons good of topology images/tutorials/guides out there. Search properly.
a18aa0 No.14724745
>>14724614
Not unless you want to make use of some very specific features of the program. Like how it can be synced up to UE4's animator now.
>>14724698
I almost miss the ridiculous thousand dollar price tags they put on their software
085fa2 No.14724759
>>14724702
Look on the wiki, on the game dev resources page.
Lots of topology related resources in the modeling section to learn from.
1563cf No.14724761
>>14724745
Who would win, subscription-based propriety software or this amethyst geode?
c004ea No.14724786
a18aa0 No.14724821
>>14723387
Don't know if it'll help any but I've been using this as my sort of key for what I want hair to look like once I get a good machine and get back to work. Thankfully the autism of Zelda's fanbase means there's already some documentation on BoTW's rendering techniques
9e3e8e No.14724858
>>14724761
amethyst a shit
d5bfb1 No.14724951
HookTube embed. Click on thumbnail to play.
WHICH ONE OF YOU MADE THIS?!?!?
ec76a9 No.14724975
>>14724951
is this rottenredditor?
40e122 No.14724978
>>14724951
>other videos are mostly babby's first existential crisis-tier blogs
>the few videos of "making games" he's using Construct 3
c15ef0 No.14725046
Tree plats, need bridges now
Also more trees
c004ea No.14725079
>>14724978
whats his youtube channel
e306c7 No.14725214
>>14724858
Doesn't really count as that's an artificial bismuth crystal compared to the natural amethyst geode
30ee27 No.14725306
leaves and buttons for the """"demo""""
53f78c No.14725499
How should I manage my Vector class?
Should I have two standard VectorF and VectorD classes who's size is set on creation or should I make Vector2F, etc for the sizes I need?
663a98 No.14725801
>>14724858
Damn anon did you design this level yourself? Looks pretty magical. The castle with a winding path up to the central tower and watchtower/ practical facilities going down along the edges right? Seems really defensible and practical to be honest. Sieges would have to go up the winding, winding path whilst they get shat on from above. Corner positions can be defended by that same winding path, however, if the enemy keeps their distance and aims for the structures. Plenty of positions to fire bolts form. I think if you lay down sand you can catch boulders they launch to keep them from rolling down the mountain side.
085fa2 No.14725827
>>14724821
Thanks anon
My actual in-game shader has been evolving into a more botw-style lately; so this will definitely help.
I especially like the specular on the hair. It looks like it’s a non-standard approach to specular.
>>14724951
>blamingTheTools.mp4
>babby’s first existential crisis
wew
1563cf No.14725933
>>14725499
SFML (2D based) has a few variant classes based on on what it uses: Vector2f, Vector3f, Vector2u, Vector2i. They support basic operations, like addition and comparison
XNA (3D based) only has Vector2, Vector3, Vector4 in floating point only, but has additional vector functions built into it, such as cross, dot, reflect, etc.
1563cf No.14725961
Also got through a few more things
>Behavior class exists again, with event hooks (create, destroy, idle, move, bump, trigger, collect), hooks still need to be called, though
>Animation works again
>Music and sounds play at start at 2s intervals, need to make it so I have an options menu that decides whether to pause and/or mute the game when it loses focus
085fa2 No.14725978
>>14724821
On second look they definitely have a standardized approach. make a shitty pic to show what i mean
a18aa0 No.14726166
>>14725978
Yes, all the hair models have geometry that cover the head and then they add to it like a niggress putting extensions into her hair.
934df2 No.14726350
>>14726166
>>14725978
I've always wondered about the possibility of a hair-specific sculpting system.
>build a custom square/rectangle/hex/triangle/n-gon grid over an imported head model, with each grid space being a clump of hair
>clumps can be grown, shrunk, cut to various angles, split, teased, braided, etc. using intuitive tools
>advanced panel lets you add ornaments, change the weight of each clump, apply custom shaders and so on
>export generates a "wig" model matching the head, or a series of vector images like Live2D
>Hairworks/TressFX compatible
>library can be accessed by vidya for improved character creation
085fa2 No.14726455
>>14726350
They have something similar for maya, but specific to a hair cards approach. For quality wise, it was used for the latest witcher.
Pirated maya just to try it, and I didn't like it at all; though it's better than doing hair cards yourself (though i hate working with hair cards in any manner).
Your approach sounds more intuitive, but specialized for for more anime/stylistic hair (non-hair cards); which is something I'd definitely use.
934df2 No.14726530
>>14726455
>never heard of hair cards
>look them up
Oh fuck, that's awful.
fdd50f No.14726637
>>14726626
The fuck kind of hair cards are those?
Hair cards usually look more like pic related.
158d32 No.14726666
>>14726626
That image looks like an example of physics controlled hair, with only the physics guides shown. Hair cards look nothing like that. see >>14726637
085fa2 No.14726680
>>14726530
Image the shit those crazy japs who made the hair for the final fantasy games w/hair cards went through… all those crazy hairstyles, and precision required.
I can imagine how it could be fun, while being easy for short hair, but it can get frustrating really fast for long hair even when using nice tools like maya's particle system, or say bez curves (technique used by FF character modelers).
>>14726637
That's the maya particle hair tool, which is using a proxy (generated) hair card to render their supplied hair alphas. It outputs hair cards like those you posted, and are definitely easier to work with (compared to by hand) due to the particle/curve style tools; though it's still a pain.
>>14726667
kek'd
>>14726666
yes satan those are physics flyaway hairs, but see above
0b20e3 No.14726728
>literally spend 12 hours baking lights in my scene
>it now looks even worse than before
>I can't even go back to how it looked before
1563cf No.14726740
>>14726728
>Didn't make a copy
>Can't manage resources with VCS's
What did you learn
9526f9 No.14726747
>>14726728
That is flat shaded. There's not even shadows. What did it do for 12 hours?
0b20e3 No.14726758
>>14726740
It's all good, I think something broke in the last parts of baking and somehow didn't finish. Luckily all the intensive parts were done already so I just had to run it again
9526f9 No.14726818
>>14726758
It still looks flat shaded…
9526f9 No.14726841
Is there any overlap between agdg8 and agdg/2? Do you share any wiki, any disagreement? I tried searching for agdg online and I just found the alternate ones. They even use the same logo.
0b20e3 No.14726850
>>14726818
Maybe it's a problem with Unity?
Not sure what to do tbh
9526f9 No.14726892
>>14726850
I mean, what's 12 hours of lightmap computation for if it's not to make some sweet global illumination model? I just see some basic lighting. Maybe run through some tutorial, I think you might have missed a checkbox.
0b20e3 No.14726899
>>14726892
I know these settings work cause I used them in another scene.
At the moment I straight up don't have the time to redo the whole process so I might as well keep it as it is.
330623 No.14726918
>>14726850
sometimes bakes take really long times, a good way to test your bakes more frequently but not wait forever is to nerf the resolutions for lightmaps globally as well as on each meshrenderer. and mark your statics.
1563cf No.14726924
>>14726841
>Wanting to use Dxscord
0b20e3 No.14726930
>>14726918
Yeah that really does not seem to change time for me, not sure why
I think I just gotta spend more time learning what these settings actually do
8f60cf No.14726933
>>14721699
rotten_redditor is the BO of /sfg/, an offshoot/ripoff/hijacking of /guro/
9526f9 No.14726980
>>14726924
I just want to know if there's any trade with that tribe.
a18aa0 No.14727087
>>14726530
Just imagine the tedium of texturing haircards to make the hair look the way it does. Now you finally understand why art assets take forever to produce to the point that FFVII Remake can't have all the same content as the original. The logistics of modern art asset pipelines simply doesn't allow it.
9526f9 No.14727151
>>14727087
Ideally you don't do the texturing manually. You use a brush tool to paint the hair polygons which are already uv mapped to a single hairpatch texture. Then you use some styling brushes to trim and comb the polys. Only big, character-style details would be manual.
that being said, it's true too much time and money is put into AAA art assets. If only they spent a fraction of that into fun, innovative game design.
1563cf No.14727189
So I had a thought about how my game's meta progression would work.
In pic related, you access levels by completing adjacent levels and you can play them in a nonlinear manner. If you can complete three objectives, the level is considered "mastered" and gets a lightbulb icon on it, rather than a gem icon. Each level can be played on easy or hard mode, and you want to beat the hard mode's gem quota, time limit, and use the hard exit in one go.
So I wanted to do something pretty much identical to that, but for each objective you complete, you get a gold card. You don't have to get them all in one attempt, you could end up with 3 gold cards. But if you manage to do the "master" thing, you'll get a bonus of 2 extra cards, for a total of 5 gold cards per level. Then, I was going to put different unlockables or even level groups behind card doors (eg need 70 to enter)
Thoughts?
158d32 No.14727276
>>14726841
> do you guys associate with cuckchanners?
085fa2 No.14727310
>>14726841
>>14726980
>I just want to know if there's any trade with that tribe.
definitely isn't any, and friendly reminder to lurk more
9526f9 No.14727325
>>14727276
>>14727310
Fair enough, but in my defense agdg is a different kind of place than the rest of /v/.
085fa2 No.14727350
>>14727151
>You use a brush tool to paint the hair polygons
Only really works well for initial placement, but this is where tools like particle hair really come in handy (esp. if it has auto hair-card generation, for non-overlapping hair cards).
>already uv mapped to a single hairpatch texture
When i used hair cards i generally had multiple variants of said overlapping uv maps w/hair alphas prefabbed to the side that i'd just duplicate
>Then you use some styling brushes to trim and comb the polys
only works well for short hair/low-pixel density hair alphas.
If the hair-card polys have major overlaps, and the pixel density is high on the hair cards; then there will be really bad alpha blending aliasing, and other such issues that require bandaids upon bandaids to sorta-fix (why not prevent that in the first place, layer them perfectly).
The ff character artists do all their stuff by hand, strand by strand, because of the flaws of the hair cards approach; to not (visibly) overlap any haircards as to not cause said quick-n-dirty result (from high pixel density hair cards overlapping)… a glitchy overdrawn/alpha bleeding mess.
Though, for stuff like really patchy/thin hair (low-pixel density that the shader has to alpha blend) it's less of an issue (short hair is easy with hair cards, layer and overlap is ok, because pixels we're working with to alpha blend are way fewer), but when you start approaching the level of multiple layers, and long hair… oh boy will the alpha blending technique bite you in the ass if you don't lay your hair cards perfectly.
>>14727325
you're good anon it's just some friendly bantz
0b20e3 No.14727401
Advise me on this /agdg/.
I'm wanting to remove all references to Unity's internal physics system and move my character exclusively with my own math since I now know enough to make it work.
Should I implement my own 3D collision? Is that a complete waste of time?
ec76a9 No.14727443
>>14727401
I don't know if Unity's physics are deterministic, but if your game relies on physics it's always better to write a system yourself - one that you understand well and can adjust easily.
I know I'll have to do something to get UE4 to simulate physics in a deterministic manner - not looking forward to it tbh, it's either gonna be a whole lot of fucking around with PhysX or cutting it out and adding my own, which makes me wanna kill myself when I think about it.
085fa2 No.14727456
>>14727189
I love thinking about the meta-aspects of games, but I really don't have enough context for your game to fully cognate what you're discussing (at a mutual level of understanding).
So, I'll just go over what is one of my favorite multiple difficulty systems for a "campaign"; halo (CE).
Even if you played it, it can be helpful to get a reminder of what it included, and how it enchanted someone of maybe another playstyle.
There were 4 difficulties: easy, normal, elite, and legendary.
Each level can have "difficulty emblems" displayed, per level, if you beat that level at those difficulties. So, if u beat it at normal and elite; then the normal AND elite emblems displayed at the level selection screen (no overwriting/erasing w/higher beaten difficulty, u did them all, they're all displayed).
However, if u beat it on easy, u have earned no emblem for the level; i.e. only displays normal/elite/legendary, giving a sense of, "that isn't how the game is meant to be played, but go on ahead, plebs".
All levels can have, up to N/E/L "difficulty emblem" displayed at once if u beat that lvl, at those difficulties, per lvl.
It also unlocked a few neat aspects, like a special ending, hidden triggerables, and such fabled things.
It was a pretty neat system, but each difficulty had significant challenges added (better AI, fewer ammo/guns/etc, more enemies, lower shields, lower ally health, etc); making it actually really fun to beat all the levels on all difficulties (especially with a friend).
As, even beating your most disliked level was awesome at all those difficulties, as those were metals to your achievement; which was really fricken neat (though just about all the halo:CE levels had amazing level design, and engagements).
This was before the achieve systems that are so popular nowadays, and having a really personal memento to your accomplishments in the form of difficulty emblems per lvl was enchanting.
It also made for some really fun weekends with my buddies, playing halo:CE, beating lvls on multiple difficulties in coop, and then switching to morrowind for some hotseat chill adventure time.
i'm gushing nostalgia while rocking some rose tinted glasses atm if u couldn't tell
>>14727401
For 3D physics unity uses nvidia's physx.
look at the physx github, and tell me you want to replicate that mess.
It's honestly not worth it, unless you need to implement something that the current physx sdk cannot do, and it will not support; as otherwise it's burgeoning into engine dev territory… and you should learn more if you want2 (though, you don't have2 re-implement physics to do so, look at their source code, experiment more, read more specialized physics engine documentation, port an open source physics engine).
You can also check out the ported bullet physics engine. As mentioned it is open source, so you can get your hands dirty there, and learn more/experiment at the low-level if you're so inclined.
0b20e3 No.14727461
>>14727443
Unity's Physics system is deterministic, but still, there's some weird shit that sometimes happens that I'd rather remove.
I already have all the movement figured out, I've got all the vectors I need for every single situation, so it's essentially a question of applying my own gravity to my character and changing the way I apply those movement Vectors.
Shouldn't be that hard, right?
0b20e3 No.14727466
>>14727456
I probably explained myself wrong. I don't need physics, like, at all. At the moment I'm just moving my character with physics because when I started off it was easier.
All I really need is gravity, and that's about it.
ec76a9 No.14727473
>>14727466
If that is the case then you're probably better off going kinematic
0b20e3 No.14727483
>>14727473
Oh shit I didn't know I could do that. I thought I had to remove the rigidbody completely.
085fa2 No.14727492
>>14727466
>Unity's Physics system is deterministic
For actual physics simulations physx isn't deterministic sadly, and is built for speed v. accuracy.
>there's some weird shit that sometimes happens that I'd rather remove.
this happens because of the above.
they offer a bunch of options to tweak, you just need to understand why the weird things happen, and find their related tweakable setting.
>>14727466
>All I really need is gravity
oh, well, definitely just apply your own gravity.
I've found it's best to not use an always active non-kinematic rb for player controllers imo (only maybe disable kinematic for jumping).
So, use a kinematic rb, definitely; apply your own gravity and such.
Note that the physics engine also includes collisions, as u mentioned you want to redo it, and that's the main part i was referencing as being a bit of a "mess"; as that's some high-level math involving building of some complex high level solvers to get near the level of physx's offered collision detection.
Thus, if u want different collision detection, I'd recommend bullet or some other open source physics engine.
>>14727443
>deterministic
I know newton is deterministic, and bullet can be made deterministic (https://pybullet.org/Bullet/phpBB3/viewtopic.php?t=3143).
They've probably got ports/bindings for at least bullet to ue bcs it's so popular.
0b20e3 No.14727501
>>14727492
Isn't collision still going to work if I keep the rigidbody but set it as kinematic?
085fa2 No.14727504
>>14727501
yup, it'll work exactly the same
918fd0 No.14727569
I hope everyone is working hard on your demos for the big day. I see so many projects in these threads that I'm excited to play and I know other people are as well. Good luck my friends.
330623 No.14727578
Anybody work with scene streaming in Unity? All the games I've worked on so far either have one master scene always loaded, or we just load the entire scene at once.
In an RPG style game this would make less sense, as you'd have a scene for a village or some shit, and then potentially scenes to determine which minor versions of the village would load.
So you'd potentially load up to 3 new scenes per screen transition. How do you keep those managed? How do you have cross-scene references (doors / load points) without going mad trying to manage them all? Where would the logic for loading a scene variation go?
You can't even slot scenes in components easily so having a global lookup object somewhere to map scene to scene transitions is some skullduggery.
All of this on top of not enforcing that every loaded scene is strictly a "map" and not potentially a battle screen or something. I've never read anything that allows for a system that is easily usable and verifiable for designers that don't know as much about programming.
0b20e3 No.14727597
>>14727578
I'm not sure what you're asking here. You want to load all the transitions scenes as soon as you load the map that connects to them?
Surely if you're using Unity you can just use Triggers with a script attached to them, and when the player enters one they get transported to whatever other scene you specify in the Inspector?
8a43ba No.14727599
Worked on internal logic for easily putting together random clothing assets that might have cloth physics or just additive meshes. I'm starting to like C# a lot more than Java.
Something interesting for VR game design news: I went to a convention today and Facebook said that 83% of time spent inside Gear VR/Oculus Go/phone VR/etc. was spent not in games but rather in media viewing. So stuff like watching movies in Hulu or Netflix. Also that only 40% of those normalfags have ever opened a VR game app on their VR device. They prefer media consumption instead. This means that VR is still not profitable as a gaming platform and might take many years to reach media consumption.
>>14725306
Protip: That's not how you profile shaders. There are 2 things that you are doing wrong: consistency for control group and misusing shaders.
When you profile for performance, you need to make sure you maintain the exact same benchmark. Looking around randomly at your tree shader is not maintaining your benchmark because you change on screen poly count which changes rasterization which changes fragment count. Not to mention you are changing visibility culling and making Unity calculate a new field of vision which makes it harder to isolate your shader's performance. You need to first find 1 static spot in your map that really stresses your shader. If you are using forward rendering: If you look at your trees from below, none of the leaf triangles overlap that much because your view shoots out to the sky. But if you stand at tree leaf height and look forward towards all your trees from a corner, your view will have several overlapping trees (lots of triangles rendering on top of each other). Find a good spot for this and just save that position for when you compare benchmarks. If you are using a deferred renderer: you would instead find a point that contains as much leaf as possible, so in that case zoom into one leaf until your entire screen is leaf.
The misuse of shaders is that for good performance shaders, you should NEVER notice variable framerate by adjusting uniforms for the same scene. If so then that means you must be using branching in the fragment shader. For your leaf shader you can again try using step() function. You don't need any if statements and you should strive to see absolute constants when benchmarking. Also look into making model LODs if you are looking to target a certain framerate.
8a43ba No.14727604
>>14727461
I'm gonna go out on a limb here and say that you don't really know what you are getting yourself in to. What makes you want to replace Unity's physics? 3D math with Vector3 classes is the equivalent of hammering nails down your pee hole.
330623 No.14727615
>>14727597
A little more complex than that. All the big N64 games had a thing where they'd have the map geometry as a scene, then would load different layouts for stuff like a plot boulder getting knocked over or having an "on fire" version of a map.
What I need to do when a scene transitions is to load the base scene, in addition to other scenes that are appropriate based on the player's progress. (e.g. the "on fire" layout for that scene)
But that logic can't live on every single entrance and exit to a particular scene, and managing string references to map scenes would break way too easily.
0b20e3 No.14727632
>>14727604
>What makes you want to replace Unity's physics?
Unity's physics are imprecise and in some situations they act in ways they shouldn't.
I've already done all my 3D math with Vector3s so I don't know why I wouldn't be able to just get my character to fall, move and jump without Unity's gravity system.
>>14727615
Ah I get it.
That's more about logic honestly.
I haven't done it in unity but I did have a 2D hardcoded RPG where I was working on this.
If I remember right, my player contained data for different flags. These were added to the player in case of specific dialogue/quest completed/special event.
When loading a map (which were just JSON data), the script would check if there were more than one version of that map.
In case the script found more than one it would check the requirements for each version to be loaded (which were essentially flags that I could cross reference with the player's flags), then it would go through whatever flags the player had trying to find the flags that were needed to load that map.
If all the requirements were met it would load the specific map.
Since you're in an engine you could do this object by object.
So on loading a map you would check where each Gameobject in the scene would be and which model would be associated with it. Depending on the player's flags you'd then instantiate the GameObject at XYZ position with W model.
Not sure if that makes sense as I explained it.
Essentially if you don't want to have 30 scenes that are just copies of each map you might want to instantiate objects when the scene is loading.
8a43ba No.14727654
>>14727632
Right but I mean specifically what do you want to replace? Is your character a single point or a capsule? Because trying to hamfist your own physics engine on top of another physics engine sounds like the silliest thing ever.
0b20e3 No.14727660
>>14727654
What do you mean what do I want to replace?
I want to have more control over gravity without having to deal with things bouncing without wanting it, or slipping, or weird velocity changes happening.
I'm not writing a physics engine, I'm just applying forces to make him fall and maybe very few other situational occasions.
I mean shit just being able to run my game out of Update without touching FixedUpdate would make a whole lot of stuff way easier and simple.
8a43ba No.14727672
>>14727660
>run my game out of Update without touching FixedUpdate
But the reason that is implemented is so you get consistent physics not tied to your framerate.
0b20e3 No.14727680
>>14727672
Yeah, but in the meantime it creates a bunch of other problems.
If the worst that comes out of this is having to lock my game at a fixed framerate then I honestly really don't care that much.
330623 No.14727688
>>14727680
doing your logic in FixedUpdate and handling input in Update and if everything physics is absolutely knackered then setting something as kinematic should be all the fucking around you need to opt out of Unity's physics if you don't have time to tune everything to play nicely with them.
0b20e3 No.14727694
>>14727688
That's what I'm saying.
Although the coliisions don't work when something is kinematic so that's probably going to be a problem.
8a43ba No.14727724
>>14727694
Maybe if you posted more technical details as to what you are trying to achieve we can help you come up with a solution. I'm pretty sure Unity's physics is robust enough. What EXACTLY is your character supposed to do that it doesn't?
330623 No.14727730
>>14727694
The conditions for colliding with Kinematic are different but they can still work properly in most situations as long as the other thing also has a rigidbody or if it's static IIRC.
If not, then whatever your kinematic thing is attach a configurable spring joint to a different thing that can't collide with your kinematic object and use some of that. A bad idea that I don't recommend but it could be useful in some circumstances.
0b20e3 No.14727793
>>14727724
There's not a single instance of something that doesn't happen, it's just a series of small things that I feel I have no control over.
I just don't have control over the whole Unity Physics system, which means in some really rare scenarios things seem to happen even though I didn't plan for them to.
The problem is that I'm basically using an absolutely gigantic system despite needing only a fraction of it.
I dunno it just kind of pisses me off.
8a43ba No.14727800
>>14727793
Ah okay. So it's more of an annoyance. Well then I wouldn't worry about it then until you start implementing and seeing measurable scenarios. Doing cleanup before doing anything is just like premature feature optimization.
8949f9 No.14727882
>>14727800
Can't you faggots come up with your own ways to express your thoughts and ideas?adding "feature" between "premature" and "optimization" doesn't count
0b20e3 No.14727895
>>14727882
Speaking in fake tech/design savvy language to look cooler is one of /agdg/'s oldest and most honored traditions
ae0749 No.14728018
>>14727882
I would honestly not be surprised if all of the people here crying about OOP/DOD had a grand total of zero experience in writing either but were just regurgitating what some faggot on jewtube said.
c004ea No.14728243
hey guys, hello!!!
whats your favorite game and inspiration
ec76a9 No.14728324
>>14727895
I'd understand if you were bitching about things like "paradigm shift" but it's honestly just a very common topic.
>>14728243
My inspirations (games I find extremely worthwhile to learn from, it can be intentionally well-crafted games or just accidental gold games) are as follows:
>Atmosphere:
1) American McGee's Alice
2) F.E.A.R.
3) BloodRayne (I'm fully aware of the campiness of it, but sexy vampire girls, fucking awesome monsters and Nazi occult research puts it in a special place in my heart)
4) Hidden: Source (and all other similar campy early 2000's secret files games)
>Gameplay
1) The big ones that I won't list because you know em (Quake, CS, UT, etc.)
2) F.E.A.R.
3) Minecraft because it opened my eyes to the value of procedural generation when done right, complete destructability of environments and sandbox as a primary gameplay element. It became shit in like Beta 1.7, I know.
4) Garry's Mod modes like Trouble in Terrorist Town. I've yet to see another game facilitate PvP mindgames in the same way as these do. It's really incredible.
I was planning on breaking it down into multiple categories beyond that but I'm really tired, anon. Anyway, these are the main ones.
4833d9 No.14728373
>>14728243
Skyrim battlefield dark souls XD best fames ever made I want to make something like it!
d11af5 No.14728646
When exactly will demoday start? I need my game at the top of the thread for maximum attention.
fbad5d No.14728766
>>14728646
The first post will be an anchor post that everyone who posts a demo has to quote, so you'll get equal amounts of attention anyway.
d46180 No.14728769
>>14727599 ☑️
> Something interesting for VR game design news: I went to a convention today and Facebook said that 83% of time spent inside Gear VR/Oculus Go/phone VR/etc. was spent not in games but rather in media viewing. So stuff like watching movies in Hulu or Netflix. Also that only 40% of those normalfags have ever opened a VR game app on their VR device. They prefer media consumption instead. This means that VR is still not profitable as a gaming platform and might take many years to reach media consumption.
That doesn't surprise me, people who are serious about vr gaming are probably going to invest in a dedicated device or wait for prices to drop.
>>14728243
>Thief 1/2 (gameplay, level design, story [especially how it's incorporated into the levels])
>Quake (level design, lighting)
>Rare's n64 3D platformers (music, level design, character design)
>Deus Ex (story, gameplay)
>E.Y.E. Divine Legs (level design, character design, gameplay)
>1080 snowboarding (music and gameplay)
>>14728324
> 2) F.E.A.R.
> 4) Hidden: Source (and all other similar campy early 2000's secret files games)
Excellent taste, anon
1b1a7d No.14729059
>demo day coming up
>4 day weekend
>shityeahtimefordevvan.exe
>bunch of shit comes up
>0 progress
>mfw
fuggg :–DDDDD
f7a57c No.14729148
>>14729059
>7 day weekend
>do loads of shit I intended to for months
none of them is devving, despite my ambitions
RIP
c15ef0 No.14729340
>>14727599
>That's not how you profile shaders
I was not profiling shaders, I was showing off that the performance is good overall, even with all trees releasing particles (originally they were capsule particles, a very bad idea) and using up a good portion of the screen. I would update the nvidia drivers and use the GPU profile/debugger had I wanted to profile the shader
Thanks for the tip anyways, I believe i am using ternary operator instead of step(), so there is one thing to improve on it outside of complete rewrite of the technique
7270ff No.14729716
>>14718613
Looks like the kind of game I'd get into.
>>14716590
> gamedev is the one thing I could never give up
I know that feel. But I'm too much of a brainlet to program and no one will join my projects, so instead I used mods and modding to make my "game". It does quench the thirst, for a while.
ec76a9 No.14729949
>>14729716
I failed every single year of HS and never went to college because I knew I'd fuck it up. If I can program, you can, too. It's not rocket science, fucking pajeets do it.
8949f9 No.14729962
>>14729949
pajeets don't actually "program" like you or I, they're more like an ant colony.
ec76a9 No.14729975
>>14729962
Yet still, they do it. The point is - programming is a vast body of knowledge but once you get into it, everything about it can be easily understood. =
1563cf No.14730029
>>14729059
20 days out from demo day, I had my schedule. I had 9 days off, and scheduled appropriate dev time for 7 days, with a 2 day buffer in case of shit happening (extra shifts, family emergencies, game nights, whatever).
I ended up with 6 dev days, and didn't quite manage to follow my schedule for things, but more or less, I got a lot more done than I normally would have.
53f78c No.14730069
Would having all my collision data stored as arrays of vertices have a noticeable effect on game speed for collision checks? Would an AABB have a noticeably faster performance?
1563cf No.14730102
>>14730069
Your first order of collision checking should be some kind of spatial partitioning. Trees of some sort are ideal for this because you can usually prune magnitudes of objects just by traversing the tree and finding an appropriate node.
From there, you want to then check the remaining potential objects to see if their collision shapes overlap (circle/radius or AABB). You can get more precise from here, if they intersect.
Personally, I think arrays are not well-suited to the task because they don't scale well, but there's probably a lot of neat tricks with them.
7b2e39 No.14730120
>>14728243
I'm making a 3D platformer (though it is in very early stages).
Basically, controls are about 50% Mario 64 and 50% from other games/original moves.
Level design and aesthetics are inspired by the Spyro trilogy, though settings are ones that are generally not seen in platformers, are original or if they are more standard, I try to still do something creative with them.
Music is from all over, really. I mostly just think of a song I've heard that would fit a stage and learn how to write something in that style. Xenoblade and Kirby are the ones I've used so far.
53f78c No.14730134
>>14730102
I'll make sure implement trees later on, right now I'm just trying to shit out something for Demo Day. Thanks anon
1563cf No.14730148
>>14730134
If it's for demo day, then just worry about shitting something out.
Unless you have thousands of verts that are getting checked every frame by every object, it should be fine.
40e122 No.14730718
>>14729716
Get Gamemaker or Construct and just like make game.
085fa2 No.14730753
>>14727660
>What do you mean what do I want to replace?
The reason this is important, is because you specified some arbitrary things you want to replace; so it's better to get an exact definition here.
Also, the physics engine covers a lot of ground here in terms of offered functionality (i.e. a physics engine isn't JUST physics, but all the components going along with it).
So colliders, joints, kinematic rb, raycasts, sphere/shape casts, and so on.
These are all implemented by physx, or physx offers the functionality to allow these to be implemented via unity at the native C++ side.
>I mean shit just being able to run my game out of Update without touching FixedUpdate would make a whole lot of stuff way easier and simple.
The issue seems to be that you're touting some major ignorance in terms of how to use fixed update correctly, and what its purpose is; in addition to examples of how tying your movement/etc to framerate.
I'm not attempting to be an ass here, but it's true from what I can observe from your posts; I've been there too, so let me explain this clearly… in a way I wish it was explained to me years ago.
Fixedupdate is there to run things separate to framerate, at a FIXED rate.
Having your movement tied to framerate results in major inconsistencies, with variable results based on your framerate, and results in shit like fallout4's super fast movement at high framerates, and super slow movement at low framerates; in addition to the issues of a "variable timestep physics engine" that leads to a plethora of potential issues (overly large delta, i.e. very slow framerate = very inconsistent and large deltas of movement = inconsistencies such as passing through colliders = BAD).
This is resolved via fixed updated, which runs it at a fixed rate, THUS, consistency; however, one must multiply their movement deltas by the fixed update's time delta (i.e. runs at 60 ticks per second, thus, fixedDelta = 1/60, and getting actual movement delta applied in fixed update is: movementDelta * fixedDelta = fixedMovementDelta.
Which is the movement scaled to a rate of 1/60, and is important because now your movement is always consistent (I won't mention why this sometimes is false, but in most cases, this is true; in addition to being way easier to use than a variable timestep).
>I'm just applying forces to make him fall and maybe very few other situational occasions.
Same, I use no actual physics from physx for my character controller, but, I still use fixed update for my movement and such because it's better to not tie that to the framerate.
>>14727694
>collisions don't work when something is kinematic
yes, they do.
The requirement here for a collision is a COLLIDER, and having a kinematic rb or even lacking a rb at all isn't necessary; as long as you have a collider.
However, the kinematic rb offers some functionality you may want2 use.
The only thing that changes for having a rb set to kinematic is the lack of forces being applied between rigidbodies.
A kinematic rb, with a collider, can still adjust the movement of a normal rb (due to say gravity, or having a force applied through contact).
The only thing here that changes is the rule of equal and opposite forces, the kinematic rb does not get a force applied back, and it only applies a force to a hit collider w/a non-kinematic rb.
>>14727793
Understand the physics engine, and you will understand what is going wrong. In most cases it's either: user error, or the issue is derived from the speed over accuracy approach of physx (they offer a tweakable setting to resolve these issues).
Such as very fast moving rbs w/colliders, they may seam to "phase through" colliders at extremely high speed, because their movement deltas are large at the fixed rate you defined for the physic's engine's timestep.
This can be resolved by enabling "continuous collision detection", which means, even with large delta it tracks between deltas applied per fixed timestep; so, no more phasing through colliders.
The same can be said for when objects collide at very high speeds, and they seem to "penetrate" eachother; the reasoning here is the penetration depth, and possibly deltas; thus try to enable continuous dynamic detection, interpolation of deltas, and maybe lower penetration depth on the physx side; wallah, fixed because I understand what the problems is, and why/how it can occur.
>>14728018
Same can be said for defenders of OOP, and the point that they haven't touched DoD or the alternative programming design patterns; it's actually a lot more likely.
Reason being that most have experience in OOP because its the prevailing meme design pattern, as most languages are designed as an object oriented language first and foremost, and it's taught on the majority of "beginners" or "starting programming" tutorials or academic institutions as "best practice for all scenarios"; which just isn't the case.
918fd0 No.14731027
>>14729059
I know that feel. I have so much necessary to the game missing and I struggle with accepting that it's a demo and not a release, and then I've only gotten an hour or two in per day for the past week.
924c11 No.14731281
>>14727401
I think Character Controller is what you're looking. Look it up on the Unity documentation.
9526f9 No.14731443
>>14731427
I would space out the three sets of strings at the start, and the "scream" voices that make up the melody should be louder, or maybe there's one more voice that's needed on top.
330623 No.14731500
>>14731427
in the intro i think it'd be cool if the higher strings in the beginning had a heavily downward bent pluck before droning on
during the intro's low droning something heavy and low should be building up along the steady droning of the strings
quieter tambourines, or more reverb on them to drown them out some – they take away from the screamy notes coalescing.
during the next recess either the tambourine or the scream should take a minor role in dragging into the strumming to make the transition less hard from loud to quiet
ec76a9 No.14731681
>>14726933
Yeah I mean that much is true but I don't think it's related to /agdg/ beyond his presence and we've kinda been trying to erase it for the past several months. He doesn't pop up here anymore, and I'll take that. By posting his faggot handle you risk gravitating him back here and I don't think anyone wants that noise.
7b2e39 No.14731725
>>14731427
The song is plagued to hell and back by echo. It's fine on the first few notes, but when they start playing a bit quicker a few seconds in, the echo merges into the notes and it all just sounds a bit crap. You should either turn the echo way down, or make it manually instead of automatically.
The melody itself is kinda hard to follow, and it's far too short. Hell, it's shorter than the intro.
8b3844 No.14732359
Hyper Light Drifter except it is an MMO with psuedo-permadeath, no levels, all stats and abilities come from items (no account-bound items).
Does this sound like something you would play?
0b20e3 No.14732403
>>14732359
>indie MMO copying a dark souls ripoff
53f78c No.14732449
>EventHandler almost done
>just need to finish off physics
>tfw
>>14732359
>no levels
Explain please.
Is it randomly generated?
8b3844 No.14732450
>>14732403
>tfw you read a post so fucking stupid you though you were on cuckchan's /pol/ board
>>14732449
No experience levels is what I meant. You don't progress by killing things and gaining EXP, you progress by having better equipment.
ec76a9 No.14732464
>>14732449
mscuse me goy, it's called procedural generation not random generation. we use procedures okay? its actually much harder than handcrafted content. from now on only use the term procedural generation not random genreation. its not random its procedural, which is very hard and also very good so pay us money.
0b20e3 No.14732466
>>14732450
>ask for opinions on an idea
>get legitimate response back
>lmao you're from cuckchan
Starting off on the right foot I see
53f78c No.14732472
>>14732450
Not a bad idea, but this >>14732403 anon is right, it sounds like a rehash of a rehash of an idea. Top-down action RPG sounds a lot better than "soulslike lul".
>>14732464
Kek, I fucking hate those faggy devs who have to use such pretentious terms
1563cf No.14732476
>>14732464
>he's still using procgen
>not using superior swift generation during your agile scrum sessions
7270ff No.14732508
>>14730718
How easy to use is it? I have to go one man army with way too many hobbies mode, but if there's a fair chance of success I'd love to.
0b20e3 No.14732511
>>14732508
Making a game is a great way to learn programming, so even if you don't know too much about it you'll probably pick it up really fast.
Anyone can make a good game if they care enough, have enough autism and just don't give up.
Honestly, give it a shot, google whatever problems you have and see where you get. You'll eventually learn how to do stuff by yourself by learning from other people's questions and docs.
899429 No.14732516
>>14732511
u sure like that meme spacing format
1e55fb No.14732520
>>14732466
Your "legitimate response" was a half-assed redflag meme that trivialized and oversimplified my post. Every game is just a ripoff of a ripoff at this point, that isn't a criticism of any kind, just a banality.
7270ff No.14732523
>>14732511
I tried in the past. Several times.
I can't take care of music, assets, writing and programming all at once, for it will take a decade to finish. But game engines have changed a lot in 15 years so I reckon maybe I actually have a chance. If Bloodstained ends up reviving the genre I'd like to do a 2d-3d metroidvania. Seems simple enough.
0b20e3 No.14732537
>>14732520
Alright lemme expand on that.
First of all, the biggest red flag from your posts comes from "MMO". I don't know what is it with people that want to start making games, but they ALWAYS start with an MMO.
But what is the problem with MMOs? Well, the biggest problem is that they're extremely complex to make. The shit you need to know to make an MMO is not entry level, and it's certainly not attainable by an indie team or a one person team in any way that would matter.
On top of that, MMOs cost a lot of money to make AND to maintain, because of course you will need to host the servers yourself.
If that weren't enough of a killer, on top of that you have the fact that the demand for MMOs is at an all time low, and that because of the inherit nature of an MMO, if there are no players playing the game already no new players will ever start.
This happens constantly with online-only indie games, so you can imagine how much of a failure making a full on MMO would be.
About the Hyper Light Drifter part, that's more personal, I'll admit that. I just think there's been a huge wave of these type of games that look "cool" but unfortunately none of them do anything very special. So you just have tons of games that are all kind of mediocre because they're all extremely derivative.
Nothing stops you from working on what you want to work on, but if you ask for opinions on why something is shit then you will get unfiltered opinions back.
I just supposed in current year people would know about the problems of making an MMO, and I wouldn't need to actually elaborate.
>>14732523
An engine won't really help you in the creation of assets and that's usually a big problem.
What I'd suggest is try to program a prototype out with placeholder assets and then trying to find an artist on twitter or tumblr to collaborate with.
I suggest checking out the #pixel_daylies (https://twitter.com/hashtag/pixel_dailies?f=tweets&vertical=default&src=hash) hashtag to find people you could work with
ec76a9 No.14732542
>>14732472
>>14732476
anon I'm very sorry but I kinda baited you. Procedural generation is often random but not necessarily so. Any random generation algorithm would use a procedure to do so, however. The difference is that "procedural" implies a set of standard procedures being taken to create a set output, but it can often be given a particular set of inputs that are fixed, which is often boiled down into a seed, which would make it not random. This particular case of term usage is valid.
>>14732537
>>14732511
I really want reddit to exit the building. Tap enter once, not twice if you'd like to fit in better. Until then, get the fuck out kindly please thank you.
1563cf No.14732551
7270ff No.14732553
>>14732537
Thanks. Great info. If all else fail I'll just join another project.
Are there any sites for that too? So far I basically lurked on /agdg/ threads until someone got interested.
8ee087 No.14732555
085fa2 No.14732556
>>14732359
Sounds neat, except the mmo part.
Though, you shouldn't give a fuck what these anons, nor what I think; though that doesn't mean you shouldn't be pragmatic with the scope of the game you're going to be developing.
In the end, you need to be making a game you want to play, and if you enjoy it then that's what matters.
If these anons, or anyone else doesn't like your concept or what is your visioned idea; well then fuck them as far as I would be concerned.
As you're making the game you want to see, and thus it would be a game for you… not for some random anons.
Also, if you're concerned about an audience, there is a crowd out there that has nearly the same taste as you; you just need to find a way for them to notice it.
>>14732520
>Every game is just a ripoff of a ripoff at this point
this why the whole "soulslike" distinction is retarded, as is shit like "rougelike"
Mostly because developers latch onto this, and games in this "genre" are shat out following the standard formula; where in another case if they just took some inspiration, and built ontop of it deviating from said formula it could've been an interesting game (so, wasted potential).
See RoR, it has some elements from DS, but it took inspiration from it then built ontop of it, and innovated said formula in an interesting way; it is definitely not a soulslike, as it innovates beyond the scope of DS.
0b20e3 No.14732557
>>14732553
Any place that has a lot of artists. Tumblr would probably work as well as Deviantart, however they're both really hard to navigate and finding new content in them isn't quite as easy.
Twitter works well because you can also check what the person is like outside of their art so you don't pair up with someone that ends up being an asshole or a SJW later on
085fa2 No.14732559
>>14732555
>(you)
>trash
see i can do it too :^)
ec76a9 No.14732561
>>14732559
C# really is just C++ for babies though.
8ee087 No.14732565
>>14732559
Good luck being tied to Microsoft. Even Microsoft owns Mono.
085fa2 No.14732576
>>14732561
>>14732565
>language wars
>so trivial it's a waste of posts
>made a joke bait post implying it's a topic with trivial points
>anons misses my point
swallowed the bait, nice
0b20e3 No.14732577
>>14732576
I somehow doubt most of these posts are from people that usually lurk /agdg/
53f78c No.14732584
>>14732576
We can all agree Java is shit though
t. javafag
8ee087 No.14732586
>>14732576
Why are you mad? It is only video games.
ec76a9 No.14732598
>>14732584
Java is fucking pajeet-tier. I don't understand how or why so many people still use it.
53f78c No.14732606
>>14732598
I learned Java because I wanted to make Minecraft mods when I was younger
085fa2 No.14732623
>>14732577
pretty much, as u know thread quality always gets worse around demo day, and lingers a bit after DD
>>14732584
>>14732598
every tool has its usage, every tool has different contexts for usage where it is superior.
However, for certain tasks some languages are objectively superior, but in other cases that superiority is so trivial or so easily mitigated/unnecessary that the convenience far outweighs said trivial superiority.
So on and so on, the point is to be pragmatic.
We've had this discussion too many times.
>>14732606
exactly, a context where using java is perfectly valid.
ee5119 No.14732673
So with Unity Asset Store having a massive sale, is there anything worth picking up? Or should I just do everything myself?
8ee087 No.14732678
>>14732673
>Unity
Save your money and just switch to Godot.
ee5119 No.14732685
>>14732678
I'd rather not switch engines after having written a quest, dialog, inventory system all by myself in unity.
515faa No.14732688
>>14732673
I picked up I2 Localization, for $20 it's not worth writing your own localization.
5f315c No.14732694
Trying to decide which Support/Specialist spell would work better for my RTS project. Anyone care to spare some advice?
>Option A: Care Package
>Drops a crate full of health kits and mana
>Can crush enemy units
>Near instant support
<Health kits can be stolen(?)
<Doesn't last that long because of its nature
>Option B: UFO
>Summons a UFO
>Has high health, can tank enemy fire
>Can abduct enemy units, disabling, then killing them
>Functions as a temporary unit
>Stays on the map much longer than the care package
<Abduction takes awhile
<Abduction doesn't actually DEAL damage
<Doesn't immediately give out health/mana
<Slow as fuck
8ee087 No.14732705
>>14732685
Then be prepared to be constrained and oppresed by Unity's every whim.
085fa2 No.14732719
>>14732705
>???
>pokemon battle music starts
>the true autist has appeared
>true autist uses "OPEN SOURCE RULES"
>it is ineffective
515faa No.14732726
>>14732705
Stallman pls go and stay go.
I'm going to have to perform some kind of triage before the demo. Just finish polishing GUI and implementing music probably. I really wanted the two maps to be polished as well but I don't think there's time.
POST DEMO STATUS
40e122 No.14732734
>>14732719
>>true autist uses "OPEN SOURCE RULES"
>that feature hasn't been implemented yet but it's on the roadmap
8ee087 No.14732736
>>14732719
>>14732726
Good goyim! Please keep using Unity for me, my profits are getting terribly low!
7270ff No.14732749
>>14732557
I actualy meant finding some projects to join as a writer, musicguy or level designer, relatively new to it.
Twitter does help a lot. There's a lot of talent there.
085fa2 No.14732752
>>14732726
looking pretty clean anon
The UI equations disappearing when u click one, or finish one is odd, though I suppose it's a "focus on this task" type of deal. Though from a game design perspective it breaks the flow of the gameplay.
It may be more satisfying if you leave the UI menu for the equation you're doing after completing them, and have a satisfying "explosion" or dissolving of the UI equation when the enemy is taken out. During that time after solving it, set it so the color slightly changes (already solved color) and is set to be more translucent (giving a feeling of, this is disabled).
Though, that would mean you'd have to leave the UI equations always visible for those UI effects to be consistently seen.
So, when u click an equation, the other non-active equations have their opacity lowered, so they're less visible, but you can still see them so u can plan ahead which one you want2 do next.
Then u can focus on your equation, but, you can also see the others so it won't disturb the flow of the gameplay.
>>14732736
My dad works at nintendo u know!
515faa No.14732799
>>14732752
Thanks anon, I'll take your feedback into consideration. I especially agree that being able to plan what equation to solve next is a missing element right now. I'm pretty sure I can make solving an equation really juicy with a bit of work. Honestly, 90% of the work on this game is just juice, at least that's what it feels like at the moment.
5f315c No.14732826
HookTube embed. Click on thumbnail to play.
>>14732726
Unless you're making the other equations invisible for "performance " reasons or some shit, why not copy the Typing of the Dead model of emphasizing whatever equation is being worked on? This'll let you plan out your next equation, or something.
I realize you're making this as a mobile game first and foremost, but I guess if you wanted to, you could probably make this keyboard compatible easily
085fa2 No.14732831
>>14732752
Just to clarify here
>leave the UI menu for the equation
meaning just the UI display of the 2+1, etc.
Not the actual answer menu, that answer menu part is fine.
Also, what happens if u get one wrong?
Do the enemies get a buff or something, and the equation changes?
Humm, I could actually imagine this being really cool with a few of the aspects from HoMM3 borrowed.
>>14732799
true you'd definitely need a lot of juice for that type of game.
515faa No.14732879
>>14732826
I think my original thought was that I have significantly more UI elements than Typing of the Dead, so I would hide the others to help the player focus. But as 085fa2 said, I think fading surrounding elements significantly might work as well, hadn't thought about it before.
>>14732831
>Also, what happens if u get one wrong?
As of right now you just lose your score multiplier but I am toying with some ideas. For example, making the enemy move/hit faster. I am currently not switching out the equation since I figure it might get frustrating for the kids, but it's on the table at least.
918fd0 No.14733022
>>14732726
>POST DEMO STATUS
ec76a9 No.14733166
So I thought I should implement voxels by making a map of a vector and a pointer to voxel object, so I don't have to worry about processing large areas of empty space (no voxels) when drawing them. Am I retarded or am I onto something? I know I can use an octree but honestly fuck that noise, at least for now.
fdd50f No.14733186
>>14733166
If you aren't using an octree I hope you don't have many voxels.
ec76a9 No.14733192
>>14733186
It's just a sculpting app I wanted to make for my sister's kid, so I'm not dealing with a lot
d11af5 No.14733236
>>14732726
>POST DEMO STATUS
I'm in the final stretch, fucking kill me.
ec76a9 No.14733240
>>14733236
this is what a shoggoth looks like, im fucking convinced.
53f78c No.14733250
>>14733236
What circle of hell has that thing crawled from?
e2a28b No.14733266
OpenGLfags help
Is there a way to repeat a tile from a tilemap over the entire surface of a polygon? I know you can repeat a texture by stretching the texture polygons outside of the texture bounds, but how about a small part of a texture? If you do the same, you'll just get other tiles into the polygon.
d11af5 No.14733277
>>14733266
Here you go: https://www.khronos.org/opengl/wiki/Array_Texture
If you insist on using a shitty non-array texture atlas, then your only options are to either break up the triangles into smaller triangles or to calculate the texture coordinates with math in the fragment shader. Both of these are shit compared to array texture.
515faa No.14733283
>>14733266
Just calculate your UVs properly?
934df2 No.14733284
>>14733250
The depths of Google's deep image analysis AI.
e2a28b No.14733300
>>14733277
Oh, I think I read something about that before but forgot all about it. Thanks, I'll look into it.
40e122 No.14734127
>>14732726
Added some cliffside houses for max comfy. Needs real textures, but stock bamboo will do for now. Also added a pause menu and an options menu, currently only lets you change the level of MSAA and quit.
1b1a7d No.14734148
>>14734127
That does look comfy, what's the game?
40e122 No.14734182
>>14734148
3D puzzle-platformer but you use a snow/sandboard to navigate the levels. May add other movement modes such as walljumps and hang gliding depending on if it's fun.
1b1a7d No.14734472
>>14734182
That looks fucking great, love it.
I think the gameplay GUI is passable in it's current state, I'm going to move on to get music in the game. Will have to iterate more on this later, try out above anons ideas about fading equations instead of hiding them.
482a3d No.14734551
>>14734472
What happens if you mess up an equat ion?
1b1a7d No.14734689
>>14734551
Right now you just lose your score multiplier, the equation stays the same and the answers stay the same.
f7a57c No.14734705
>>14732726
>POST DEMO STATUS
demo is likely kill but progress is made
1563cf No.14734834
>>14734705
I find the pauldrons covering an invisible, nonexistant limb, and still deforming, to be endearing
f7460d No.14734835
>>14733250
it's just a frog with a filter
d11af5 No.14734897
>>14734835
no its real I recorded it in my backyard with my phone
c15ef0 No.14735040
Still need to make more 7 or so trees with their platforms and the community center in the bigger platform
c15ef0 No.14735045
Is there any fully rigged with IK characters to make mecanim compatible animations rolling around?
I don't want to settle for this "You walk on grass to cut it" solution, and setting one up will take too long and I will miss demoday
Also, I could get some replacements for the running, talking and crouching animations too, especially running in non-forward directions
c15ef0 No.14735050
>>14735045
>running, talking and crouching
running, walking and crouching
085fa2 No.14735094
Anyone have experience with the scriptable rendering pipeline as of yet?
Is it stable enough to worth investing time into it as of yet?
>>14735045
>fully rigged with IK characters
Just use the rigify blender addon, it takes around 10-15 minutes to perfectly align the rig to the character, and it then generates the IK/FK rigs.
There's some steps u need to take to make it compatible with unity though (just search: unity docs rigify).
>Also, I could get some replacements for the running, walking and crouching
For quick animations look at the maximo animations. They're mocapped so they're pretty quality, and serve as quality placeholders.
Make sure they're skinned, and in unity format when downloading. Then just ctrl+D to separate from the prefab, as u probably already know, but just to be safe i'll mention it.
f7a57c No.14735109
>>14734834
ha, I'm glad you like it. My modelbro did a really goot job at animating this
I'm currently working on properly texturing the armour. I think I'll have to add a zoom option to make the Holy Mother visible
c15ef0 No.14735129
>>14735094
>Just use the rigify blender addon
We tried this when we were making the characters, we couldn't figure out how the use the goddamn character thanks to super outdated manual, actually, the first result in google for the search was a 404 page (https://docs.unity3d.com/Manual/BlenderAndRigify.html)
I am not allergic to trying again, but considering there are only two days left, I'd rather download a working model instead
c15ef0 No.14735208
Grass Patch ready for demo day
085fa2 No.14735215
>>14735129
>first result in google for the search was a 404 page
man they're breaking even docs in this latest beta, here's the old doc link:
https://docs.unity3d.com/2017.4/Documentation/Manual/BlenderAndRigify.html
>we couldn't figure out how the use the goddamn character thanks to super outdated manual
Is it getting it into a unity compatible format, or using it in blender?
085fa2 No.14735237
File: 01bd8252e6c4c56⋯.png (Spoiler Image, 260.98 KB, 618x649, 618:649, 01bd8252e6c4c566772bea0b6e….png)
>The first deferred shaded video game was Shrek, an Xbox launch title shipped in 2001.
>it's ogre when u start using deferred shading
>DID NOTHING WRONG
couldn't resist
c15ef0 No.14735238
>>14735215
>Is it getting it into a unity compatible format, or using it in blender?
getting it into a unity compatible format
using in blender is trivial, so much that makes me wonder why there is no auto-import for it already
085fa2 No.14735263
>>14735238
ah, yeah that link i provided gives u the steps.
pretty easy todo, I'd say 20-30minutes maximum w/o detail weight painting.
save this vid for when u do it: https://www.youtube.com/watch?v=dYiAd_08-0k
as a guide on properly setting it up + generating.
after that use the rigify to unity guide
Anyways, it can wait until after DD, u don't need a rig with IK/FK for humanoid animations (auto-retargeted mechanim animations); just a proper humanoid rig hierarchy (https://docs.unity3d.com/2017.4/Documentation/Manual/Preparingacharacterfromscratch.html).
8b3844 No.14735333
How to 3/4 perspective art? What are the rules? How do I make my pixel art conform to this? Think Legend of Zelda but a little less top-down.
40e122 No.14735414
158d32 No.14735680
> I managed to squeeze in the time to implement the tile map drawing before DemoDay
Yay. Thanks again to the anon that provided me with the tiles I made use of
515c2b No.14736477
>C++ project
>gotta connect to (MySQL) database
>every library and connector are broken, don't have good Find___.cmake files and requires code that fucking stinks
>by the time I find this all out, I'll have already written a better find module
I've been at this shit for fucking hours.
Advice?
330623 No.14736557
>>14736477
Must you connect to an external MySQL instead of SQLite
Can you use a more complicated solution that comes with interchangable SQL drivers, since they probably already wrangled it?
515c2b No.14736581
>>14736557
This is the backend of my application which logs data and permanently stores it in that SQL database. This DB will then be queried and the latest samples retrieved to be presented to the user.
I would love an interchangeable SQL driver if they all didn't suck for C++.
I could use SQLite, but I'm sure integrating it into my project on fucking windows is gonna suck as well. I swear, every project out there has some fucked up CMake garbage that's half broken. What the actual fuck.
8c0cb1 No.14736630
>>14732726
>POST DEMO STATUS
Working on some final stuff before demo day. Tonight and tomorrow is trying to wrap up the demo with some bows and ribbons so it at least has some extra feel to it. Still very basic, and is more of a proof of concept than a demo. Don't expect too much. The plan is to have a lot more to show next demo day though. We almost went full teaser trailer for this one, but ultimately it felt like people would be more interested in actually playing a tech demo, even if it is kinda bare.
253087 No.14736762
>>14732726
Started working on it earlier this afternoon and I'll probably be done tomorrow because it's fucking garbage.
96c193 No.14736847
Finally got my script working on adding cloth meshes at runtime so customizable outfits are possible. The physics for it are really finnicky but I don't think I'll ever run into a situation where she sways back and forth really fast like this stress test.
>>14735208
Looks like fun!
1563cf No.14737201
>>14732726
>POST DEMO STATUS
Completely non-interactive. Did a bunch of refactors for behavior and stuff. Have to sleep now, then do an 8 hour shift, so I probably won't get back until like 8 pm tomorrow. That'll leave me with some time to pull an all nighter and try to power something out, but I'm not hopeful.
I've been trying to make a GameState object that's just a POD, with a number of StateController things attached to it that either do the main update work, or else have a number of aux functions that object behavior can hook into.
Anyways, here's my current build if anyone wants to laugh at my code
https://github.com/JackalsoftGames/AutumnWillow_sfml/tree/Pre-demoday-prep
0b20e3 No.14737467
YouTube embed. Click thumbnail to play.
This might be the most useful Unity talk I've ever watched
I cut down my Light Bake settings by hours
8949f9 No.14737473
I want to make nice pen art for my game. How do I draw with pen?
0b20e3 No.14737483
>>14737473
I think you might want to ask this to people that draw, not people that make games
8949f9 No.14737489
>>14737483
most people here don't make games
0b20e3 No.14737502
>>14737489
You can leave any time
8949f9 No.14737504
>>14737502
make sure to tell us when you drop your project, I'll let you know how my art is going
0b20e3 No.14737511
>>14737504
Since you can't draw with a pen I doubt you'll be much further along than I am.
But I'm sure your positive and friendly attitude is sure to get you to places.
8ee087 No.14737521
Had to submit an example project to demonstrate a bug in Godot. Hope they like it.
https://github.com/godotengine/godot/issues/18605
8949f9 No.14737522
>>14737511
I don't need a "positive and friendly attitude to get me places", I'll go there on my own
40e122 No.14737611
>>14737521
You madman, you actually did it.
515faa No.14737659
8ee087 No.14737690
Tiles imported straight from Rome, but what is with the strange inscriptions on them?
96c193 No.14737765
Have you guys noticed that in Unity, if you make an animation with shape key animations, that animation won't work on your mesh unless you apply it to a mesh that has those shape keys? Or is it just me?
I was inspecting the Animation tab and I think it has something to do with how Unity binds the animation to the target skeletal mesh internally. Shape key tracks seem to precede bone tracks, so even if it has the same bones, because it has shape keys that precede the track list, it can't bind 1-to-1 because it compares shape keys to bones so it just fucks up and can't bind and play the animation. Can someone hopefully prove me wrong?
dabc19 No.14737858
Is there a way to draw an empty circle in SFML? I can only get circles filled with a solid color.
0b20e3 No.14737874
>>14737765
Isn't this how it's supposed to work?
>>14737858
I don't know about SFML but you're probably looking for "Stroke" and some kind of transparent setting for "Fill"
515faa No.14737920
>>14737858
Can't you do a circle with transparent fill color and an outline?
dabc19 No.14737950
>>14737920
That's what i'm trying to do but it fucks up really hard for reason i don't understand.
It moves the circles position when i change the radius dynamically and everything just goes bonkers.
96c193 No.14737954
>>14737874
>Isn't this how it's supposed to work
Yeah but this is talking about animations with bone and shape key animations. So even if the bones match, shape key amount differences break the binding.
515faa No.14737962
>>14737950
Sounds like there's some other problem in your code, is your origin set at the center of the circle? Are you using negative outline thickness to not expand the radius of the circle/accounting for said expansion?
dabc19 No.14737973
>>14737962
Are you telling me that by default the origin is not the center?
515faa No.14737990
>>14737973
Should be pretty easy to test, but it might be the top left corner (0,0), for consistency.
ec76a9 No.14738144
>>14735237
My brain is telling me that Tetra has suffered enough, but my heart is telling me to bully him for fun. Any more pics on the subject?
f7a57c No.14738211
>>14738144
Listen to your brain
0b20e3 No.14738228
Can I replace a State Machine with a substate machine without having to remake all my animation transitions?
I've got about 50 transitions coming out of my idle and I'd rather kill myself than write every single one of them by hand.
Using Unity
330623 No.14738245
>>14736581
If you're not married to a particular SQL source then I'd build a simple-ass API to do all the data munching for you server-side.
Also fucked up CMake garbage is why I avoid C++ and I typically end up having to fix up the CMake files in some capacity when I need them.
330623 No.14738250
>>14738228
You can just drag the transitions into a sub-state machine no problem. As long as you're not using the sub-state machine's default entry point (you won't be at this point) you will be fine. Just make sure you're using version control in case you fuck something up.
0b20e3 No.14738255
Realised I needed more moves that increased horizontal speed. Took the opportunity to add a sleeping animation.
>>14738250
I don't think I can drag already connected transitions to a new state machine honestly
Is there some kind of shortcut that allows me to do that?
e2bb69 No.14738450
>>14737690
mmm… those are some nice floor tiles…
ec76a9 No.14738515
515c2b No.14738575
>>14737858
https://www.sfml-dev.org/documentation/2.4.2/classsf_1_1CircleShape.php
oughta work:
>sf::CircleShape::setOutlineThickness( whatever );
>sf::CircleShape::setFillColor( sf::Color::Transparent );
Don't forget to set the outline color
515c2b No.14738596
>>14737973
"The default origin of a transformable object is (0, 0)."
That's the top left corner of the square's bounding box in SFML's coordinate system.
https://www.sfml-dev.org/documentation/2.4.2/classsf_1_1Transformable.php#a56c67bd80aae8418d13fb96c034d25ec
Found yer problem, ho!
f25504 No.14738823
>Making demo for Demo Day.
>In the process, figure out a solution for major problems in movement, collision, and drawing.
>Almost all of the work I did for the demo needs to be redone to use this new solution, and lots of other shit needs to be tweaked.
>It's so much better that not having it in the demo for tomorrow would feel like an embarrassing half-measure.
Guess I'm staying up all night again.
40e122 No.14738843
>>14738823
What is the official start time for demo day? UTC 00:00? That gives us 8 hours left.
f7a57c No.14738850
>>14738843
nah, it's usually amerifag time
e796f0 No.14738919
Has anyone read this book? Is it any good?
i guess it doesn't matter because I already bought it but i want some opinions
085fa2 No.14738939
File: 512d30dec6c60c0⋯.png (Spoiler Image, 228.6 KB, 1483x639, 1483:639, 512d30dec6c60c05ee5e531a32….png)
File: 992a418f4dc09fc⋯.png (Spoiler Image, 1.04 MB, 1200x1455, 80:97, 992a418f4dc09fc8143356f951….png)
>>14738919
it has some good info.
i skimmed certain sections to get some low/mid-lvl knowledge for engines; which I was lacking at that time.
Definitely gave me a few jumping off points, and from there I did my own specific research online.
>>14738144
use them wisely
f3cb12 No.14739036
How do people handle game design like combat systems etc. Just a spreadsheet and some guessing?
1b1a7d No.14739092
d11af5 No.14739143
HookTube embed. Click on thumbnail to play.
>>14739036
Understand what you're trying to design and what you're trying to achieve.
When you're making an action combat system, you usually have the following goals:
>You want players to quickly react to things happening, possibly triggering an adrenaline rush. To achieve this, the game should be responsive and the time given to react should be small, yet still do-able for your target audience. (Action)
>You want players to quickly analyse the situation and evaluate the next course of action. To achieve this, your game should have clear advantages to some approaches, and make switching between them on the fly easy. For example, a weapon swap should be a single button press away, or maybe your weapon has an alt-fire for long-range use. (Tactical)
You'll find that almost every game out there uses a mix of both. Even something as simple as One Finger Death Punch, which on first glance is just "QTE the game" (see embed), but actually has some tactical decision making in it:
>Performing an attack moves you in that direction. Hitting enemies on the right will give you slightly more time to react to enemies on the left.
>Occasionally you're giving a throwing weapon which you can throw in either direction to instagib an enemy. This allows you to pick off enemies you have issues dealing with.
>Occasionally melee weapons drop on the ground for you to pick up, giving you bonus range. Deliberately attacking in a direction where there is no enemy will count as a miss, but can be utilized to move towards a melee weapon and pick it up.
Instead of throwing in a combat system, ask yourself what experience(tm) you want to deliver.
Most people just want to see cool shit happen on the screen, which is why the industry has so much QTE garbage.
c15ef0 No.14739162
>>14736847
We are having the same problem with our clothing
65a458 No.14739187
>>14715211
Some code is just unavoidably shit though. In video games, there are parts where you're just going to have a bunch of nested if statements, there's no way around it.
f773f1 No.14741974
YouTube embed. Click thumbnail to play.
>tired
>hungry
>retarded
>trying to understand this through the finn-jew accent and tuneless generic chip music playing at the same volume