[ / / / / / / / / / / / / / ] [ dir / animu / arepa / asmr / leftpol / tacos / vg / vichan / zoo ]

/v/ - Video Games

Vidya Gaems
Email
Comment *
File
Password (Randomized for file and post deletion; you may also set your own.)
* = required field[▶ Show post options & limits]
Confused? See the FAQ.
Flag
Embed
(replaces files and can be used instead)
Oekaki
Show oekaki applet
(replaces files and can be used instead)
Options

Allowed file types:jpg, jpeg, gif, png, webm, mp4, swf, pdf
Max filesize is 16 MB.
Max image dimensions are 15000 x 15000.
You may upload 5 per post.


<BOARD RULES>
[ /agdg/ | Vidya Porn | Hentai Games | Retro Vidya | Contact ]

File: 05807eacae74613⋯.png (835.2 KB, 661x544, 661:544, ye olde agdg.png)

File: 2638106755175d2⋯.png (1.5 MB, 1920x1080, 16:9, AGDGlogo.png)

File: 17bfe2b3f6fcbfa⋯.png (126.34 KB, 2604x3139, 2604:3139, agdg jlmk.png)

3cc89b  No.15445185

Classic Edition

Resources

>>>/agdg/

>>>/vm/

>#8/agdg/ via irc.rizon.net

>Dev resources: http://8agdg.wikidot.com/resources

Links

>Wiki: http://8agdg.wikidot.com/

>Beginner's guide: >>>/agdg/29080

>Previous thread: >>15408260

Announcements

>QUARTERLY DEMO DAY SCHEDULED FOR November 11th

>Please contribute to the wiki if you can

3cc89b  No.15445204

File: 09beac6fd70e0db⋯.webm (8.32 MB, 1208x514, 604:257, 2018-09-13 13-33-22.webm)

Ok faggots, gonna shill some free asset I've found in the asset store so you niggers don't end up buying something that this already does

https://assetstore.unity.com/packages/tools/animation/sox-animation-toolkit-110431

I am using this for hair, took quite a while to get the numbers right. One other thing this does is handling collision (since it does have a place where you can set colliders), but I still need to investigate how it works.

It seems to work quite with inverse kinematics, where it sets a weight at a certain distance, it tries to keep the weight at a certain point relative to the root of the bone, the difference between the current position and target position is used to rotate the bone. This also works with anything, really, does not need to be a skinned mesh.


0b94da  No.15445241

https://www.youtube.com/watch?v=F6VerW98gEc

godot getting a new tilemap feature.


05f180  No.15445245

File: d27ab4bbed6a949⋯.png (155.08 KB, 370x259, 10:7, waterfox_2018-05-08_06-44-….png)

Where should you be advertising your game while in development if you wish to make it a profitable release?


3cc89b  No.15445250

>>15445245

reddit probably


0b94da  No.15445255

>>15445245

look up where people that maybe interested in your game may gather.

you should have at least a fanbase before trying to monetize your games.


ffc71c  No.15445314

>>15445245

Reddit or Youtubers


3cc89b  No.15445381

YouTube embed. Click thumbnail to play.

>>15445245

Early access and giving it for free for every single curator that seems even remotely interested in this type of game.

As dirty as it is, it seems that Early Access games make more money than ready on release games, at least according to the faggot in embedded


05f180  No.15445510

>>15445381

I'm guessing like kickstarter there are a lot of early access news sites which talk about it.


ec2955  No.15445524

>>15445314

Sometimes it's hard to figure out who to send games to. I know that the few Jewtubers and streamers I know wouldn't care for that type of game.


fdd87e  No.15445702

File: 2a36361cd705083⋯.mp4 (6.27 MB, 1280x720, 16:9, 2018-09-13 20-44-32.mp4)

It's back

working on key features


67d81d  No.15445716

I still need to implement drae call sorting so i dont have a bajillion texture swaps. I don't even have the bajillion texture swaps implemented yet


a0405b  No.15445718

File: 282093dac4ce76b⋯.png (298.59 KB, 539x295, 539:295, cool.png)

>>15445702

Nice work anon. Great choice of sound effects.


154ece  No.15445727

File: d9a64189574ea43⋯.webm (3.73 MB, 640x360, 16:9, purring chicken.webm)

>>15445702

>pet the chicken


3cc89b  No.15445738

>>15445702

are you going to make a cow level?


fdd87e  No.15445747

>>15445738

>>15445738

I kinda have an idea how/where to hide it

no one may be able find it


636ea0  No.15445752

File: 135e75f0671f97e⋯.png (Spoiler Image, 708.26 KB, 640x739, 640:739, ClipboardImage.png)

>>15445702

Do these chickens drop meat or full plate armor after being killed? Or maybe both?


be7812  No.15445757

>>15445381

It's because of the hype machine. Selling an incomplete game works on normalfags because they like being excited about the next update. The promise of more to come makes them feel like they're getting more for their money (even if what they're paying for now is basically a ripoff). Additionally, normalfags are idiotic beasts driven by the need to socialize so early access is another plus because it gives them something to talk about. They can blabber on about how much they loved or hated the latest update.


fdd87e  No.15445761

>>15445752

neat idea, though I'll likely keep things in town 100% non-violent the chicken level, on the other hand…


3cc89b  No.15445762

File: b825fc8ad7614a0⋯.jpg (207.05 KB, 1280x720, 16:9, smug olivia.jpg)

>>15445747

>He thinks devs can hide things in the age of datamining

Not telling you to not try, but don't be surprised if it is discovered on the first day


67d81d  No.15445763

>>15445762

Depending on how much Hype™ his game has, people might not even bother


fdd87e  No.15445766

>>15445763

>people might not even bother

this is what I expect


3cc89b  No.15445768

>>15445761

There was another game (dungeon siege i think) that did chicken level, if you are doing it instead of cows to be different from Diablo then go for another farm animal like pigs


be7812  No.15445790

>>15445762

There are games from a few years ago that people are still finding secrets in. Sure, people can datamine that there is a secret mega boss that unlocks an alternate ending, but it's another thing for players to legitimately trigger that.

Or you can do the ugly Edmund McMillen thing where he made a proprietary file format for BoI in an attempt to keep dataminers from discovering all the secrets right away and spoiling the whole thing for people.


e2b1b9  No.15445810

File: 7b95b9d7703f3df⋯.jpg (287.37 KB, 1600x785, 320:157, art example.jpg)

So I'm looking for art styles and I always like this style on this picture. Is there a proper name for it or it's just "broad strokes"? I want to search for tutorials that teach me how to paint like this.

>>15445524

Search for let's plays of popular games similar to yours, for example, if your game is a metroidvania search for super metroid or castlevania sotn if your game is more retro or nier: automata if it has anime titties, or hollow knight or ori for more grounded game. Also use keywords like "review" or "stream" for reviewers and streamers, respectively. Look into the views of their videos and the subscriber amount, they could give you an idea of how popular that creator is.


3cc89b  No.15445816

>>15445768

It was a chicken 'room' rather than level. The chicken level was a custom level filled with the same chicken enemy that spawns in that room.

The chicken room was unlocked by beating all of the multiplayer campaign dungeons and taking their key item, then with the key items in hand you can open the door to the chicken dungeon. It was pretty fucking neat game. I wish Microsoft still made good games


636ea0  No.15445821

>>15445810

Modern impressionism


05f180  No.15445822

>>15445810

That's just a photo with a really bad filter over it. It's a form of abstraction so look for that.

>>15445768

Pigs are lame. Do a chicken, duck and goose level. Each with a different immunity on higher difficulty.


cb2ea4  No.15445854

>>15445822

It's an oil painting you dummy.


e2b1b9  No.15445858

>>15445821

Thank you that's exactly what I wanted.

>>15445822

>That's just a photo

No, it seem's to be a real panting, I took it from this blog

http://artisttbeckham.blogspot.com/


154ece  No.15445893

>>15445768

just do quails instead :^)


3cc89b  No.15445925

File: dfda5bccc578b94⋯.png (2.15 MB, 1134x834, 189:139, rapist bird.png)

>>15445893

>>15445822

What about killer Kakapos?


67d81d  No.15445938

Hypothetically, if someone wanted to make a Resident Evil clone, rather than doing 3D and transforming it, couldn't you just take a bunch of photos of existing rooms in public streets and spaces (eg library) and heavily retouch then, then downsample them to a pixelated background?


3cc89b  No.15445945

>>15445938

Yes, there are filters to pixelate photos for this kinds of purposes.

There was a very robust one a couple of years ago being talked about here, but I can't really remember the details


27fc5d  No.15445966

>>15445938

I'd like to ask you why you think you couldn't.


7a15d8  No.15445997

File: 0ba29d4d0ead630⋯.jpg (1.5 MB, 3888x2588, 972:647, Capybara_2.jpg)

>>15445925

>Not carnivorous capybaras


67d81d  No.15446021

>>15445966

Because as far as I know, that shit was originally prerendered 3D, not photographs


67d81d  No.15446044

File: 6175c392a743b97⋯.jpg (1.24 MB, 1836x3264, 9:16, 20160711_143231.jpg)

File: 685e5f13c1419a7⋯.png (37.66 KB, 320x240, 4:3, 20160711_143231 copy.png)

Yeah, that looks pretty decent, if that was a project I wanted to do in the future


3cc89b  No.15446066

File: c519dc102e77e5d⋯.jpg (92.06 KB, 843x1034, 843:1034, Tsuyu Reeeeeeeeeeeeeeeeeee….jpg)

>>15446044

>vertical photo

REEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE


0b94da  No.15446097

If you had notch type of money, what would you do?

I would try to build some sort of european castle or a japanese house, but more like a cultural center for diferent kind of arts, like a sort of conference house for artist, maybe as a kind of legacy after I die.

Is not like a small european castle would cost less than modern houses on jew jork or some shit.


67d81d  No.15446104

>>15446097

Actually, Italy has a ton of aging castles that are not being maintained. Apparently they have a land ownership grant if you present a modest proposal to restore/maintain it over a period of time. Even something halfassed like "check for weatherworn bricks and put $150/mo towards new masonry" supposedly gets you one.


b83677  No.15446110

>>15446097

Make another internet that's hard as balls to use to weed out normalfaggotry.

Companies are forbidden from using it, especially Google and Co.

Have the servers be located in space so they don't fall under any country's regulations.

Just a pipe dream, but it would be a fun project nonetheless.


0b94da  No.15446123

>>15446104

I live in south america, we don't have castles.

>>15446110

that's a bit harder than making a building.


27fc5d  No.15446128

File: 0402668823a4162⋯.png (112.4 KB, 513x441, 57:49, yeiee2p4.png)

>>15446097

Start an official company called AGDG, hire a bunch of talented developers and artists, then steal everyone's ideas from these threads and finish the games faster and better than any individual dev can realistically do.


b83677  No.15446129

>>15446123

>that's a bit harder than making a buildin

Just takes Elon Musk tier autism and money, I'm sure there's people out there like that.


0b94da  No.15446134

>>15446128

that seems pretty doable and mainly if you had 1MA skill.


8b122a  No.15446167

>>15445810

It's impressionism. Lookup Claude Monet.


fdd87e  No.15446169

File: 671f5bb5ec05863⋯.png (172.47 KB, 664x481, 664:481, ClipboardImage.png)

>>15445204

potato knishes

>>15445816

>It was a chicken 'room' rather than level. T

wew, that doesn't sound too similar as to make me a total fraud

>>15445822

>Pigs are lame

Not kosher enough for ya?

there was the pigasus in another arpg anyway.

>>15445925

>>15445997

cute but too exotic

>>15445893

>quails

that's more like it. pheasants, maybe? nah, let's stick to chickens

>>15446066

not even a proper 4x6 photo. Otherwise it would be acceptable

>>15446044

still, a pretty cool location for a map


8b122a  No.15446172

>>15446097

Build a payment processor, domain registrar credit card system, and bank that wouldn't deplatform people over their views.


3cc89b  No.15446196

>>15446172

Then base it on Silicon Valley and slowly replace all your employees with libtards :^)


67d81d  No.15446198

>>15446169

Photos might have a 4x6 resolution, but N64 had 320x240 for most things


46c219  No.15446276

File: 63a34b7dd93b269⋯.png (422.71 KB, 1000x961, 1000:961, ClipboardImage.png)

>>15446097

Create a company for the purpose of creating the first proper AIs. AI holds incredible potential, but I think it's a genuine threat, and nobody's treating it with the care that it deserves. Someone's going to do it eventually, so there's no point in stalling it, and we might as well do it on our own terms. We need to make sure that someone who at least cares about the future of earth and humanity is the one that makes it happen. Almost every entity working on it is some kind of snake in the grass who low-key (and sometimes blatantly) loathes humanity, with the possible exception of Elon Musk, who isn't working on it directly. I'm not such a fool as to assume that I'd be the one to succeed in imparting some divine, brilliant benevolence to a machine, but someone has to do something. Ideally, whatever form it takes would at least be enough to mitigate or prevent the threat that other AIs represent.


154ece  No.15446313

>>15446169

>nah, let's stick to chickens

do some silkie chickens or shamos


3cc89b  No.15446359

>>15446123

>we don't have castles

we have tons of coastal fortresses though. same shit but for boats


3cc89b  No.15446365

>>15446172

>and bank

You really want some nails in your neck, don't you?


6f779d  No.15446421

>>15446169

Replace every peasant in the game with a pheasant if you hit a certain trigger.


1e3977  No.15446538

>>15445790

>Or you can do the ugly Edmund McMillen thing where he made a proprietary file format for BoI in an attempt to keep dataminers from discovering all the secrets right away and spoiling the whole thing for people.

>and spoiling the whole thing for people.

You're full of shit, nothing beats the hype of legitimately discovering a secret, you can't do it in the age of internet where everything is leaked on day 0, the last time I've felt hype was in Rain World threads were people collectively were discovering stuff and helping each other, better than "here are all the secrets and how you can unlock them, that's everything you'll ever get in this game"

I will probably do the same, either making a proprietary file format or embedding the secrets in the binaries and encrypting them with red herrings everywhere.

If it lasted long enough I'll be satisfied to add even more stuff, of course, I agree that some secrets in BoI are pure trash and impossible to find without a guide and even then it doesn't work that easily, that's what bad about them.


67d81d  No.15446581

>>15446538

One thing I liked is branching scenarios, though. Think of a game like Clocktower 1, where the mansion always has the same rooms, but are mixed up a little bit. As well, certain jump scares may or may not be present in a given run.


67d81d  No.15447519

File: 4f4855c3d7437f4⋯.png (69.68 KB, 810x625, 162:125, 500 draw calls.png)

Made a bit of progress

>Graphics helper object has an internal custom vertex buffer

>In addition to passing the draw calls into the buffer, it also tracks the texture being passed

>If it's the same texture, then it increments a counter, otherwise it starts a new counter

>Textures aren't sorted, but it will group them as necessary

>Optimized this could be ~3 draw calls with texture sorting, but right now it drops it from 500 calls to 336, which is about a third of the calls, so it saves a proportional amount of textures


8ae8af  No.15447761

File: 17445ce8e9783c4⋯.gif (2.23 MB, 600x300, 2:1, Axe__.gif)

File: 163b5293b3af09e⋯.gif (1.53 MB, 600x300, 2:1, Mace__.gif)

File: e4e4539b5ea7f66⋯.gif (1.94 MB, 600x300, 2:1, Sword__.gif)

Shitty gifs for >reasons


67d81d  No.15447890

File: 0cc42afb29f6d57⋯.png (99.96 KB, 808x627, 808:627, woo.png)

>>15447519

Cleaned up the helper a bit, but polite sage because it's essentially the same update


6ee5ae  No.15448002

>>15445381

I watched this video last night. I can't fucking stand the British "ERM" every 5 seconds.

He made some good points. One main thing is that putting your Games on Steam doesn't mean anything. Early access seems to work for games because instead of having to hype your product then have a release, you have a constant stream of players who will advertise your game via word of mouth.

/v/ had put me off early-access but I think I will go for it. Who cares about integrity when it just lessens your chances of succeeding?


6ee5ae  No.15448031

>>15446097

I would buy some land, build a sophisticated house with a pool and a complete gym, then marry my cousin.


6ee5ae  No.15448073

>>15446276

I don't think you actually appreciate the threats of AI.

It's not going to be terminators, it's going to be computers generating their own languages and passing signals in such a way that people can't comprehend. Eventually they'll be so far advanced they could do whatever is completely unfeasible to people today. All it would take is a single bug or some side-effect of some process/goal in legitimate AI programming and they could completely change the world in a day. It's not terminators, it's impossibly difficult to calculate pathways of lasers bouncing around to kill people from miles away. It's forming some new weapon that annihilates the planet ..


8ec782  No.15448356

>>15448312

That's a mighty lewd image you drew up there.


f48cd6  No.15448417

File: 75f3842329c0b77⋯.png (5.62 KB, 495x588, 165:196, ballspringforces1.png)

>>15448356

does this do things to you?


000000  No.15448544

>>15448179

Your quaternion is around the forward vector which means that the newHop vector is always pointed along the forward vector which you then set to zero (+z is forward in unity) and normalize. You're ultimately setting the transform's forward to zero which I assume Unity doesn't like.

If you want random vectors in the general direction of the target you could do something like this:

float angle = UnityEngine.Random.Range(-45.0f. 45.0f);
Quaternion quat = Quaternion.AngleAxis(angle, transform.up);
Vector3 meToPlayer = player.transform.position - transform.position;

transform.forward = quat * meToPlayer.normalized();


67d81d  No.15448555

How come some rendering systems completely fucking implode when a camera's Up vector is equal to (0,1,0)?


000000  No.15448576

>>15448555

Probably gimbal lock.


67d81d  No.15448607

>>15448576

No, it results in NaN exceptions and usually throws an exception. Speaking mostly from XNA, but I know a few others react poorly to this


7f7464  No.15448611

File: fcb6596d8cea9de⋯.png (38.97 KB, 962x642, 481:321, poide_2018-09-13_22-53-53.png)

I am busy cleaning up my vulkan code, I expect to re-write a whole lot of it because most of it was designed to get results without a real understanding of how to improve it, when I was writing my Vulkan graphics system last summer. So, I have to rewrite it to render things "correctly" because the Vulkan graphics system has by far the worst performance.

I was looking through my code and I found out that whenever I updated a string, I would delete it's corresponding VBO and make a new one, instead of updating it. And I think that in my original renderer I would end up doing this to about every line of text every frame. So, I'm re-doing a huge amount of it until it's acceptable. I won't stop until I can get higher performance than the OpenGL renderer.

Still what i'm making is probably not the best way to do things. Just better than what I had before.


7f7464  No.15448666

>>15448611

and of course, posting my code is the best way of debugging. I wrote VK_FORMAT_R32G32B32_SFLOAT when I should have written VK_FORMAT_R32G32_SFLOAT


c362c0  No.15448687

File: 2fc0994c7ad792a⋯.jpg (167.1 KB, 500x457, 500:457, 1379042692012.jpg)

>>15445204

>Inverse Kinematics

>2 bone ponytail

Nigga you don't need any of that and I think you are overcomplicating it. Do this:

Say your bones are A and B of the ponytail. A is the middle, B is the end.

1.) [oldAPos] and [oldBPos] are the current positions

2.) distance clamp A to the root of the head.

3.) distance clamp B to A

4.) [newAPos] and [newBPos] are the new current positions

5.) Add velocity to A and B based on new minus old positions. And add a little gravity.

6.) Easy physics

7.) Bonus: Constantly apply a force on A to make the pony tail curves outwards instead of constantly be dragged by gravity downwards.


f48cd6  No.15448886

>>15448544

Thanks, that fixed it


67d81d  No.15448908

Anyone mind taking a quick look over the code for my SpriteBatch object and seeing if I did anything retarded?

>Can throw vertices and rect regions at VertexBatch until it fills up

>At this point you need to clear it or draw it or whatever

>SpriteBatch automatically draws and clears its internal VB if it fills up, which is why it tracks a RenderTarget etc and has a begin/end draw call

>Flushing is a combination of drawing and reseting the buffers

>It stores the textures and vertex counts in a list for future sorting (not implemented yet)

https://hastebin.com/etufepebad.cs


46c219  No.15449041

>>15448073

No, I get that. I didn't post X because I think he's realistic. The events of his series would be preferable to what could be coming.


c65cb4  No.15449139

File: 90e29cc3de53cc0⋯.webm (4.07 MB, 800x600, 4:3, demo 2018-09-14 10-08-28-….webm)


c362c0  No.15449502

File: 40164aeda72cffe⋯.webm (6.44 MB, 600x600, 1:1, new_eating_donuts.webm)

First pass at eating donuts.


6f779d  No.15449539

>>15448607

>>15448555

If you have a forward-vector and want to create a the right and up vectors from it, you cross it with the world up-vector to get right, then cross that with forward to get up or in another order, I just woke up and am too lazy to do cross vector math right now.

You need to put an if in there to check if worldup == cameraforward and use another axis instead, else you get a null vector for right, but you can also just limit how far up the player can look.

Now I can't guarantee that's the exact thing they're computing in the engines you talk about, but there may be other reasons they're crossing the camera look vector with worldup, and that's the most likely thing I can think of for problems.

>>15448687

They'll probably use the same system for ever character's hair, including a lot of longer hair, so might as well use the same system for everything.

>>15449502

CUTE


c87481  No.15449844

>>15449502

>feeding gray burned donuts to your loli


a0a47a  No.15449856

File: 0f1450fd086952d⋯.jpg (53.91 KB, 506x480, 253:240, 1447136393220.jpg)

>>15449844

>he thinks it's an actual donut


983f3f  No.15449888

>>15449502

The animations are very expressive and blend together very well. Did you do that all on your own?


845f35  No.15449912

File: 47f706877a815ef⋯.jpg (173.76 KB, 536x565, 536:565, 1467e51af5ad45cb9f6296f448….jpg)

>>15449502

no bully


3cc89b  No.15450046

File: 7ceb6706f00b3cb⋯.png (528.18 KB, 800x600, 4:3, adam hat.png)

File: 98c608ca722989c⋯.png (196 KB, 581x610, 581:610, adam1.png)

File: aee4d4282a62a7b⋯.png (123.4 KB, 581x610, 581:610, adam2.png)

File: dd34aabcf587868⋯.png (144.08 KB, 581x610, 581:610, adam3.png)

File: 449ebdf446e49a3⋯.png (217.54 KB, 581x610, 581:610, adam4.png)

>>15449502

At this point just rename the project "Bully the Shinobu"

I love it

We gotta decide if we are going for cel shading or not, if we do we will need an alternative to the UBER shader


a5c6a0  No.15450062

YouTube embed. Click thumbnail to play.

>>15450046

>"Bully the Shinobu"

Time to make a spiritual successor to Rising Zan.


a0a47a  No.15450101

File: 056b67e0b993cc6⋯.png (58.05 KB, 1366x724, 683:362, single reference multiple ….png)

File: d5f05933a124c8e⋯.png (466.92 KB, 1366x718, 683:359, single reference multiple ….png)

File: 671e77633fa7621⋯.png (466.8 KB, 1366x722, 683:361, single reference multiple ….png)

finally did a thing i should have done a while ago

cross scene reference storages can now reference one object but have their own id to avoid issues with multiple storages referencing the same object

example from before:

we got 4 froggo doggos, same reference

tell npc to attack one of them, works fine

save and reload

he's now attacking a different one because he didn't bother to remember which froggo doggo he's attacking

and the only way to avoid that was to make a new prefab+reference for each froggo doggo

how it is now:

game starts and every cross scene reference gets it's own dictionary for instances+ids

the storages assign ids incrementally, if the preferred id is not specified (-1) they just assign the instance count+1

tell dude to attack froggo doggo

save - not only is the reference to the froggo doggo saved, but the id is saved too

load - enemy is froggo doggo, specifically tell him that he has to attack the froggo doggo with the saved id

hope all of that made some sense

tl dr; i don't need to make every single character unique


9e370f  No.15450495

File: 8ade64e2ade7938⋯.gif (66.76 KB, 256x192, 4:3, thinkign.gif)

hi I'm making a renpy game and want to give the player an option to "ignore" a certain character, basically they press a button and they never see that character's dialogue until they turn him back on

the large size of the script means I can't just put an if statement wherever he speaks, so how do I go about doing this?


3cc89b  No.15450512

>>15448687

well, afaik the thing I am using is hardly as complex as actual inverse kinematics, but look as if it was and uses a similar concept.

Either way, it's working fine, and I really really doubt this will even use 0.05% of the frame time, so it's fine even if it's overkill

It's a free asset, I didn't code this thing, I am just using it. Anything I say about how it works is purely speculation based on the inbuilt debug visualization settings

It also was way easier to set up than Unity's joints and works much better than all of my attempts using all of the joints


3cc89b  No.15450526

>>15449502

Steal the donut back before she can eat it.

Headpat her into the abyss


0b94da  No.15450532

>>15450495

use godot.


9e370f  No.15450549

>>15450532

but I'm a brainlet


6317fd  No.15450560

>>15450495

You either put a condition check before every say statement or you write two different scenes with and without the character, are the other characters not acknowledging the existence of that character?


0b94da  No.15450576

>>15450549

godot programming is literally a python clone, is like not more dificult than renpy code.


be7812  No.15450611

>>15446538

Maybe my formatting was wrong. I was not trying to imply datamining all the secrets out of a game is a good thing. I just think it's a little nuts that he decided he needed to make a proprietary format to keep people from digging through the files. Which all of that work went out the fucking window anyways once he put official mod support into the game.


6b6a9a  No.15450767

I'm trying to work out how major state transitions should work with ECS.

I've read a bit, and the concept of spaces seems to come up a lot. As I understand it, a space is a disjoint set of entities, and spaces aren't supposed to communicate with one another. I'm planning to use entityx, and some of my descriptions will use terms specific to that framework.

My approach, as it currently stands, is as follows:

>a global integer will correspond to the current state

>I'll have a separate entityx instance (which defines the systems to act on the entities, and the entitymanager holding the entities) for each state

>in main, each state will correspond to its own loop, looping while(gamestate == <whatever integer). It will call the update() method of its corresponding entityx instance.

>will have global sets containing groupings of relevant entity handles

>upon transition to state change, will call a system which registers entities from the relevant union of these global sets with its entitymanager

>each entityx instance will have different systems that it calls upon the entities registered with its entitymanager. In a 2d jrpg, for example, there will be a different set of systems for a battle mode and an overworld mode.

>a system which, upon detecting a state change is imminent, will synchronize global sets with entities in current entity manager. Maybe I'll use something like a flag component to figure out where entities are supposed to go.

not sure if i've missed anything, sorry for convoluted explanation. The reason I feel like I need to store entity handles globally is so I can access entities in different states (like if I open an inventory manager and change my toon's sword, it should affect the toon entity on the other states too). Maybe I can also use systems to check for input on menus and such, but I don't really know.

Am I retarded?


6b153a  No.15450811

>>15450767

Sounds like you're over complicating it.

Usually you'd just apply a component flag, and if an entity has that component; you execute that system (or disable that system executing for that entity if lacking said flag).

I.e. isInBattleMode -> execute battle mode systems, etc.

Better to think about the "spaces" as contexts (separating entities with different contextual functionality), which are separating entities/component/system functionality (purely for the sake of usability/mental compartmentalization), but you can definitely use them across contexts (such as accessing a static input entity via a global singleton).

As you need input components from the input context for the game context systems (using input component to move player in game context's movement system via input DIR, etc), and so on.

Also look at entitas, it already has all of that solved, and has a quality wiki for explaining these things + examples.

It's originally build for C#, but its been ported to quite a few languages.


0b94da  No.15450848

File: ee8dec3fed3e283⋯.png (282.06 KB, 1440x900, 8:5, 2018-09-14-132238_1440x900….png)

making the background level.


0b94da  No.15451013

File: 399948bf5ac561a⋯.png (83.84 KB, 400x400, 1:1, building one.png)


6b6a9a  No.15451040

>>15450811

I might be over-complicating it, but I find your post comparably confusing. If I have a space that I name "battle", and I somehow access an entity in "battle" during the "cutscene" or "overworld" state, isn't that somewhat poor form? My idea would be something like putting the playable characters in one space, enemies for a level in another, and menu entities in another space. That way, if I'm navigating to an equipment screen for my characters, I can take the union of playable characters and menu widgets, which is pretty self explanatory. I'm not really sure how I'll handle only needing a subset of one of those groups, though, like maybe two monsters in a battle system.

I understand that systems operate upon a subset of the entities owned by the entitymanager that they're passed, depending on whether the entity owns the components that are relevant for that system to operate.

My confusions is over managing the state of my game, where a given state fundamentally changes which entities are relevant, and which systems act upon those entities at a given time. If I have a JRPG and transition to a battle, I need to draw a battle menu, some kind of background, the enemies I'm in battle with, and my toon. The old draw system that I would've been using while I was walking around wants to draw a bunch of stuff that I no longer want to draw. All of the systems that were acting upon those entities need to stop; I don't want overworld entities to move, be drawn, etc; i expect the state to be as it was when I exit combat. If I'm using the same systems for these two states, I think I'll just have to bloat them with too many checks on the flags of entities, and of course I'll have to meticulously flag everything that I don't want to be drawn/moved/whatever, and remove those flags upon transition from that state. A system like movement wouldn't even be relevant in a turn-based battle, so I'd be flagging EVERY entity to stop them from moving, without disabling the system, which seems crazy. I can't just delete them because I want them to be there when I get back.

Of course, if I only want a subset of one of my spaces, I'll have to flag them (like for battle mode), and then my new systems will have to query the entitymanager only for entities which are flagged, but that seems like the lesser of two evils.

I'm not really understanding what you said about input. I should have entities for input too? Like, an "input" component with the key I'm listening for, and then have the CheckInputQueue system loop over the entities with input components, and emit an event if a key that I'm listening for has been found?

I'll take a look at entitas.


9e370f  No.15451142

>>15450495

figured it out myself, you can redefine say statements into

def n(what, **kwargs):
if narrative:
narrator(what, **kwargs)


0b94da  No.15451233

File: 7a4610545dd46f4⋯.png (623.07 KB, 1440x900, 8:5, 2018-09-14-143802_1440x900….png)


e81a21  No.15451718

Are small stats less pleasant in rpgs? Each stat is between 0 and 100, same with skills, I'm using a bell curve distribution for the rolls

so 50 in stealth has 50% chance of success while 65 has 80%.

I enjoy small numbers in my rpgs, I would've gone with a smaller cap but the current range allow nice fine tuning while not coming up like an odd number, what the player can see can be different tho, it's easy to multiply the numbers by a factor to make them look bigger, but it'll be awkward for humans, starting out average means having 50 in skill which gives a 50% for success, and multiplying values to make them bigger will just make the average grow.


6b153a  No.15451797

>>15451040

It seems there was some miscommunication/misinterpretation.

Contexts are separating the systems/components/entities (i.e. a gameEntity -> uses game context components, and operated on by game context systems).

Your idea of "spaces" are different, and more like "views" in the MVC design pattern (model-view-container); which isn't related to the ECS design pattern (in a purely textbook definition). It is more like an extrapolation for the application of ECS, and not directly written in how to apply ECS within a textbook fashion.

>My idea would be something like putting the playable characters in one space, enemies for a level in another, and menu entities in another space.

So basically, spaces are "versions" of game modes. Changing states switch versions of the game mode; got it.

> If I have a space that I name "battle", and I somehow access an entity in "battle" during the "cutscene" or "overworld" state, isn't that somewhat poor form?

For you idea of spaces it sounds like it would be poor form.

For my original assertion, in relation to contexts, no.

Although they're separate ideas, so, yeah.

Ok, so it sounds like you need multiple different "spaces" which use separate entities, components, and systems.

To handle the problem of requiring a common entity between ALL spaces, you can have a "meta entity" that stores data you need for all "spaces" (inventory data, game state, etc); i.e. a "playerMetaEntity" -> the entity that stores state/meta data that's common to all states/game modes).

When you change spaces you just change a "states" enum in a gameStateComponent of the meta-entity (battleSystemState, overworldState, etc).

Using a switchGameStateSystem, you would switch "spaces" (due to a state change in the meta entity's gameStateComponent), and would enable that space's entities, and disable the old state entities.

Basically, each space that is separate has separate entities. Switch spaces via a system, which is triggered when you change the state enum in a meta entity containing a state switch component.

To disable all entities of a particular "space" really depends on how you want to approach it. Though, it could be as easy as disabling all entities that have a particular state type component.

>I'm not really understanding what you said about input. I should have entities for input too?

I do, yes.

Although, they're unique static entities, and can be accessed from any context via a global singleton.

>Like, an "input" component with the key I'm listening for, and then have the CheckInputQueue system loop over the entities with input components

Sort of, and it's mostly for ease of use.

I have an entity per major category of inputs that can be separated into different entities (controller, kb, mouse).

I poll my inputs and update the component of the particular entity that I polled input for (keyboard input goes into the kb entity components), and store that inputs data in the related component.

After that I coalesce my inputs into a single entity.

You could skip a step and just combine all inputs into a single entity w/o separating into different entities, but, I prefer separating into different systems/entities as there's quite a few inputs per type of input mode (with different things like deadzones, interpolating inputs, etc).


6b153a  No.15451798

>>15451797

>MVC design pattern (model-view-container)

should be

>(model-view-controller)


be7812  No.15451820

File: 03608197b6b2d1b⋯.png (2.79 MB, 2108x975, 2108:975, 4472 - tagme.png)

File: edcde9986b22abd⋯.jpg (1.21 MB, 1028x768, 257:192, 4701 - tagme.jpg)

File: e561a941f498990⋯.png (3.52 MB, 2995x1059, 2995:1059, 5061 - tagme.png)

File: 8d4dfebcef10ab2⋯.png (2.99 MB, 1330x2174, 665:1087, 5063 - tagme.png)

>>15451013

>>15451233

Your buildings should have some side detail. Have some official Evangelion screencaps for reference


3cc89b  No.15451930

File: 7bc80a0da4c5cb1⋯.webm (3.45 MB, 540x524, 135:131, 2018-09-14 18-49-00.webm)

Bug or feature? You decide!


fdd87e  No.15452063


1ca5e7  No.15452086

>>15451930

Have hidden codes throughout your game and have it be accessed by on of those codes. Along with bobble head and low gravity for stupid fun.


993d02  No.15452121

>>15451820

Man that's some great art.


6b6340  No.15452125

>>15451820

>>15452063

I'll try my best but is like asking a mona lisa to a begginer.


fdd87e  No.15452157

File: c794ef604b6a2c3⋯.png (6.05 MB, 2829x1407, 943:469, building.png)

File: 22312e2dc595fbe⋯.png (984.76 KB, 2829x1407, 943:469, building2.png)

File: a6363b6624c1080⋯.png (1.78 MB, 1781x1013, 1781:1013, ClipboardImage.png)

>>15452125

You really want to re-draw the textures you're provided for free? Just make them repeatable, apply a filter and you're done. I went from pic 1 to pic 2 to pic 3 for a Ludum Dare in a few minutes.


6b6340  No.15452170

>>15452157

I just learn by making mistakes, usually I first find out what doesn't work.

I'll try again for the other buildings.


be7812  No.15452271

>>15452125

It's more that the sides of the building are barren of detail. Sky scrappers rarely have flat nothingness on any side of them.


e13445  No.15452688

Fucking hell, digital drawing on a mouse feels like total ass. I really feel like I will have to buy a drawing tablet because using a mouse is like I'm in kindergarten again trying to learn how to write. Does anyone here knows how to draw with a mouse? I'm genuinely curious how you do it.


a11c51  No.15452772

File: 7d793aeccec8ccd⋯.webm (12.31 MB, 1280x720, 16:9, Potato Knishes OFFICIAL-I….webm)

>>15446169

>potato knishes

THERE IS NO ESCAPE


be7812  No.15452776

>>15452688

If you have to draw with a mouse. Get something that supports vectors. You don't need a tablet to draw when you can manually control everything about your line work. Otherwise I'd recommend just ponying up the cash for a basic bitch tablet. It's what, $50 USD these days? Even the cheapest Wacom pays for itself if you use it regularly.


67d81d  No.15453237

>>15452688

Now if you're doing texture work, get a tablet, since it's basically drawing. If you're doing pixel art, I like a mouse more, since I manipulate shit with keyboard more ofter.


67d81d  No.15453618

File: 9210d8c41b99533⋯.png (861.9 KB, 808x627, 808:627, ClipboardImage.png)

Small progress: Attempted to do some clip rectangles, and messing with render states.


74fd78  No.15454576

>>15449888

Everything from scratch including models, animations, and math. Nada used from asset store. I wanna be the polar opposite of Yandev.


fdd87e  No.15454617

>>15454576

>I wanna be the polar opposite of Yandev

words to live by


159281  No.15454644

File: 40bea9cea779cf7⋯.mp4 (5.42 MB, 800x600, 4:3, puzzle_game.mp4)

Wait a minute, that flame..!

>>15451233

Try modeling the geometric details of the building (indent the windows, extrude pillars, vents, etc.), and texturing it after that. I think it will look better that way.


6b6340  No.15454971

File: 8ad9bb1d03831d1⋯.jpg (10.59 KB, 480x360, 4:3, hqdefault.jpg)

0-3 dev

4-9 learn to make music today


409d55  No.15455016

File: c31c708ccbf724b⋯.webm (1.26 MB, 1280x720, 16:9, RevengeOfTheTVBots.webm)

Haven't been back to this can in quite some time. Nearly done porting my game from Java/LibGDX to C++/OpenGL. The ranged enemy AI is ready to go and by the end of the weekend I should have melee AI too.


6b6340  No.15455063

File: f832048996a21c2⋯.png (589.4 KB, 1440x900, 8:5, 2018-09-15-094100_1440x900….png)

looks like some cheap doujin game, I love it.

;)


3cc89b  No.15455069

File: 7aab4a2d879fe59⋯.png (271.69 KB, 810x599, 810:599, disgusting and sad.png)

>>15455063

>being proud of mediocrity


6b6340  No.15455074

>>15455069

>he doesn't like doujin stuff

shit taste.


8e4a29  No.15455086

>>15455074

You can't possibly say "shit taste" to someone after posting that >>15455063


6b6340  No.15455097

File: 302cbba861db13c⋯.jpg (130.6 KB, 640x480, 4:3, mqhEaGz.jpg)

>>15455086

It looks like doujin shit bro.

doujin stuff is literally amateur as fuck.


000000  No.15455099

>>15449139

Did you rip peasants from Gat out of Hell?


ec2955  No.15455101

>>15455063

>>15455074

The buildings look too low-poly for the model, the model looks too high-poly for the buildings, the flat sea is a joke too, it all just gives me whiplash looking at it.


ec2955  No.15455109

>>15455097

I can point out a dozen problems with that at just a glance, but it's leaps above yours, mate.

And stop being proud of looking "amateur as fuck", that's something to be ashamed of.


3cc89b  No.15455118

YouTube embed. Click thumbnail to play.

>>15455097

Doujin works usually have plenty effort and love behind it

Yours scream laziness and arogance

This is doujin


6b6340  No.15455124

>>15455118

It just becomes really frustrating you're happy with your results and everyones just shit on it.


3cc89b  No.15455128

>>15455124

You are deluding yourself, deep inside you know its not good enough

Also if you come here you know you will get the harsh truth, if you want a hugbox to go reddit


8e4a29  No.15455130

>>15455124

You should expect it. At least those aren't personal attacks


6b6340  No.15455133

>>15455130

>>15455128

I already told you guys I'm really trying my best I can do.

>but is shit


8e4a29  No.15455139

>>15455133

In that case, keep trying.

You can both be happy with your results while knowing you can do better or will do better as you get better


6b6340  No.15455147

>>15455139

https://www.youtube.com/watch?v=aAiWX9Fi64E

this was my best effort like four years ago.


ec2955  No.15455155

>>15455139

This is quite important.

Though most times, the realization that I can do better sinks in when I do better. I'll write a good song, like it, then write a far better song and think quite low of the other one, going back to adjust it and so on.

Speaking of which, I really need to go back and adjust my models. It's not as good as it should be, though I'm not sure how to adjust it exactly. Maybe later.


4cc6df  No.15455157

>>15455133

Do you expect a participation prize?

If it's shit, IMPROVE IT.

Enough of your "but this is the best I can do, pls pity me :(" bullshit, you should always be striving to get better. These threads are based around people building their own skills and helping others to improve as well, are you really surprised people are calling you out when you come in with that level of work and make no attempt at bettering yourself despite receiving actual criticisms and advice?


8e4a29  No.15455161

YouTube embed. Click thumbnail to play.

>>15455147

Hey, I remember that bowl.


6b6340  No.15455166

>>15455157

It just becomes really tiring to hear people shitting on something you made everytime.

yeah, I know I'm shit.


3cc89b  No.15455171

File: 3c2268619f88f7b⋯.png (76.7 KB, 605x824, 605:824, shirt.png)

File: 2fb0dddb3ed2965⋯.png (129.2 KB, 700x575, 28:23, red reference.png)

>>15455133

This was my best 1 year ago, it was almost a copy of Red from pokemon sun&moon, the topology was shit and it had several flaws in the joints, but at the time i didn't know how to improve, But I remade it over and over again.

If you look at my model today it has none of the flaws it had and it even looks older, as if it grew up.


ec2955  No.15455172

>>15455166

If you're tired of being told you're shit, improve.

What do you expect, people to lie because you're trying super duper hard? Go to Reddit for that.


a0a47a  No.15455173

>>15455166

dont post it if you cant handle the obligatory bants


6b6340  No.15455177

>>15455173

>>15455172

>>15455171

I just posted some video of what I was able to make four years ago.

I've improved.

It just become tiring to hear: oh well, improve, when you had improved a lot.


ec2955  No.15455179

>>15455177

You know, anon, going from knowing nothing about coding to printing HELLO WORLD is also an improvement. Doesn't mean you deserve to be coddled for it.


8e4a29  No.15455181

>>15455177

>implying you can ever stop improving


6b6340  No.15455186

>>15455181

>>15455179

ok, I'll shut up.


fdd87e  No.15455465

>>15455016

wow, how long has it been, bro? Great to see you back


409d55  No.15455501

>>15455465

Over a year for sure, last time I was around I don't remember there being a general thread here on /v/, I also don't think there were demo days on this chan yet.

Game probably hasn't changed much since then though, most of the last year has been spent rewriting the engine in (very C-like) C++


ec2955  No.15455514

>>15455501

The threads have been around for a while, for sure, but Demo Dubs Days were started on 11/11 last year, so it's been around a year since inception for sure.


31957e  No.15455558

>>15455016

What is the game? I remember playing a demo of it.


dc1fba  No.15456219

File: 5b18657c83c96a4⋯.jpg (65.71 KB, 285x428, 285:428, disgusting.JPG)

File: 28e21fd18e76a07⋯.jpg (26.8 KB, 652x435, 652:435, landa.jpg)

>>15456153

>people unironically posting Extra Credits calling them educated

>programming is hard

>uwu

>blogposting everywhere

Words cannot express the disgust I have for the state of that site and its inhabitants. If you use that site kindly piss off and remain there.


dc1fba  No.15456231

File: 7c5d476216e60a7⋯.png (225.36 KB, 795x379, 795:379, 7c5d476216e60a7215edec43bd….png)

>>15456153

>even cuckchan tells him his art is shit


6b6340  No.15456242

File: b2c9d2346759824⋯.png (350.15 KB, 1216x384, 19:6, mountain.png)

>>15456153

>>15456219

>>15456231

I honestly have no clue what I am doing.

I finished watching a painting tutorial for mountains and I think it looks better than my 3D attempt.


dc1fba  No.15456256

>>15456242

It does look better but it looks like you just copied someone using a technique without understanding its purpose. Anyway, consider fucking off to cuckchan.


6b6340  No.15456274

File: 9e04177c7382d9c⋯.png (129.58 KB, 1440x900, 8:5, 2018-09-15-160124_1440x900….png)

>>15456256

Downscaled seems to look fine.


154ece  No.15456404

https://archive.fo/hbKRe

>I just want to let you know that, after discussion, we will be removing the term "slave" from future versions and replace it with puppet. Master/Puppet was the best that we could come up so far that does not make reference to a hierachy relationship between two human beings.

Godot, pls


67d81d  No.15456420

>>15456404

That's okay, it's open source so you can just rename them!


00e375  No.15456431

File: 76d4b06fc69c5cd⋯.jpg (32.92 KB, 460x428, 115:107, cac86df1583e7b097ed3316bf8….jpg)

>>15456404

>if we make this tiny change maybe the SJWs will stop bugging me

Poor Juan, he doesn't know what's coming.


00e375  No.15456462

File: 7b06c5635c948b3⋯.png (65.13 KB, 311x454, 311:454, Screenshot_2018-09-15_15-3….png)

File: 1518ea8e8d7fb6a⋯.png (135.26 KB, 750x592, 375:296, Screenshot_2018-09-15_15-3….png)

File: f7ddff51e60d03d⋯.png (40.18 KB, 607x276, 607:276, Screenshot_2018-09-15_15-3….png)

>>15456404

>look up complainer

>Steven Universe shitter

>they/them

>crossdresser

>Rust contributor


f1485f  No.15456476

>>15456404

Godot no!


27fc5d  No.15456479

File: 3858da2f6da9d24⋯.png (242.06 KB, 435x470, 87:94, 3858da2f6da9d24694bc90b02b….png)

>>15456462

Is this satire? Please tell me it's satire.


3cc89b  No.15456484

>>15456462

For once I am glad I went the proprietary slave route


27fc5d  No.15456529

>>15456484

>not going the enginedev route

Free from all shackles except your own sanity


ec2955  No.15456536

>>15456404

And this is why you enginedev. You can't jew yourself.


00e375  No.15456541

File: 53c5a62e7720f0d⋯.png (211.71 KB, 594x1232, 27:56, Screenshot_2018-09-15_15-4….png)

>>15456479

If you check the tweet he's responding to (https://archive.is/Rd6lm), it's several months old and he initially argued against it. I guess the guy got fed up with their bitching and made the mistake of giving in.

This won't end well.

>>15456484

>I may be an shit-covered communist transgender potatofucker but at least I call my sissy faggot a slave


ec2955  No.15456555

>>15456541

How long until we can't use the word banana, as you can say that someone is going bananas, and that's trivializing mental illness? Or we have to slaughter all pigs and cows, as those animals are sometimes used to make fun of fat people?


67d81d  No.15456569

>>15456555

You can't slaughter cows, they're sacred to Hindus! Fucking bigot


993d02  No.15456573

>>15456555

Going bananas is offensive because it references the nature of primates, and a black person of color is often discriminated as a primate. :^)


ec2955  No.15456585

>>15456569

Did you just call me a BIGot? How dare you body-shame me?!

>>15456573

PriMATE? Are you implying that only humans capable of reproducing are normal?! I cannot even begin to tell you how offensive that is!


be7812  No.15456608

>>15456420

>Update your project to latest version

>Bunch of shit breaks because of this pointless word change

>It's open source so you just change it back

>Everything's working again

What a way to waste everyone's time


993d02  No.15456630

>>15456608

It shouldn't normally work like that, but C++ is a fucking dumb language held together with Pajeet paste and the souls of corporate technicians.


74fd78  No.15456665

>>15456404

Shame. It happened to python a few days ago as well. Python has been dying and rotting structurally from the inside for a long while. Guido van Rossum, the original creator and maintainer of Python jumped ship a couple months ago. Basically python is ripe for taking over and is being usurped by crossdressing homos.


3cc89b  No.15456666

>>15456529

>>15456541

How did you two manage to interpret the exact opposite of what I've said?


00e375  No.15456683

File: 530c664acf92a84⋯.png (1.1 MB, 1280x850, 128:85, 5d57d574ea98ee8c0c69f8a5ca….png)

>>15456666

>Super Satan misreads two posts and accuses them of misreading his own


7f7464  No.15456711

File: 04e526d10f1e476⋯.jpg (119.46 KB, 1280x820, 64:41, photo_2018-09-08_15-57-46.jpg)

>>15456404

>>15456462

>tfw your own engine will never have contributors besides yourself

>tfw you will never have to deal with any of these people

Open Source "culture" was a mistake. If you want to make something, you should just make it yourself. No engagement, no contributors, no community. The next person to work on my engine will hopefully just be another 1MA programmer forking it to make his own improvements.


46df19  No.15456788

File: 753da9a1f22d69e⋯.jpg (27.68 KB, 540x438, 90:73, 15c13284325a4fb028a0f607d9….jpg)

>>15445185

>tfw you have cool ideas, but you have no time,talent or effort

at least you guys are having fun tho :^)


27fc5d  No.15456803

File: 1e9db50b8c7acd1⋯.jpg (32.58 KB, 400x399, 400:399, 193.jpg)

>>15456711

Okay, listen for a second, this is going to sound retarded, but instead of caring so much about "open source culture", maybe we should encourage for "software DIY culture" or something.

Instead of throwing open source software at other people, you provide resources and tutorials and sample code to teach how to make things yourself. Even if your program is proprietary or closed source, you can have detailed explanations and tutorials on how you made things. Eventually there will be detailed educational material for making almost anything.


9c5eea  No.15456833

>>15456803

Culture? I don't need to no culture, you idiot. You piece of dirt. Who are you? I step on you. I clean you off of my shoes at night. I step on you, and then I throw away the shoe. That shoe that I just threw away, that's worth more than your worthless life, mister. And I am pissed because I loved that shoe. You dirt. You piece of subhuman shit.


00e375  No.15456869

>>15456711

You can still use the cathedral development model and keep your source code open. Not releasing source code at all is very shortsighted and makes reviving old projects or porting them to new platforms a bitch, all because some autist didn't want anyone to see or "steal" his super special code.

>>15456803

There's already loads of that stuff floating around the internet, people just have to get their asses off goybook and look for it.


27fc5d  No.15456911

>>15456869

>There's already loads of that stuff floating around the internet

Only because it's accumulated over a long time, and it's still lacking in many areas. If "showing how you did it" was as popular as "sharing what you did", it would be a whole different picture.


96c013  No.15456918

File: 29d95778cfbcfb6⋯.png (117.15 KB, 1282x863, 1282:863, fagtard talk.png)


67d81d  No.15456921

>>15456911

So you're saying the solution is to show others how to do something. To "teach" them, perhaps. If only there was an organized institution in which prospective students could learn from "teachers"


f40276  No.15456930

>>15454617

so never, just like make game?


27fc5d  No.15456933

File: a50d210dabb7bd3⋯.jpg (42.71 KB, 601x454, 601:454, 1648952.jpg)

>>15456921

Have fun learning about how racist programming is in (((school)))


6b6340  No.15456967

File: 3a265e727f798aa⋯.jpg (479.33 KB, 2560x1920, 4:3, IMG_20180915_180452.jpg)

I now can shitpost and dev at the same time.

:)


67d81d  No.15456979

>>15456933

Already did my time 6 years ago, loan paid off, I just need to make game


7f7464  No.15457068

>>15456869

I never mentioned what license I was using, just that nobody can contribute to my project but me. I will make the code available when I have finished the engine to an acceptable state that a substantial demo can be made.

There is a huge amount of gray area between not releasing the source code and having "open source". When you present it like a binary choice, you're making it look like Open Source is the only way to share what you did, which isn't true at all.

>>15456803

That would be more useful to have, even if you would end up duplicating work. It would be better to both release the code and try and teach people how you made it.


ec2955  No.15457118

>>15456711

I would take that as an insult, personally. If someone else needs to improve my engine, that means that I'm a failure as a programmer.


7f7464  No.15457161

>>15457118

You're never going to make something perfect. Even if you did, that doesn't mean that it won't need improvements later.

The Doom engine had to be improved so that it could work on other platforms than DOS and NeXTSTEP. That isn't the fault of John Carmack, because when he made the Doom engine, he didn't have access to those platforms. Most of them didn't even exist yet. But that doesn't make a failure.


40695c  No.15457215

>>15456967

Fuck off


67d81d  No.15457241

File: 6611469ebf001a2⋯.png (40.92 KB, 474x450, 79:75, ClipboardImage.png)

Anyways, back on topic, I'm having a bit of troubles making my SpriteBatch class work the way I want it to. The client code should just throw shit at it and have it work, they shouldn't need to worry about the underlying vertex buffer - but what happens when it fills up? Right now I just store the render target and draw to it when I'm out of buffer space, then reset it.

The whole point of this class is to accept a ton of draw calls, then sort them and display the stuff to the screen. The sorting part is tricky, because the vertices need to be arranged into quads and sorted appropriately, but I when I mess with it too much, performance tanks. My draw call looks like

>SpriteBatch.Add(Vector2 position, FloatRect source, Texture texture, Color color, Transform transform, float depth) { … }

Instead of passing all that data and storing discontinuous stuff and making it a nightmare to sort, I was thinking of just having a separate Sprite class (which name conflicts with SFML's but I wn't be using it), which wraps everything up nicely, pic related.

One downside to this is that I can't take advantage of sprite sorting AND reuse a sprite. This means that I can have a number of entities with an associated sprite, but tracking up to 128x128 tiles with their own objects that take 60 bytes each? No way


00e375  No.15457290

>>15457068

>I never mentioned what license I was using, just that nobody can contribute to my project but me. I will make the code available when I have finished the engine to an acceptable state that a substantial demo can be made.

>There is a huge amount of gray area between not releasing the source code and having "open source".

That's the Cathedral development model and it's still open source even if you don't accept others' contributions. It was used by prominent freetard projects like Emacs, GCC, and Cube 2 Sauerbraten back in the day, but nowadays most freetard stuff leans towards the Bazaar model.


7f7464  No.15457419

>>15457290

I was just talking about how you present the software release as a binary between not releasing source code and making it open source, and how that's not a real choice you have to make. I know about the cathedral and the bazaar.


67d81d  No.15457424

File: dfcd20d06eb2240⋯.png (30.52 KB, 1348x982, 674:491, ClipboardImage.png)

>>15457241

Also I'm not sure if I should scale just the background to the window size or what. Right now I'm working on a rough UI system to get stuff displayed, but I have the lowest supported resolution at 800x600, but if I make the window big I can either scale everything or not.


993d02  No.15458170

File: 14a5949a3993938⋯.mp4 (2.61 MB, 1920x1080, 16:9, scene_editor.mp4)

Have you guys seen Xenko? It's an open source (MIT) C# engine. Honestly, its features look way better than Godot's, shame it only has so few people in its community.

https://xenko.com/


6f4958  No.15458299

File: 5bd497a222b5a24⋯.webm (8.98 MB, 480x360, 4:3, angry shoe man.webm)

>>15456833

Are you the guy from this video ?

>>15456921

If only said institution would offer anything of value that you cant find online for free and without marxism then your idea may have worked.

>>15456967

You should REALLY consider scrubbing the metadata off images that you post online you Columbian shitposter.


993d02  No.15458308

>>15458299

You think that's bad? He's some jungle spic named Cristian.


409d55  No.15458427

File: 29c42bc664e2815⋯.webm (1.02 MB, 1280x720, 16:9, DO-DO-DO-DO-DO-DO-DO-DO.webm)

Ever get that feeling like you've just blundered into the wrong neighborhood?

>>15455558

The name is Ctesiphon, and its intended to be a 90s-style single-player shooter. I've participated in a few of the demo days back on the other chan.


f1485f  No.15458533

>>15458427

You said yoy made your own engine? Lighting looks pretty good.


409d55  No.15458560

>>15458533

Yeah, it was previously done in Java with LibGDX, but I kept getting frustrated by awkward OOPisms and eventually decided to redo the whole thing after hitting a really annoying garbage collection issue.

The lighting on the world is just a quick bake from Blender internal, with dynamic lighting on entities using the same algorithm to match.


409d55  No.15458963

Having been out of the loop for so long, can anyone tell me what happened to Tetrachrome? The dev is saying he's going to delete his tumblr page.


a0a47a  No.15458971

>>15458963

yokotaro made his game and tetrachromedev went apeshit


409d55  No.15458978

>>15458971

Shame. There's plenty of space for more booty-simulator action games in the market. Hell, he had a completely different art-style too.

I suppose all of this is very old news by now though.


567b8c  No.15459004

>>15458978

Yeah fag where’ve you been?


409d55  No.15459018

>>15459004

Like I said above, I ditched my entire codebase and started over in C++. At the same time the progress general over on the agdg board was so slow it wasn't worth lurking in so I just went back to 4daggy instead.

Now that I actually have more visible stuff to post, I'm back.


f1485f  No.15459035

File: c93b020c9078708⋯.png (1.16 MB, 1200x1455, 80:97, 1510861591899.png)

>>15458963

I feel bad, but it always gets a smile on me.


caf381  No.15459070

Does anyone remember the name of the dev/project that came here about a year ago and got shit all over? He was from CDProjeckt Red and had some anime girl tank washing simulator or something he was working on. Curious to see whether it went anywhere or if it's just vaporware.


1692fb  No.15459102

>>15458963

>>15458971

>>15459035

>imagine being so deranged and paranoid that when a game that happens to be tertiarily similar to yours is made you think you got ripped off.


1ece94  No.15459112

>>15459102

Even if he didn't get ripped off it's gotta be crushing knowing that someone beat you to the punch at your own idea and everyone would assume your game was the ripoff if you finished it. And Yoko Taro did post here, you never know.


f1485f  No.15459117

>>15459102

I was terrified when Elite Dangerous and Star Citizen were announced. Later I realized no one was going to make the game I wanted to, but the dismay was real.


620e18  No.15459166

>>15459035

He did a better job than a tumblr tranny would of, it should be grateful


fdd87e  No.15459170

File: b4bf400fd96eb01⋯.png (1.41 MB, 1199x760, 1199:760, ClipboardImage.png)

File: ad130b5894fb278⋯.mp4 (918.02 KB, 600x386, 300:193, DnK4opwWsAAK0K9.mp4)

File: 1b46bad9a17a440⋯.mp4 (803.64 KB, 700x462, 50:33, Df0cXtOUwAEEPrt.mp4)

File: e9b288bd8258e43⋯.mp4 (2.74 MB, 1280x720, 16:9, Rn5_tGbo7XwdJc8h.mp4)

>>15458963

>what happened to Tetrachrome

lives on in our hearts.

>>15459070

they are still making progress and it looks good, though primarily it's animations and no tank combat but clearly got driven away from here, they were just shilling, not taking part in any discussion.

https://twitter.com/TeamFalldog


004396  No.15459190

>>15459170

Any gameplay at all so far?


ece486  No.15459288

>>15459102

>tertiarily

>terti

You mean tetra.


be7812  No.15459452

File: ad3416f7d2f3368⋯.png (115.43 KB, 1232x822, 616:411, BreastPhysicsExplaination.png)

>>15459170

Was clearly just shilling and working at CD Projekt Red means he's no amateur. If the timeframe of when he said he worked there is correct he was involved with both Witcher 2 and 3. Meaning he has no place on a development thread meant for people who are working on their first few games with nothing to support development except their dreams and their free time.

>>15459112

That discussion's already been had. If you line up the time Yoko Taro visited here with how long they said they developed the game, unless they were horribly behind schedule, 2B as a design would have existed many months before his visit. The contract with Platinum would have already been signed and finalized. There's no reason to suspect that Yoko Taro stole the idea of some polish amateur developer.

There's also nothing to indicate he used this website beyond posting in his own thread. He hasn't mentioned it since he had that thread or anything. I do not think he's a regularl enough user to even know we have a game development thread.


6c0a06  No.15459470

>>15459070

Did they get shit all over? I remember that dev posting for a while but I don't remember anyone shitting on him. What happened?


27fc5d  No.15459476

Is there a good tutorial for handling windows in Linux? I'm basically looking for the equivalent of Win32 api, no OpenGL.


be7812  No.15459479

>>15459470

See >>15459452 and >>15459170

He wasn't contributing to any development discussion and wasn't an amateur. He just used these threads to plug his game and cross-posted on reddit/cuckchan/twitter at the same time. He was just using these threads as a marketing front.


bfc546  No.15459486

>>15459476

XWindow?

You could look at the GLFW Linux source code. It's MIT license.


455635  No.15459500

>>15459476

X11, but it's rather low-level and you'll have to deal with extensions if you want to support multimonitor setups and the like. Depending on your needs, a GUI framework like GTK might be the better choice.


27fc5d  No.15459510

>>15459486

Reverse engineering source code is several magnitudes harder than reading a tutorial or comprehensive and well put together documentation.

>>15459500

>GTK

>toolkit for creating graphical user interfaces

Sounds like bloat.

Looking up X11 gives promising results at first glance.


409d55  No.15459557

>>15459470

>>15459510

Using X11 directly isn't that hard, just remember to use the Xinerama extension for proper multi-monitor support like >>15459500 said. Without it X11 treats your entire combined desktop as a single "screen".

See Shadow Warrior Classic Redux for an example of doing this wrong. It's fixable if you can get to the options menu or change the config files, but on a two-screen setup it fullscreens on a single monitor but acts as double width in-engine.


409d55  No.15459586

File: e9b8b6abeb4737f⋯.webm (1.38 MB, 1280x720, 16:9, PortComplete.webm)

Now this is podracing! With the charger dudes getting their attack animations back, the C++ rewrite is now fully complete.


00e375  No.15459590

>>15459476

Most Unix vidya devs use SDL for a reason, anon. You can use raw xlib/xcb but it's a mess.


3cc89b  No.15459816

>>15459170

Swimsuits and tentacles are a very dangerous combination anon


3cc89b  No.15459819

>>15459586

Very nice


6b6340  No.15459960

File: 3678b60c0996cc9⋯.png (1.08 MB, 2464x900, 616:225, 2018-09-16-130120_2464x900….png)

second attempt at doing more mountains.


a0a47a  No.15459999

File: 25ace5728f83f48⋯.png (170.18 KB, 400x400, 1:1, 25ace5728f83f48834666613d0….png)

>>15459960

stop being such a fuken weeb


b83677  No.15460007

>>15459999

NUMBERS


6b6340  No.15460009

File: 46a4c9b1136e5e4⋯.png (83.21 KB, 697x148, 697:148, mountain.png)


ec2955  No.15460011

File: d93f10816e03d4d⋯.gif (1.97 MB, 395x192, 395:192, d93f10816e03d4d64b78b91cbd….gif)


428cc1  No.15460012

>>15459018

> I just went back to 4daggy instead.

Stay there


00e375  No.15460032

YouTube embed. Click thumbnail to play.

>>15455063

>>15459960

>>15459999

It looks worse than Evangelion's N64 game and uses loads more resources. Leave it to third world shitters to make crappy "low poly" games in heavier engines designed for PBR instead of using OpenGL 1.x engines for toaster friendliness.


993d02  No.15460046

File: da49482d0932d42⋯.jpg (149.25 KB, 771x960, 257:320, da49482d0932d42288fa72aee0….jpg)

>>15460032

Damn that game looks rad.

Wasn't there a PS2 game as well?


6b6340  No.15460057

>>15460032

godot is toaster friendly though.


993d02  No.15460066

>>15460057

Yeah. Godot is coming out with an opengl2 renderer for shitphones.


00e375  No.15460070

>>15460057

Not toaster friendly enough for what you're making. Even when OpenGL ES 2 support returns in Godot 3.1 it still has an awful lot of overhead for something you could do easily in the id Tech engines or something like PLIB/Irrlicht. This is almost as bad as Dustdev using Unity for his meme lowpoly shooter with higher system requirements than fucking Crysis.


993d02  No.15460074

>>15460070

Godot has more documentation and more people using it though.

Working with Irrlicht is a pain because you're thrown into 2005 xbox360 dev territory.


7f7464  No.15460078

>>15458560

how does it do collision detection and culling?

>>15460057

Godot requires GLES 3.0. None of those are toaster friendly. OpenGL 2 runs on some toasters but they still struggle to render stuff that uses the full featureset. You'd have to write a renderer for OpenGL 1.1, to be compatible with all the toasters, and it still isn't going to be fast as just using your own engine to do this.

>>15460070

Even id Tech 3 is more than he needs. It runs a VM for the gameplay code. You can just write your own engine with lower system requirements.


6b6340  No.15460088

>>15460078

you can get a GLES 3.0 GPU for like 20 bucks.

unless you're doing retro console dev, there's not point of using Opengl 2.1 unless you plan on giving support to thirld world pc.

Open gl 2.1 was made in 2006.

Open GL 3.1 is from 2010.


154ece  No.15460093

>>15460074

>Godot

>documentation


00e375  No.15460113

>>15460088

Yes anon, but it's the principle of the thing. Making crappy low poly games which don't run on authentic retro hardware is an insult to the devs who spent loads of effort making their vidya run decently on old PCs and consoles while still looking nice. People also make desktop programs like (((Discord))) bundled inside fucking web browsers and use the same "it's CURRENT YEAR+3, we don't need to worry about optimization or performance anymore" excuse. This approach is everything wrong with modern programmers and also why everything is so fucking slow despite our hardware.


6b6340  No.15460124

>>15460113

maybe I'll enginedev later when I finish this game so I can make a kof clone in SDL, sounds fun.


993d02  No.15460127

>>15460093

Maybe you'd be right back in 2.1, but even then it still had more than Irrlict/Ogre3D.


be072a  No.15460130

File: 28ce85a4e960067⋯.webm (1.98 MB, 720x404, 180:101, accidentalbullethell.webm)

Made a character that can deflect back projectiles.

yes screen shaking is still shit

yes all assets are recolored cc0 sprites from opengameart


993d02  No.15460136

File: 9e8fba8e7070f46⋯.png (30.67 KB, 281x530, 281:530, 9e8fba8e7070f46f8a8db4603d….png)

>>15460113

You're expecting high expectations from people who are one man game devs.

Why make a game when you spend two years making a shitty tech demo for your shitty engine- reinventing the wheel each time? Who fucking needs that?

What do you want people to make? Actual video games or shitty tech demos? If its the latter, head on over to >>>/tech/


7f7464  No.15460137

>>15460088

A GLES 3.0 GPU for 20 bucks is going to comply with the spec but isn't going to run anything using those features at a reasonable speed. The date something came out doesn't mean very much either. OpenGL 2.1 came out in 2006 but intel integrated GPU's only supported OpenGL 2.0 at least until 2008 or something.

Either way the point is that if it CAN run on those old systems, it should be able to run on them. There's no excuse. You aren't doing anything that needs all this new hardware.


6b6340  No.15460138

>>15460113

for some reason there's pixel bleeding on my mountains but godot devs mention there shouldn't be and that it will be fixed in 3.0


993d02  No.15460143

>>15460138

3.0 is already out though. It's even in stable branch.

Unless they mean 3.1, they're probably wrong.


6b6340  No.15460146

>>15460137

but most people even in my thirld world nation has pcs that are newer than 2008.

2008 pcs are basically e-waste tier at this point.

I'll try to enginedev maybe in the future for fun.


6b6340  No.15460148

File: 6b7b701c7c489bc⋯.png (358 KB, 1440x876, 120:73, godot.png)

>>15460143

for some reason there's pixel bleeding even if I have the fix option selected.


6f4958  No.15460154

>>15460113

Thats a great way to waste a ton of time just so 2 users can play your game on their dusty 486.

You have to find a method viable from both a performance and a work-efficiency perspective.


993d02  No.15460159

>>15460148

Try compiling the engine source straight from the master branch. If it still persists ask them why changing slave to puppet is so important when textures don't even fucking work.


6b6340  No.15460168

>>15460159

maybe I should enginedev as well.


00e375  No.15460173

>>15460136

I'm not pushing enginedev right now, I'm just fed up with lowpoly shitters jumping for the latest popular engine instead of looking around to see if something older might fit their project better.

>>15460154

>install id Tech fork

>just make game using an existing engine

>wow I wasted so much time making a game everyone can play


f1485f  No.15460179

>>15460070

Which game is Dustdev's?

>>15460159

>changing slave to puppet

Is there nothing to look up to in this world?


993d02  No.15460180

File: 7534360dd6987a0⋯.jpg (66.34 KB, 504x459, 56:51, 7534360dd6987a03a9da1703a9….jpg)

>>15460168

Do it.

It's critical to make sure you have a great grasp on programming before you do anything else like art.

Start with C and SDL and work up from there.


00e375  No.15460191

Invidious embed. Click thumbnail to play.

>>15460179

vid related


00e375  No.15460193

>>15460179

>>15460191

It's called Dusk, I'm retarded


6b6340  No.15460200

>>15460180

I guess I should learn CS.

I am tired of not being able to solve my issues with game engines.


f1485f  No.15460202

>>15460193

>>15460191

Oh right. They're not agdg though, right?


e1aa57  No.15460210

>>15460078

>You'd have to write a renderer for OpenGL 1.1, to be compatible with all the toasters

That's exactly why I abandoned Godot, that and the performance, my pc only supports 1.1 due the shitty integrated graphic card and intel being shit at making drivers, on linux it should support 2, even with that I managed to run godot using software openGl, for 2d it was bearable on a low resolution screen.

I really miss their animation editor, but the ship has sailed a long time ago.


00e375  No.15460215

File: 2ebeef5af93d2c6⋯.png (345.61 KB, 583x641, 583:641, Screenshot from 2017-12-27….png)

>>15460202

Nope, just some bearded numale who doesn't get classic FPS level design.


6f4958  No.15460216

>>15460173

The time spent on learning idTech is wasted as soon as you move on to another project unless all you want to do is retro-lowpoly stuff.

Experience in Godot or whatever else > Having your first experiments run on old hardware.


993d02  No.15460219

File: 8c7f056fccbbc79⋯.jpg (32.63 KB, 379x499, 379:499, 41gHB8KelXL._SX377_BO1,204….jpg)

>>15460200

Read the K&R book first if you want to learn C. That's how I did it.

After that check the /agdg/ wiki for a more books, you'd probably want something like OpenGL or SDL to write with to get graphics.


f1485f  No.15460231

So is Godot data driven ECS or shitty iterated OOP?


7f7464  No.15460236

>>15460146

>>15460154

Even if you have a PC newer than 2008 it still might not support OpenGL 2.1. When something is released it is going to be produced for a few years and then phased out.

Really the point is just that better software will run faster and on more computers.

>>15460210

What is your GL_VERSION and GL_EXTENSIONS strings? Even if something says it only supports 1.1, there are still probably a lot of features that you still have and can use with extensions. I have my engine set up so that if needed it can fall back to a fully 1.1 compatible renderer, and it just queries extensions for everything else.


993d02  No.15460240

>>15460231

They have their own scripting language called gdscript which is basically python.

They also support C#


6b153a  No.15460247

>>15460240

They're asking about the design architecture


f1485f  No.15460258

>>15460247

>They're

Serious question, though. How well does Godot scale with enemy numbers?


993d02  No.15460259

File: cf3f2cb9dfa6d16⋯.jpg (511.99 KB, 1024x904, 128:113, 436909ac514772ac86c70d8df8….jpg)

>>15460200

Another reason to get into enginedev is Vulkan. Vulkan is basically the biggest shit right now that's making OpenGL deprecated. All versions are cross compatible and is optimized as fuck.

Take a glance at this. It's honestly ten times better than OpenGL and Godot won't be getting a Vulkan renderer in a long ass time.

https://vulkan-tutorial.com/


00e375  No.15460265

>>15460231

Definitely OOP, but at least GDScript doesn't support multiple inheritance. Godot is designed around its scene system and making heavy use of instancing.

>>15460259

>Godot won't be getting a Vulkan renderer in a long ass time

They're adding a Vulkan renderer in 3.2, but if 3.1 and its GLES2 renderer is any indication that could take months.


993d02  No.15460271

>>15460265

It took an entire year for 3.0 to come out and my guess is that 3.2 will come when Anno makes the fourth movie.


f1485f  No.15460275

>>15460265

First it was the slave/puppet cowardice, then it was lack of vulkan (which I thought it had), and now it's not even ECS. This day seems determined to sour me towards Godot.


00e375  No.15460306

>>15460275

They initially went for GLES3 over Vulkan because MoltenVK wasn't open source yet, but now that it's FOSS and the retards at Apple are dropping OpenGL support they're making Vulkan a priority for 3.2.

>Why support Apple at all?

Mostly mobilefags.

>not even ECS

The scene instancing system is still decent and fits some games nicely. I'd recommend it over Unity and Game Maker anyday for non-pixelshit 2D vidya Love2D and Orx are debatably better depending on the game and your preferences and it would be nice for 3D vidya if occlusion culling wasn't delayed to 3.2.


f1485f  No.15460381

>>15460306

But what about 300+ 3d fighter dogfights?


e1aa57  No.15460407

File: d255cbf0af26a6b⋯.png (21.64 KB, 1120x385, 32:11, 0916212014.png)

>>15460236

Oh I actually have 1.4, keep forgetting it, long time ago I tried to use SFML then dropped it because I couldn't use render textures, but now they're usable even if they're employing some slow hack to make it work for old machines, shaders are obviously out of question, is your engine 2D or 3D? how do you deal with unsupported shaders?


6f779d  No.15460446

>>15455501

>>15459018

>Over a year for sure, last time I was around I don't remember there being a general thread here on /v/

>At the same time the progress general over on the agdg board was so slow it wasn't worth lurking in so I just went back to 4daggy instead

Except these threads have been right here from the very start, and have always been faster than the /agdg/ board, which pretty much only exists to post updates about your own project. The only thing that's relatively "new" is the threads on /vg/, although I personally never look there and don't know if they're very active.

Ever going back to 4cuck for any reason means you're a faggot, don't do it ever again.


6b153a  No.15460490

>>15460275

If u want an engine that have a native ECS API look at Unity (the goal is "performance by default" (compiling to native code, and optimized per CPU/etc), and to be able to multi-thread everything possible with the ECS job system in addition to other neat stuff).

They've been developing the ECS API for awhile now, but as of yet I don't think it's production ready; it's also sort of a mess in terms of usability (thread dedicated to improving API/usability on the unity ECS sub forum, so it's being worked on).

It's being headed by that infamous DoD C++ talk guy who has trouble breathing during his talk (forgot his name too lazy).

Though in terms of politics Unity really isn't any better as they're slowly getting invaded by SJWs.


f1485f  No.15460498

>>15460490

I really wanted to go for FOSS. Shit.


b6f799  No.15460507

>>15460130

I don't know why NPCs like screen shaking.


ec2955  No.15460508

File: 8ce13bd2b1f31d2⋯.png (5.83 KB, 1830x102, 305:17, ClipboardImage.png)

Is this how Skynet begins?


455635  No.15460517

>>15460507

It gives explosions a satisfying impact. From a gameplay standpoint it's kinda annoying though.


7f7464  No.15460523

File: 510030fd24712e4⋯.png (580.14 KB, 640x480, 4:3, sigma2_2018-09-16_16-04-10.png)

>>15460407

My engine is 3D. I don't use shaders in OpenGL yet just because there is not any effects that I need to do with shaders at this time. Eventually I will have to use a vertex shader for dynamic vertex lighting, and a fragment shader, and stuff like that, but not yet. Right now I am rewriting my Vulkan renderer, it requires shaders but the ones I am using are pretty basic right now.

By the way, for you shaders are definitely not out of the question, your card supports a vertex and fragment shader, see the GL_ARB_vertex_program and GL_ARB_fragment program in your extensions string. It wont be like using GLSL but it is something you can work with.


905529  No.15460526

>>15447761

WHAT IS THE STATUS OF THIS GUY??? WHERE IS HE!


905529  No.15460541

>>15459170

Litterally what I want.


6b6340  No.15460560

File: f6d2459aa0b40dd⋯.jpg (7.22 KB, 200x169, 200:169, 1521529127966.jpg)

I'm still not sure if wasting years making an engine is the best idea.


00e375  No.15460574

>>15460381

>300+ 3D fighter dogfights

I haven't tested anything on that scale, but with clever use of MultiMeshInstance and C/C++ in performance-critical sections it should be possible. You'd get better performance with a specialised engine or one built around DoD, but for the latter you have few options aside from Unity's new ECS job system and obscure stuff like https://github.com/dbartolini/crown

>>15460560

In many cases, it isn't. Enginedev makes the most sense when you have very specific needs and know how to solve them, or if you're making 2D vidya.


67d81d  No.15460578

>>15460560

>>15460560

It is. You learn so much, from how to handle different data types, optimization, and compatibility, etc. It literally makes you a better programmer.

Compare with say, Unityfags. I'm not bashing on anyone in here of course, but you get some libtard on Twitter who makes the modern equivalent of a babby's first gamedev in Gamemaker with it, and then they load up on soy and circlejerk like they've accomplished something. I mean I guess they have but it's crude compared to what genuine learning and effort will accomplish


993d02  No.15460579

File: 54528138b8e2de9⋯.png (245.17 KB, 1290x750, 43:25, Screenshot_2018-09-16_14-2….png)

>>15460560

If you want full control over your game then go for it. It was daunting at first, but in a few days I got a nice test area running with full keyboard and mouse controls.

There's nothing really anything bad that could come out with it. You'll have a better understanding of vector math and how graphics work on a computer.


67d81d  No.15460582

>>15460574

Why does 2D gamedev fall into enginedev territory? Are you >implying that 2D games are somehow more complex because they need a custom engine, and that they are better than 3D games since you have fewer limitations to explore gameplay with?


6b6340  No.15460585

>>15460579

>>15460578

>>15460574

How do I learn?

Do I simply try to build a project and learn on the go, googling stuff or do I try to learn first all the CS shit I need before?


993d02  No.15460591

>>15460585

Read some books and try some examples.

I personally learned through by look at what other people did in their source code, and when I had a question about what a function did or why X did thing, I went on an IRC or looked at the documentation.


7f7464  No.15460599

>>15460585

You should at least know the basics of programming. Then start learning how to draw graphics. It will all make a lot more sense after you can do that.

>>15460582

2D engines are pretty simple to make. So it's not that big a deal to just write your own.


993d02  No.15460601

YouTube embed. Click thumbnail to play.

>>15460585

A good selection of videos for basic 1.1 OpenGL are these.

Python is easy as fuck to learn and all the OpenGL functions are in C libraries so you'll barely get any performance issues.


6b6340  No.15460606

>>15460591

>>15460599

>>15460601

I don't mean programming, but topics like data structures, compilers, and all the math i need.


e1aa57  No.15460607

>>15460523

I never considered using extensions, but I may give it a try, good luck with your engine.


993d02  No.15460618

File: 2e26e73a2d9a348⋯.jpg (152.7 KB, 1080x1080, 1:1, 2e26e73a2d9a34834878a5aa96….jpg)

>>15460606

Just look on Google for resources for graphics programming.

Here's a good start that came up on the first search result.

https://medium.com/vrtigo/resources-for-beginning-graphics-programming-c0da724381bc

Remember, there's no set guide for this. A lot of you'll be doing is picking apart information that you need to suit your goals. Once you become better though, some of this stuff comes natural. Don't feel bad if day all consists of studying. Any form of progress is great.

Feel free to ask any questions to sigmadev. He knows a lot of shit.


3cc89b  No.15460630

File: 898461fdd291c33⋯.png (42.42 KB, 485x737, 485:737, hair elly.png)

File: be1ded8306430b6⋯.png (117.73 KB, 474x868, 237:434, new hair.png)

Added strands, so its not so solid, still looks like it is missing something, maybe another level of shading?


7f7464  No.15460644

>>15460607

I make sure that it's all optional, so it can fall back if it doesn't have the extension I am using. It works pretty well, and it is really the only way to use your hardware's full potential. You can get some really big speed boosts using some of those.


b1b7ea  No.15460843

>>15452776

Hmm, I gave vectors a look but the problem is that the art I want to make is going to be just as shitty to do on vectors as it will be with mouse+krita. I'm still not 100% set on the art style I'm going to use so I'm not going to discard vectors yet.

In the end I think I will buy a tablet before making the decision, but I'm a student on a third world country, so technology isn't cheap here. I will wait until the end of the year when my sister to come back from Canada, maybe she can find it cheap over there and then bring it back to me.

>>15460630

>still looks like it is missing something

Looks fine to me. Maybe if you had a reference of a hair style you wanted to steal recreate then it would be more obvious what's missing.


8ae8af  No.15460860

File: f0ec5b646875692⋯.gif (852.29 KB, 900x341, 900:341, 2018-09-10_19-47-54.gif)

>>15460526

What do you mean? I'm right here :)


905529  No.15460890

>>15460860

The name of the game.


953206  No.15460910

>>15460860

wow

at what point do you reach the eldrich god level in the game?

>>15460130

>filename

ya know, that might actually be a cool concept

>>15454644

wow, that silky smooth framerate


6b6340  No.15460916

I just realized there's a 2 year technician degree about programming that is online and is offered by my goverment main trade school.

Should I get it?


8ae8af  No.15460953

File: 98b743f0ed75f01⋯.png (47.2 KB, 834x688, 417:344, Fireworks_2018-09-16_18-46….png)

>>15460890

It's Aethyr.

>>15460910

I mean it's set in a world that's very much "touched" by eldritch gods. So the entire game is the eldritch god level. Different temples are spookier though. As for meeting an actual god, who knows?

Don't mind the shitty art


dc71a4  No.15461037

>>15460860

Apart from speebot and the waifu dude i've seen your game for the longest time in here, i seriously respect your dedication. Your game is probably the only "modern" indie pixelmeme i'll play once it's out.


be7812  No.15461067

YouTube embed. Click thumbnail to play.

>>15460046

Yes but it was terrible. Graphics aren't great and it was an "RPG" laden with QTEs. The only thing good about it is the huge amount of canon lore it has. A lot of what we understand about Eva today originates from this PS2 game.

>>15460843

Ask her if she can look into Huion tablets. Those are cheap and just as good as Wacom.


905529  No.15461112

>>15460916

Do you want it? Will it help you get a job you want? etc have you done research?


3cc89b  No.15461190

File: 43f3bfb272d1d77⋯.png (1.51 MB, 1500x965, 300:193, elise.png)

>>15460843

I do have a reference, but i can't recreate something at this level yet

>steal

what? no! what makes you think that? should i give her glasses?


6b6340  No.15461429

>>15461112

programming is a good safe job, I already know how to code and getting a degree will be nice.

I may later think about making some game like cave story.


00e375  No.15461535

File: 1f25a11161d5bc9⋯.jpg (660.17 KB, 1500x1080, 25:18, fortnite cancer.jpg)

File: 1ad1e81230d85ab⋯.jpg (487.8 KB, 1240x735, 248:147, fortnite cancer 2.jpg)

File: 2acc2d8098d7524⋯.png (330.05 KB, 1133x1016, 1133:1016, Screenshot_2018-09-16_18-4….png)

Want to make money? Why not take up SFM Fortnite porn as a side job and earn a sick $73 a month?


be7812  No.15461572

File: 4a4967feaf1fc7b⋯.png (157.49 KB, 345x345, 1:1, Disgust.png)

>>15461535

I'd rather go back to the convenience store


67d81d  No.15461588

You faggots probably saw this already but this is important because it related to the future of Linux and other large FOSS projects

Basically a bunch of people continue to cuck out in the tech sector and put inclusivity ahead of professionalism

>>15461279

>>15461535

>$73/month

>Misread as $73K per month

Phew


7072da  No.15461593

File: e3f7498d3e3f535⋯.jpg (99.38 KB, 448x537, 448:537, 1480573337610.jpg)

>>15461535

Aren't there people who make furry porn for like 30k a month


b6f799  No.15461596

>>15461588

>inclusivity ahead of professionalism

Not my problem. Every difficult problem has been solved, it's just a place for a bunch of college educated bastards to waste their lives without achieving anything actually difficult. and they paid way too much to do it


29a99f  No.15461609

>>15461593

>give up your dignity and make thousands of dollars

>make less but keep your dignity

Every fucking game that came out in the past was corrupted by furries.

For example Corruption of champions was taken over by furries so they decided to create trials in tainted space, they claimed it wouldn't have furries and what happens soon after, rather than fucking aliens you got a bunch of fur shit.


7072da  No.15461617

>>15461609

I don't know man, If I was guaranteed 400k a year from furry porn I'd probably just do it anonymously or with a fake name.


6f779d  No.15461629

>>15461535

>3D porn artist

>I want to get a new monitor so I can stop using my parents TV for my pics

That's a reason to not donate to him, that's hilarious.


6b6340  No.15461683

Serious question, should I do 3D porn?


67d81d  No.15461693

>>15461683

Look at this guy for example

http://twitchyanimation.tumblr.com/

>377 "coffees" purchased for him

>$350/mo on top of that

>Plus commissions

>All he does it reuse animations and models aren't even done by him


6b6340  No.15461714

File: e13b91c7e68bb71⋯.png (259.83 KB, 593x635, 593:635, 1521320229225.png)

>>15461693

I'm still not sure if money is worth my dignity.


409d55  No.15461718

>>15460078

>how does it do collision detection and culling?

I'm using Bullet 3D for detection and broadphase, as well as for handling debris objects like shell casings and broken glass. Bullet's pretty good but there have been just enough frustrations with using it that for my next project I'm going to strip it out and do the whole thing myself.

>>15460012

no u

>>15460446

Whether they were or weren't, I didn't remember going anywhere but the board.


be7812  No.15461760

File: fe62da03985c4a9⋯.png (2.02 MB, 1660x1660, 1:1, 1449010663126.png)

File: 00f1e6732683ffc⋯.jpg (1.03 MB, 1314x2557, 1314:2557, 1447530800134.jpg)

>>15461593

I don't know about that. But I know there's people who make roughly $1,500 a drawing for furry porn.


29a99f  No.15461811

>>15461760

how do you do tax information with this?

>(IRS) So how did you come by this money?

>I draw dragon and furry porn


67d81d  No.15461828

>>15461811

Generally the sites such as Paypal (if merchant account), Patreon, etc all keep tax T4's which are sent to you at the end of the fiscal year, usually around February or so


5fe2f5  No.15461955

Faggots&Deviants are a cash cow you can milk like goy citizens.

>It is truly incredible how much money you can earn off of npcs


be7812  No.15462328

>>15461811

You put in your taxes as an "Art Commission" and then they'll just take from it as typical income tax based on how much you made over the course of the year.

Source: I know a furfag artist who makes like $600-$800 a year in furshit commissions and claims he's not a furfag because 1/5 of his drawings are of normal humans.


409d55  No.15462472

File: 56406a4e599b1ab⋯.webm (1.56 MB, 1280x720, 16:9, BobbingPolish.webm)

Some more subtle progress today, I did the maths to rescale view-bobbing/footstep rate so it's faster when moving slower than running pace, and slower when moving faster. Having it be a linear relationship directly to movement speed was bugging the hell out of me for ages.


67d81d  No.15462477

>>15462328

It really depends on how much you consider yourself a furry or interact with the community. For example, if an artist said they weren't a furry, but strictly drew furries for money, that's reasonable, no?


b1b7ea  No.15462531

>>15461190

Well the two things that I see different are the shading and the complexity. Your hair only use a few colors to give the shading effect while Elise's hair has a much more smoother shading, I think this might be more of a engine-side of things, where you can't just bake this level of shading on the hair. Elise's hair is also much more complex with more visible chunks of hair and points, yours looks more like a striped cloth. But I think it's necessary to see it in-game with proper lighting to finish comparisons. Glasses make everything better,except my face****


f1485f  No.15462534

>>15461714

It's not. You'll hate yourself.

>>15462472

After playing Doom 1/2 two years ago I have an itch for FPS that no AAA can satisfy. Dusk and Amid Evil kinda did it, and I hope yours will too.


409d55  No.15462601

File: b2335db7090efc7⋯.png (1.67 MB, 1543x6892, 1543:6892, Doom.png)

>>15462534

>After playing Doom 1/2 two years ago I have an itch for FPS

Well the whole Doom->UT2004 era of shooters is my muse for this project. I'm particularly basing enemy design decisions off of what I've learned playing Doom WADs. Even games that supposedly take inspiration from that era tend to fall into that stupid locked rooms design philosophy from Painkiller and Serious Sam and end up ruining the experience.

If you haven't yet, look at pic related and the Cacowards for megawads and weapon packs. The Shadow Warrior weapon/enemy pack and Kontra Kommando's version of Aliens TC are also rock-solid. Even if no other game scratches the itch, there's enough WADs in the wild to last a lifetime of playthroughs.


c65cb4  No.15462827

File: 870c88d0eac03d0⋯.png (83.6 KB, 1366x724, 683:362, localVariables2.png)

File: 4f7597fecf4e68c⋯.png (79.25 KB, 1366x724, 683:362, localVariables1.png)

finally bothered to add local bool/float variables

can't believe i hadn't done this

was just setting stuff in scriptable objects until now since i could treat them as a public variable, but it becomes a hassle when i want to do the same stuff per npc


5d0a45  No.15463276

>>15462601

This is exactly what's wrong with those games. Hell, Painkiller has great weapons I find, but god damn those level/locked room designs.


8e4a29  No.15463307

File: 51aef33c019d1ab⋯.png (10.42 KB, 512x157, 512:157, uncloaking.png)

File: 6bfaebce4bac664⋯.png (157.5 KB, 1538x2048, 769:1024, slime.png)

>>15458963

>The dev is saying he's going to delete his tumblr page.

>not deleting everything out of the blue in a fit of despair

>giving your followers enough time to save and archive your works


7871a8  No.15463330

>>15457241

I think you're going about this the wrong way. Think in terms of the data you're working with and how to manage it. Here's what I'd do:

A sprite is 4 vertices and a texture. We can store this either as 6 vertices drawn using glDrawArrays(GL_TRIANGLES) or as 4 vertices and 6 indices drawn using glDrawElements(GL_TRIANGLES), or maybe even as 1 vertex draw as a point primitive. GL_POINTS can have size > 1, set using for instance a vertex attribute, which will generate fragments in the shape of a square, equivalent to one built from 4 vertices. You can get the position inside the square for doing texture mapping in the fragment shader. Each vertex needs to have it's world space position as an attribute, as well as UV coords for texture mapping. These should just be (0,0), (0,1), (1,0), (1,1) for top left, bottom left, top right, bottom right, or multiples of that for repetition (good for ex. large background walls). We do camera movement by simply passing a uniform for the camera offset and add/subtract that from the position attribute.

To draw many sprites using the same texture, we can make our buffer(s) contain space for n sets of vertices+indices. Since even modest hardware can draw millions of triangles, this will let us draw all the sprites we'll ever need. However, we don't just want one texture, we want many different. One way of doing this is to either allocate one buffer set for each unique texture state or keep multiple state's texture data in the same buffer and only draw the parts relevant for the current state. However, this means we'll be doing an awful lot of state changes for very few triangles. This will probably cause a lot of pipeline stalling, where the GPU is essentially working on only one triangle at a time, instead of the tens of thousands in parallel it can.

We should get rid of the state changes. There are two ways of doing so: texture atlas or array textures. A texture atlas can be built from rectangles known in advance. I've found a writeup for how to do it somewhere, but can't find it right now. There are lots of resources online and there are tools if you don't wanna write it yourself. . This works fine if you don't need texture repetition, i.e. were a quad is so large it's texture repeats over it's surface. If you need that, an easy fix is to just subdivide your sprite. If that doesn't sit well with you, you can add a second set of texture coordinates to indicate the sub-rectangle on the sprite sheet to sample from and compute the wrapping manually, i.e. do fmod on the local coordinate and offset+scale to the sampled rect.

(1/2)


7871a8  No.15463331

>>15463330

The other option is array textures, which lets you just pick the texture to sample from using an index. It's technically still one texture state, you just have the ability to do one level of indirection on the data samples. There is one important limitation of array textures. Each mip level must have the same dimensions. Essentially, each texture must be the same size. This is ok if all textures are the same size anyway (like in minecraft), but they probably aren't. If they are all power of two, you can get around it. The restriction is applied per mip level and GL allows you to omit any mips you don't want, at zero cost in sample overhead or memory use. So if your textures are all one of 8, 16, 32, 64, you simply set the array texture as 64x64, then all textures of those dimensions go into their own slots at mip 0, 32x32 goes into their own slot at mip 1 and so on. You get a sort of sparse 2d array. When sampling, GL will automatically sample from the best mip available. If only one is available, that's the one that will be used. mip 0 is completely optional. The problem then is if your textures have different aspect ratios. I don't think there's a solution to this besides padding. Each non power of two texture gets padded to the closest greater size and you handle the padding with UV coords. If you're sensitive about wasting memory, a texture atlas is more efficient.

If you do all this, you'll be able to draw everything with one call. In fact, you only need to do one state change each frame: camera pos. Everything else is on demand. Sprites need position updates (which for enemies is probably each frame, so ok, that's a per-frame state change too), update vertex data in host memory and upload to device memory. Batching this to use as few glBufferSubData() calls as possible is pretty easy. Just list all the updates according to memory pos and collapse adjacent. You can try to keep live sprites that are expected to move close to each other so you'll only need one data upload of your buffer, by for instance having two regions in your buffer, one for static and one for dynamic. Maybe even add a single extra draw call and keep static and dynamic data in separate buffers. Need to delete/hide a sprite? Zero out it's indices, GL will cull the degenerate primitives. You can over-allocate your buffers so you don't have to re-make them all the time.

All this should give you a very high-performant 2D sprite render engine.

(2/2)


67d81d  No.15463345

>>15463330

>>15463331

I should mention that I'm using SFML with C# so I don't have the ability to poke OpenGL stuff directly (I mean I can but I don't know how and shouldn't need to). Most of SFMLs performant drawing is done by calling a render targets most general draw function

>Draw(Vertex[], int start, int count, Primitive type, Render states)

I use quads which creates the additional triangles for me automatically and render states includes the texture data and blend settings as well as shader and transform.

Anyways I'll give your post a thorough read when it's not fucking 6am


6ab532  No.15463348

>>15463330

>as 4 vertices and 6 indices drawn using glDrawElements(GL_TRIANGLES)

or just as 4 vertices and drawn using glDrawArrays(GL_TRIANGLE_FAN), which is probably the cheapest approach possible. Least amount of vertices, doesn't rely on features such as point size or a slow geometry shader.

>To draw many sprites using the same texture, we can make our buffer(s) contain space for n sets of vertices+indices.

Or use instancing. Shove the instance-specific data either in a uniform or a second vertex buffer.


67d81d  No.15463352

>>15463345

Also I don't think I can work with array textures through SFML. I already had plans for sprite atlasing which is fine, but in some cases I might keep the same state information but change the draw order for depth sorting which would need me to do stuff


7f7464  No.15463354

>>15463330

You can make it faster with glMultiDrawElementsEXT since it reduces all of your calls to glDrawElements between texture rebinds into one call. glBufferSubDataARB isn't guaranteed, query GL_ARB_vertex_buffer_object before loading it and have a fallback.


27fc5d  No.15463368

File: 5dd27e18f497295⋯.png (62.44 KB, 416x259, 416:259, vwgprcm5.png)

>videogames are cancer, make your own

>game engines are cancer too, make your own

>actually operating systems are cancer as well, make your own

I'm getting ready to give up


6f779d  No.15463399

>>15463368

Computer hardware is cancer, make your own.

>tfw I'm actually learning CPU architecture on the side to see if you can make a true TempleOS machine


7871a8  No.15463591

>>15459476

I've looked into this quite a bit. I haven't actually written any code, but here are my findings anyway:

Userspace applications communicate with the X server using a serial protocol. You can do this yourself by opening /tmp/.X11-unix/X0. You write to it to ask X to do stuff like spawn windows or submit GL commands (in direct mode it uses shared memory and ioctls for GL performance) and read from it to receive events. However, there are libraries to do this for you. The two I know of are Xlib and xcb. Xlib is old and busted, xcb is new and hawt. For instance, the only way of getting events from Xlib is a function that blocks. There's no polling. So to have an application that isn't terrible, you have to use a thread. With xcb, you query the event queue and explicitly retrieve N events. If that many events aren't available, you get an error reply (or a partial response, I haven't checked the details). In general Xlib tries to hide the fact that the X window system is a client-server model by blocking client requests until the server responds. xcb gives you a connection object and an api to build requests, batch them, submit them when you want and a handle to the submitted query for polling request completion and response.

The problem is that in order to do GL, you have to use GLX, which is built on top of xlib and only works alongside it. So you can't go xcb pure and do hardware-accelerated 3D graphics. You have to use both. Xlib opens a display and sets up GLX, then you extract an xcb connection from that which does input handling. But afaik you're still stuck with Xlib page flipping, which as far as I understand it is not a good fit if you want full control of when you want to present your frames. For instance, I think that if you just do the naive approach of submit all your calls, then flip, then loop, you'll end up queuing multiple frames. The driver limits the queue depth to maybe 5 or so, but 5*16.6=83ms latency between keyboard and what you see on the screen. Not very pleasant. I have no proof this is actually happening, but I feel like some games have input latency issues and I suspect this is why.

Khronos has a windowing api called EGL that look like the solution. It's supposed to be multi-platform, so it should work across everything, even Windows. It's supposed to let you create a window and draw to it. Though afaict it's not about input, only drawing. However, I'm not sure it's actually widely implemented. EGL 1.4 is implemented in Mesa, so any Linux install using that should work, regardless if the system is using X or Wayland. In fact, it seems Wayland is very happy with EGL, since it doesn't have GLX natively, though I think there's some compatibility layer, since almost everything is still using X. So that gives you AMD hardware as long as they're not using AMDGPU-PRO for userspace. Luckily, that's only used by workstations and has worse performance than Mesa, so no gamer should be using it. The only other option is fglrx, but that's deprecated like crazy and radeonsi has better legacy support and performance, so nobody should be using that either. As for NVIDIA, it's either nouveau+Mesa, which has crap performance because nouveau can only run hardware at boot clocks, which is like 100MHz, or nvidia binary driver, which is an all in one package. I'm not sure about EGL support there, but I do know that for Wayland, NVIDIA is pushing EGLStreams, which is an extension for EGL for doing buffer management, while everyone else wants GBM, which Mesa implements. So I assume you'll be able to create a window though EGL on nvidia using Wayland. As for X, all I can find info about is using EGL to circumvent X for doing headless rendering. However, I still think it will work with nvidia. For Intel, there's only one kernel driver and you use Mesa for userspace, so that should work too. Last, EGL is the standard method of context creation on Android, so free mobile support?

There is an important caveat in that apparently EGL was originally conceived to expose GL ES on mobile paltforms, with desktop and regular GL being an afterthought. So it's possible you won't have access to full GL through EGL everywhere, even if the platform supports it through some other API. In general, anything EGL on desktop seems like a mixed bag. I'd need to experiment a bunch to know for sure. I've also found indications that you might need a heap of libraries to make it work, like ANGLE and various SDKs. Not a fan of that, personally.


7871a8  No.15463608

>>15460137

Actually, many 10+ Intel iGPUs don't even go past 1.5 on Windows. On Linux, the same hardware does support 2.1. At least it did, but that either has or is going to get disabled, because apparently Chrome was detecting and using those features on said platforms, which was slower than using fixed-function, so of course rather than fixing Chrome, they gimped the driver. True story, there's a Phoronix post about it that I cba to dig up atm.


7871a8  No.15463718

>>15460585

The way I learned was by installing SFML, going through the tutorial for how to get a window to pop up, looking up the incantation to compile the code with gcc, then kept following the SFML tutorials to the end. Once I knew how to accomplish the basic stuff (making a window and getting input), I followed http://www.opengl-tutorial.org/ to understand how to do OpenGL. When I needed to split into multiple files, I learned how to set up CMake. Eventually, I knew enough to start making progress with a minecraft clone (that I haven't worked on in over a year). For me, what works well is to have a large long-term goal with an small short-term goal to move in what seems the right direction. I never bother outlining every step to get there or listing all the systems I'll need. But I do read a lot of stuff, even stuff that isn't useful for what I'm currently working on.

You can never predict or plan a software project, unless you've done the exact thing you're making before and already know everything you'll need. Software architecture is BS, it's just people with enough experience to know how to build common things repeating the same steps that worked the last time. It's like if a construction worker would call himself an architect by building a permutation of a modular design, like a house with an extra floor. Real software engineering is exploring the unknown. So just go at it. Start exploring and trust that you'll eventually reach the destination. Any dead end is just another valuable learning experience. Point is, humans learn by playing, so do that. Don't try to study, that just kills your motivation. Approach it with confidence and joy. Just start hacking. Just like make game.


7871a8  No.15463832

>>15463348

>GL_TRIANGLE_FAN

Trifan/Tristrip is an option for single sprites, but that's not what I was aiming for. I don't see an obvious way of extending that to n sprites (i.e. 4*n vertices). The goal was to not do one drawcall for each sprite, but to draw multiple sprites with one call.

>>15463354

>glMultiDrawElementsEXT

Isn't it easier to just build a vertex+index buffer and be done with it? If you're drawing a thousand sprites, multi draw will still have some overhead in squashing together all your drawcalls into one. With a VBO+IBO, that work is done once and never again.

>between texture rebinds

Surely you can't tell GL to rebind textures inside a multidraw? By inside I mean in between two of the draw calls. Of course, if you use an atlas or arrays you can do that, but again isn't it easier to just go with a single buffer pair with everything you need?


6b6340  No.15463857

File: 6fee053afea75b3⋯.jpg (61.93 KB, 658x613, 658:613, 1520995436142.jpg)

>worried I am not good enough for steam

>suffer from depression about not being good enough for steam quality

>some faggot upload futanary scat to steam

>mfw


6b6340  No.15463945

File: 2214432e972582b⋯.png (470.74 KB, 1440x900, 8:5, Escritorio 1_001.png)

>godot alpha doesn't have pixel bleeding

BASED.


a0a47a  No.15463966

>>15463945

the shadows only make her look worse

can't you at least make it all vertex lit?


6b6340  No.15463970

>>15463966

I have no clue what I am doing.


6ab532  No.15464118

>>15463832

>draw multiple sprites with one call.

Yeah that's why you use instancing.


3cc89b  No.15464720

File: a3844fe25ee259f⋯.jpg (44.66 KB, 473x410, 473:410, 0fbf6d00839320c4899b2e88b0….jpg)

>>15459170

add this type of swimsuit


a0a47a  No.15464737

File: ccbadd9bd1e98c5⋯.jpg (42.13 KB, 640x469, 640:469, ccbadd9bd1e98c55c2a12db552….jpg)


3cc89b  No.15464758

>>15463857

>I am not good enough for steam

It's very hard to not be good enough for steam, what are you making? A evangelion fangame?


6b6340  No.15464767

File: 65b8886661e09e2⋯.png (631.95 KB, 1280x512, 5:2, sky.png)

>>15464758

yeah, I'm making an evangelion fan game, but I think I suffer from impostor syndrome and I just got very depressed because I suck at making games, even at my best effort.

pls rate my cloud.


2a9ea5  No.15464795

>>15464767

It looks fine


6b6340  No.15464815

File: cf6ea7c91ff73dd⋯.png (439.96 KB, 1026x630, 57:35, Evangelion_003.png)


3cc89b  No.15464822

File: 1f8f28cb89e686a⋯.png (658.95 KB, 1280x512, 5:2, ClipboardImage.png)

>>15464767

I believe that there is no such thing as a maximum quality you can achieve regardless of how skilled you are.

Skill at making a thing only reduces the time needed to reach a certain level of quality.

Keep improving the thing you did until it gets good. Not only it will make you better at it, the thing will look better.

Remake the thing from scratch if need be, just keep improving the thing until it is good rather than complaining about how bad you are at making the thing.

Your cloud looks very dirty, by that I mean the parts where there is no cloud you can clearly see the brush marks highlighted in red to be extra obvious what I am talking about. Besides that, it could pass as some impressionism artwork, it does not look bad, it's just not something you would normally put in a game


7f7464  No.15464864

>>15463591

>For instance, the only way of getting events from Xlib is a function that blocks.

You can do it without blocking like so:


XEvent event;
while(XPending(sys->disp)){
XNextEvent(sys->disp, &event);

>>15463832

You can use degenerate triangles between sprites to have them in the same triangle strip. Not a good idea for this, though.

The multidraw removes overhead because you have to communicate with the OpenGL server. You can't rebind textures inside of a multidraw. It's signifigantly quicker than many draws to drawelements. multidraw is not an alternative to VBO's, it is a compliment to them. Of course you also can't change the matrix applied to the triangles inside of a multidraw so it might not be the right thing to use for enemies. For the game world, I think it would be the right thing to use.


b1b7ea  No.15464906

File: 01838b8b547d261⋯.png (858.83 KB, 1000x806, 500:403, teddy bear resized.png)

3 days to do this, entirely on mouse, but I feel I learned a lot. The most important thing is that bezier curves are your friend, wife and lord savior, no way to make a proper straight or curved line on mouse.

This is the first time I painted since high school and it was actually kinda fun, almost makes me wish my game was entirely 2d. Now I need to look for anatomy lessons so I could do a portrait


6b6340  No.15464923

>>15464906

buy a tablet.


6f779d  No.15464981

>>15464720

>saying "you should add this" to the guy who just told you that the guy working on the project you're saying it should be added to doesn't post here anymore


3cc89b  No.15465090

>>15464981

>thinking that I read the post and didn't just look at the webms


61a3ae  No.15465239

>>15464906

>no way to make a proper straight or curved line on mouse

There's always a way anon, even tablet users use giant canvas then shrink it to the right size for the final product, that should make any curve appear smooth, generally you can use illustrator or inkscape for lineart and flat coloring, it automatically smooths your lines and you can control the width for that brush feel, you can also edit the line to fine tune it

t.couldn't afford a tablet so been using AI for a long time.


3cc89b  No.15465270

>>15464906

I've found that playing doodle or die and the like will get you to improve quickly from absolute dog shit to barely not terrible, just from constant practice. It's also quite fun


a0a47a  No.15465287

>>15465270

>>15465239

>>15464923

>>15464906

in all seriousness, just get the cheapest drawing tablet to test how it feels

i made the mistake of getting an intuos 5 years ago and only then realizing that i actually hate art, especially digital


3cc89b  No.15465823

File: 421a39ef3ff3a23⋯.png (117.94 KB, 485x737, 485:737, hair elly.png)

Too bright maybe?


f48cd6  No.15465888

File: 25a347c44729eac⋯.png (8.51 KB, 530x230, 53:23, Capture.PNG)

File: 4ecce851eb3455f⋯.mp4 (1.74 MB, 1920x1080, 16:9, Desktop 2018.09.17 - 15.52….mp4)

why is this speeding up when I have it slowing down?

does a Navmesh agent make "Speed" do nothing?


f1485f  No.15465911

>>15465888

how about

<transform.postion - target.transform.position

instead of

>target.transform.position - transform.position

?

This is a guess, I have no idea.


f48cd6  No.15465956

File: 53c7e01e0edbe7b⋯.png (21.5 KB, 540x354, 90:59, Capture.PNG)

>>15465911

Only the Nav Mesh Agent speed does anything but when trying to do

NavMeshAgent.Speed = .5f;

it says Speed is not a definition of NavMeshAgent.

==EVEN THOUGH== you can change the value in the component window.


f48cd6  No.15465966

>>15465956

nevermind

agent.speed is what it is


4bf096  No.15466128

File: 0130d1798f68aaf⋯.jpg (63.12 KB, 1024x1024, 1:1, Example.jpg)

>>15464906

Get pirate the program "Lazy Nezumi Pro". This program automatically smoothes your mouse movement. Enables you to draw lines with your mouse cleanly and works even better in combination with a Tablet if you decide to get one later. Pic related is a quick example i made with the mouse so you get the idea.


a0a47a  No.15468235

do most unity character controllers use navmesh or rigidbodies or something else?

my artist is shitting on my for refusing to use physics when the current navmesh controller works fine and he wants a second oppinion


181033  No.15468290


4bf096  No.15468364

>>15468290

Line smoothness/quality is the same. Nezumi has more options available though but they're mostly geared towards Tablet so not much of a difference if you're only using mouse. The only thing i dislike so far about shark is that your mouse speed gets slowed down for some reason, which makes selecting tools and shit a hassle, you can only avoid that by turning the tool on and off with a hotkey based on when you need it. If that doesn't bother you you're good. However if you get a Tablet i would definitely take Nezumi.


6f779d  No.15468391

>>15468290

>>15468364

If I remember correctly, Nezumi also has some additional features such as perspective support tools.


27fc5d  No.15469042

>>15468391

>perspective support tools

That's the graphics program's job.


a0a47a  No.15469080

Streamable embed. Click thumbnail to play.


fdd87e  No.15469331

>>15469080

did he died

poor froggo doggo


6f779d  No.15469402

>>15469042

Graphics programs only add help-lines generally, this thing actually snaps your lines to the perspective grid


f48cd6  No.15469612

File: 2d383739cffcb79⋯.mp4 (2.4 MB, 1920x1080, 16:9, Desktop 2018.09.18 - 13.55….mp4)

File: 6800007dfbccdd5⋯.png (15.56 KB, 1158x848, 579:424, want.png)

File: 873e58ededc3eb6⋯.png (15.46 KB, 768x472, 96:59, asdfasd.PNG)

File: bb6106117f95a47⋯.png (19.04 KB, 588x509, 588:509, asdfasdf.PNG)

Trying to get object to continue moving after my Swipe() method like it's carrying momentum.

Instead it restarts Swipe the instant it gets to the edge of the attacking distance.

I want it to start the Hop() method as soon as it touches the edge so it will circle around back


161830  No.15469674

>>15463970

Yes, it's in the material flags.

>>15469402

Adobe Illustrator and Photoshop (I think) do this with "perspective grid". >>15469612

Make the enemy check that it's facing the player before attacking.


3cc89b  No.15470037

File: 48893fcaacd032f⋯.png (225.34 KB, 1340x798, 670:399, elly1.png)

File: a6c9cc09d2380ce⋯.png (109.79 KB, 1340x798, 670:399, elly2.png)

What do you guys think? My brother is autistic screaming that the wings should be black


ec2955  No.15470080

File: e70af2be0568da1⋯.png (226.76 KB, 1340x798, 670:399, ClipboardImage.png)

File: 9660835f2015bf7⋯.png (217.16 KB, 1340x798, 670:399, ClipboardImage.png)

File: 76468a4a367828b⋯.png (239.48 KB, 1340x798, 670:399, ClipboardImage.png)

File: b16fc26a3977fd4⋯.png (233.6 KB, 1340x798, 670:399, ClipboardImage.png)

File: 5dbd973e2b7a4c8⋯.png (241.61 KB, 1340x798, 670:399, ClipboardImage.png)

>>15470037

It could work depending on the color scheme for the rest of the body, but it doesn't really work for me. I've tried a few color schemes, but couldn't really get one that worked right, so you might have to go for black after all. The blue wasn't too bad, though, you might be able to do something with that.

Also, her face looks strange, but it might just be the angle. She looks either drunk, or mid-blink.


3cc89b  No.15470117

>>15470080

Ok i guess it will have to be black then

She is the serene eternally tired-looking bookworm that starts reading during the afternoon and when she checks the clock it is 4am, and she gotta open the library at 7am

What program did you use for those edits? that was quite quick


ec2955  No.15470149

>>15470117

>She is the serene eternally tired-looking bookworm that starts reading during the afternoon and when she checks the clock it is 4am, and she gotta open the library at 7am

Might just be strange angle, though the eyelashes make it seem like she's starting to blink. Not an expert though, so I don't know.

>

What program did you use for those edits? that was quite quick

Gimp. Just selected the wings by color, moved them to a separate layer for convenience and played around with the Colorize tool.


3cc89b  No.15470460

File: 6f35c46da371863⋯.webm (8.34 MB, 1226x526, 613:263, 2018-09-18 19-10-13.webm)

Finally, I have something to show.

Now there is an animation for ledge grabbing, although it's still rough and the feet don't move.The hands only go up if you are moving.

I will probably refine it by changing the rotation of the hands with it's position, probably pointing to where it wants to go or something like that.

The clipping issue will probably be solved by letting the hat be knocked back, and when idle, Adam will put it back on his head, but that's pretty low priority.

I am not sure what I want to do with the whole being able to look up thing, I did it to test what height would look best for ledge grabbing, now I wonder if it should be a feature


e28359  No.15470864

in C++ (maps), or C# (dictionaries) is the find method stochastic or what?


6b153a  No.15471278

>>15470864

From what I know a lookup uses a hashed key, which accesses hashed entries that are in buckets (can quickly find which bucket to search through, and do a constant time index access), and are constant in lookup time.

Neat aspect is that dictionary are always the size of a prime due to the internal hashing algorithm used (i.e. if u init w/size 10, it does size 11 on the backend, etc).

There's some implicit boxing/unboxing in a generic C# dictionary if you don't use a equality comparer though. Using an equality comparer will speed it up the baseline "constant" access time by a bit.

TL;DR

Constant lookup time for C# dictionaries, and it uses hashing/buckets


0f1e49  No.15472268

>>15470864

>http://www.cplusplus.com/reference/map/map/

>Maps are typically implemented as binary search trees.

Due to the ordered requirement. For unordered hashmap, there’s http://www.cplusplus.com/reference/unordered_map/unordered_map/


74b189  No.15472585

File: 545484b309be826⋯.jpg (74.08 KB, 540x702, 10:13, nZtocZM.jpg)

I'm having so much trouble with love2d that I'm considering just using flash. I feel so retarded, spent days trying to combine smooth tile-based scrolling to a map loader that generates the map from a string which makes it easier to design maps when you can assign an image quad to G for grass or R for roof, etc. Made no breakthrough on getting both to work together, took a break from it for a few days, came back to it and still can't figure out how to combine these things. I wasn't even able to split up the code for tile-base scrolling into separate files to keep the main file clean.


6f779d  No.15472644

>>15472585

>smooth tile-based scrolling

It's either smooth scrolling, or tile based scrolling. It can't be both.

>combine scrolling and map loader

What kind of map data did your scrolling tutorial I presume you're using one considering your apparent skill level use, and why is it hard to make the loader just load its data into the exact same format?


3cc89b  No.15473286

>>15472585

just learn to load the data from Tiled

https://www.mapeditor.org/

I am pretty sure there is a editor plugin to literally all engines at this point


74b189  No.15473844

>>15472644

The tutorial was https://love2d.org/wiki/Tutorial:Efficient_Tile-based_Scrolling so what do I know.

The tile scrolling tutorial generated a map just using random(0,2) with 3 image quads, then had a function to add the quads to the tilesetbatch with:

tilesetbatch:add(tileQuads[map[x+math.floor(mapX)][y+math.floor(mapY)]],x*tileSize, y*tileSize)

but the map I'm generating from a string, using the method from this tutorial: https://github.com/kikito/love-tile-tutorial/wiki

I'll spend some hours trying to fix one error, then get a new error and spend a few more hours figuring out why I'm getting that error.

>>15473286

That's just for designing maps, while useful is not pertinent to solving my problem. Also the library someone made for loading Tiled maps is not being developed any longer.


dd22b3  No.15473897

Game with test animation system anon here.

Reworked memory layout for better projectile management as well as unit management, that will help me when I'll do collision detection.

Also did a basic firing script and now the enemy test unit move (still teleports around) and fire as scripted.

Next in no particular order:

- Collision detection, which will leads to "onHit" evaluation which leads to "onDeath" evaluation which leads to particle effects

- Instantiate units and projectile from base definition stored somewhere

- Ships design (barely started) and 3D modeling


8ab27a  No.15473910


127945  No.15474003

What's the best way to handle isometric maps? I want to have a big map with multiple Z levels and I don't know how to handle drawing order for walls and objects




[Return][Go to top][Catalog][Nerve Center][Cancer][Post a Reply]
Delete Post [ ]
[]
[ / / / / / / / / / / / / / ] [ dir / animu / arepa / asmr / leftpol / tacos / vg / vichan / zoo ]