[QUOTE=Deseteral;48877907]What's the game about? Looks very interesting...[/QUOTE]
It's still kinda a wip and an experiment right now, but it's gonna start of like a basic RPG that slowly falls apart and gets more and more corrupted as the story goes on.
[QUOTE=sirdownloadsalot;48877932]It's still kinda a wip and an experiment right now, but it's gonna start of like a basic RPG that slowly falls apart and gets more and more corrupted as the story goes on.[/QUOTE]
I had the exact same idea! I'd love to see how you do it
Working with raw OpenGL for the first time.
Decided to go with OpenGL ES 2.0, SDL2, glm and Emscripten.
Also, decided to wrap everything I do in a convenient RAII-oriented C++14 interface.
10 hours in, after tears and misery, I finally got sprites to work!
[IMG]https://pbs.twimg.com/media/CRCI4jGVEAATgBe.png:large[/IMG]
[URL="https://github.com/SuperV1234/Experiments/tree/master/emscripten"]WIP source here.[/URL]
[editline]11th October 2015[/editline]
[vid]http://fat.gfycat.com/TeemingAlarmingAcornbarnacle.webm[/vid]
Sprites!
(Position/origin/rotation using glm matrices.)
[QUOTE=sirdownloadsalot;48877903]
Test screens for a game that uses corruption as a gameplay element[/QUOTE]
I'm thinking of using glitch effects/corruption to signify a gameplay element in my game as well. Got any tips on how to do those effects well? The ones you're doing now in the last image look pretty cool
[QUOTE=Falcqn;48874150]Give the spaceship a halo so it's more obvious that you're playing as Jesus :v:[/QUOTE]
[IMG]http://i.imgur.com/RXryTgn.gif[/IMG]
say no more
[QUOTE=adnzzzzZ;48878791]I'm thinking of using glitch effects/corruption to signify a gameplay element in my game as well. Got any tips on how to do those effects well? The ones you're doing now in the last image look pretty cool[/QUOTE]
Right now a lot of it is pre-made. The GUI is sprite based. But the static bits and bobs are individual sprites that are being randomly selected and distributed around the left and bottom.
Most of these effects are just photoshop effects.
[QUOTE=Phobos001;48876718]:toot:
'Finsihed' the first two levels so far! Here is the first level in its entirety, with the combat mechanics as well! (Not amazing because it's teaching the fundamentals of the games mechanics, but yeah)
[media]http://youtu.be/8vOz3a7Vafg[/media]
Just gotta figure out how to get FRAPS to record it without having to go fullscreen with DWM (HTML5 WebGL etc.)
[sp]Also, the health orb is cut depending on the angle of your mouse movement and where you struck it. Call it a poor mans version of the MGR: Revengence slicing mechanic[/sp] :excited:[/QUOTE]
Dude, this looks awesome! My only complaint is that there seems to be no skip button for the only-dialogue cutscenes, even you were waiting for it to switch. Other than that I love this concept, the choice of font for the bad guy was nice too.
On my game, I got basic bullet physics implemented, with convex shell shapes:
[vid]https://dl.dropboxusercontent.com/u/357850863/ShareX/2015/10/2015-10-11_14-23-30.mp4[/vid]
I do have regular mesh shapes but I haven't quite figured it out yet:
[vid]https://dl.dropboxusercontent.com/u/357850863/ShareX/2015/10/2015-10-11_14-27-57.mp4[/vid]
It's still buggy as hell but hey, it's better than nothing.
[vid]http://giant.gfycat.com/MinorWideeyedGerbil.webm[/vid]
10000 entities.
Every entity has it's own hue (float), color (vec4) and MVP (model-view-projection mat4).
10000 draw calls (no batching).
50-60 FPS [B]on Emscripten.[/B]
I'm satisfied... for now.
Just FYI: I couldn't get more than 60 FPS with 1500 RenderCopyEx SDL2 calls.
And RenderCopyEx sucked in terms of flexibility.
I implemented a sprite renderer in OpenGL ES 2.0 (and started an OpenGLES / SDL2 / Emscripten C++14 library). :D
[vid]http://zippy.gfycat.com/FreeDelectableIndianskimmer.webm[/vid]
Handsome Matt Told me to stop posting gifs. So I killed him.
Been adding randomly generated cities to SkyMercs, woo. [url]http://skymercs.com/blog/[/url]
[img]http://skymercs.com/blog/wp-content/uploads/2015/10/city_pic.png[/img]
If there was a server that allowed application devs to register their application and then get an app ID, and if it also allowed clients to register an account and get an unique account ID
One could write a chat application of sorts and register it on the server, and if you didn't want to host your own rendezvous server because that chat application would be peer2peer and it's a pain in the ass when both peers have dynamic IPs and you want them to contact each other without knowing the IP.
So how would the clients in said application communicate? Well both of them would register an user account and get an unique account ID (and allow it to be used with the application ID), and one client would know the unique ID of the other client.
When both of them turn the chat application on, they communicate with the master rendezvous server and update their current IP with it, so when client A wants to message client B, it queries the server by the unique user ID and the server returns currently valid IP by which one can message the client B.
There would be other useful features where one can query all user IDs that are registered with an application ID, NAT punch-through and stuff. It would eliminate the need to host any kind of server if you wanted to write simple networked applications with a large user count.
Would you use it?
[IMG]https://facepunch.com/fp/ratings/tick.png[/IMG] - Yes
[IMG]https://facepunch.com/fp/ratings/cross.png[/IMG] - No
That's really just glorified NAT punching IMO
It's one of those things you'd only use if everyone else did - with that said, services with such a prerequisite have done just fine, so...
[QUOTE=cartman300;48880582]If there was a server that allowed application devs to register their application and then get an app ID, and if it also allowed clients to register an account and get an unique account ID
[...]
Would you use it?
[IMG]https://facepunch.com/fp/ratings/tick.png[/IMG] - Yes
[IMG]https://facepunch.com/fp/ratings/cross.png[/IMG] - No[/QUOTE]
Only if it's properly DoS-resistant.
It really depends on the precise API though, and whether it would give me the option not to rely exclusively on this service.
This will probably sound boring and/or pointless to most of you, but I'm working on a program to catalog my massive video game collection. It uses Microsoft SQL, so I could get some experience using it. Work in progress, of course.
[img]https://photos-6.dropbox.com/t/2/AAAZT-PzzfMTdPI95LPitl6SbACXqbOqMzMOOeky8_l-mw/12/27247197/png/32x32/1/_/1/2/Screenshot%202015-10-11%2016.40.08.png/EOXOwRQYiDMgAigC/E50SbcyXegdb10JGMtPAIpkwH7fqv51PVRfbKrgSxbQ?size=1280x960&size_mode=2[/img]
Working on a lock-free queue, no problems so far, gotta do more testing before I can be sure.
[code]
// snipping code ...
[/code]
[B]Edit:[/B] Several problems with this code, I would not suggest using it.
Randomly decided to implement 2D dynamic shadows, turned out looking pretty neat.
[IMG]http://i.imgur.com/LBcDsTN.png[/IMG]
Idea for a game: MMO that has your generate bitcoins as you do quests. the quest reward is equal to the amount of bitcoins generated while doing that quest. the value of bitcoins ingame is worth a proportionate amount for the items in the game so users transfer bitcoins to you for work they did in exchange for meaningless vanity items.
[QUOTE=proboardslol;48883254]Idea for a game: MMO that has your generate bitcoins as you do quests. the quest reward is equal to the amount of bitcoins generated while doing that quest. the value of bitcoins ingame is worth a proportionate amount for the items in the game so users transfer bitcoins to you for work they did in exchange for meaningless vanity items.[/QUOTE]
Now introducing P2W AMD hardware. Back when it was actually reasonable to mine Buttcoins on Nvidia hardware (or was it?) I was still playing games while it was going so it's definitely possible to do both at once.
I think what I described is a middleground between Folding at Home, illegal botnets, and Habbo Hotel
[QUOTE=thatbooisaspy;48880283]Dude, this looks awesome! My only complaint is that there seems to be no skip button for the only-dialogue cutscenes, even you were waiting for it to switch. Other than that I love this concept, the choice of font for the bad guy was nice too.[/QUOTE]
Gotcha covered man! Thanks for the feedback! :smile:
[media]http://youtu.be/UgI_6oCL1jA[/media]
This is also a preview for level 3 (mode7 shader made by R0J0Hound that I abused for this effect), but it's still WIP because I wasn't as productive as I was yesterday :ohno:
[sp]By the way, the game is called BULLET-POINT now. Good or bad?[/sp] :v:
Got started on balancing the game and here's how I'm doing it
[IMG]http://i.imgur.com/dA08upo.png[/IMG]
Each thingy has 2 parts, min and max. Operating in ranges makes sense because you want there to be multiple builds that are weak at some points and strong at others, or that progress differently, etc. The ranges have to be wide enough to allow for these builds to exist but not so wide to the point where the difference between each build can mean that you'll progress too slowly.
The multiple 1, 2, 3, 4 on the side are progression points. These progression points can be levels, some amount of resources spent or whatever else that represents different stages of progression. Right now I'm doing only 4 because it doesn't make sense to do more when I don't even know what the resource will be exactly and how it's going to gate stat progression.
So that I can have control over how fast/slow the game can be I started with the hits part. And there I set that on the first tier you can take at most 4 hits to kill an enemy and at least 2. So in the first tier it'll be impossible or very unlikely to kill an enemy in 1 hit. After this I defined player damage. By itself player damage number doesn't mean anything, but I have to set it anyway otherwise I can't calculate other stuff. So I just set some numbers that I think look OK. After this I can calculate the range of HP enemies should have based on how many hits it should take as well as player damage. I do something similar for player HP and enemy damage, the only difference being that since the game is heavily AoE based, for enemy damage I also take into account the pack size. Right now this calculation is mostly wrong because the model for this should take into account a looooooooooot of things, but since I don't have many numbers yet I can't do it right. Luckily nothing really depends on enemy damage yet so it's OK for it to be wrong as long as it feels right-ish. After this I do a similar type of calculation for elite enemies.
And then I started incorporating time into the model by taking into account the attack speed stat and the duration each attack animation takes to finish. per_hit_delay is a delay that I add to each attack to model kiting (a lot of the times so you don't take damage you move around after hitting the enemy once), since it adds some amount of seconds to each attack. I've since made it lower to 0.2 because sometimes you just attack a bunch of times in a row and I found that 0.2 got closer to the results I was getting in the game as I'm testing this. Anyway, this feeds into the animation attack delays, which is basically just a calculation that goes like (animation_no_of_frames*animation_delay)/attack_speed + per_hit_delay, for both min and max. And so with this I get accurate values for how much time it takes for each attack to finish. Then I use these numbers to calculate DPS of each attack, which is basically player_damage/attack_delay, but can get more complicated based on the modifiers of each attack.
Now since I have time in the model I can also calculate the amount of time it takes to clear a pack of enemies with each attack. The formula I've found that matches the reality of the game closely is (pack_size/no_enemies_hit)*(max_hits)*(attack_delay). no_enemies_hit is an attack dependent number that will return the maximum number of enemies that can be hit by that particular attack. For instance, with pierce = 3, the ranged attack can only hit 3 enemies per attack, with pierce = 2032329, it can only hit enemies in a line, so this depends on how the pack is organized at the moment, but I want max and usually max is the pack_size. And so with this I can calculate how much time each attack takes to kill a pack and now with both DPS and CLS for each attack I can start to compare them better, see how I can change some stats to make each attack scale in the way I want and so on.
For instance, it seems that the ranged attack is better than the melee attack in every way because it has higher DPS, better CLS and it's also a ranged attack, so you're safer. Seeing this I can do a number of things to make melee attacks more viable, like increasing their AoE size (which would increase their no_enemies_hit which would increase their clear speed), add a damage multiplier to the attack's formula so that it deals 1.5 damage or something, add an attack speed multiplier to it so it can repeat faster, and so on. Basically however I want this particular attack to be on par with other attacks I can do it because I've successfully reduced them to numbers. As my model gets more complex and more accurate I'll have a few more numbers to compare, but at the end of the day it's the same idea, in that you wanna keep multiple types of attacks/builds within the available ranges, and since all those numbers are being calculated from those ranges by definition they'll be inside, so yea.
Also implemented path finding and optimized a bit. Here's about 150 entities running their behavior trees + pathfinding:
[VID]http://giant.gfycat.com/RightPersonalGilamonster.webm[/VID]
Npcs jump around and look at you when you get close to them
[vid]http://fat.gfycat.com/CreepyShrillGuineafowl.webm[/vid]
I'm getting ready to implement spatial partitioning for my entities. At the moment I'm just too lazy to do anything, however I have been thinking about it.
Should I let my objects loop through the objects in the cell they're in to determine collisions(or whatever needs to be done with stuff in the same cell) or is that not the way to do it?
For all you python devs, please scream with me at the horrors of what i've done:
[code]
from ohgodwhy import decorators
class A:
def __init__(self, y):
self.x = 2
self.y = y
@decorators.noself
def b():
print(self.y * self.x)
a = A(2)
a.b() # Prints 4
[/code]
[QUOTE=TheEyes;48884054]For all you python devs, please scream with me at the horrors of what i've done:
[code]
from ohgodwhy import decorators
class A:
def __init__(self, y):
self.x = 2
self.y = y
@decorators.noself
def b():
print(self.y * self.x)
a = A(2)
a.b() # Prints 4
[/code][/QUOTE]
Aaaand why would you do this if I may ask? :v:
[QUOTE=TheEyes;48884054]For all you python devs, please scream with me at the horrors of what i've done:
[code]
from ohgodwhy import decorators
class A:
def __init__(self, y):
self.x = 2
self.y = y
@decorators.noself
def b():
print(self.y * self.x)
a = A(2)
a.b() # Prints 4
[/code][/QUOTE]
If I understand correctly, you're replacing the verbose self parameter with an even more verbose @decorators.noself annotation?
You know there's an actual problem when just one coworker and myself (the intern) are having a problem with everyone constantly using singletons everywhere. 'no its cool what could go wrong bro'
The word [I]"Singletons"[/I] sounds like the title of some horrible sitcom about a group of people who don't have significant others.
And I think of that every time I write one.
Sorry, you need to Log In to post a reply to this thread.
