I'm not sure there's a hardcoded limit on the number of objects, but like any engine, you can run into precision issues depending on the range of scale you want to use. I'm not using Unity's single precision vectors for positioning...I'm using a custom double precision coordinate space for positioning of the stars, that boils down to single precision in the rendering system...but the camera is centered around the origin always, so the viewer gets the most out of default precision available regardless of where they are in the double precision coordinate space. If that makes sense.
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
Indie Game Development Discussion Thread | Of Being Professionally Poor
- Thread starter chubigans
- Start date
- Status
I'm not sure there's a hardcoded limit on the number of objects, but like any engine, you can run into precision issues depending on the range of scale you want to use. I'm not using Unity's single precision vectors for positioning...I'm using a custom double precision coordinate space for positioning of the stars, that boils down to single precision in the rendering system...but the camera is centered around the origin always, so the viewer gets the most out of default precision available regardless of where they are in the double precision coordinate space. If that makes sense.
That sounds like the objects are not actually positioned absolutely, then, which is an advantage over other engines (like again, UDK, which I believe would require any object to have an absolute set of coordinates).
I still haven't delved too much into shaders, so anything I can read up on the subject is great.
I'm currently having some issues with some transparent background planes that get all pixellated around the transparency edges at certain camera angles despite having turned mipmaps off, making the texture interpolation linear and setting its anisotropic level to zero. It seems to happen when the camera is a bit too far either to the left or to the right of the plane (I'm using a regular camera perspective), but I can't get it to not look pixellated.
I'm currently using a very simple shader (since that's all we need), and when I tried to use the default Mobile/Transparent/VertexLit or simply Transparent/Diffuse, the planes disappeared randomly depending on the camera angle, even though the pixellation was no longer there as longs as the planes were actually visible. This is my shader:
Code:
Shader "Custom/Transparencia" {
Properties {
_MainTex ("Base (RGB)", 2D) = "white" {}
}
SubShader {
Tags {"Queue"="Transparent" "IgnoreProjector"="True" "RenderType"="Transparent"}
ZWrite On
Alphatest Greater 0
Blend SrcAlpha OneMinusSrcAlpha
Pass {
SetTexture[_MainTex]
}
}
FallBack "Diffuse"
}Here're some screen captures of the error and some object settings:
Out of the box Unity has one coordinate space, and you can treat everything as having an absolute position, translating the camera around the space as normal.
But it's all single precision. If you want higher precision positioning you have to roll something yourself.
So basically I use two spaces - one keeps track of absolute positions in double precision. The other is the default Unity coordinate space in single precision, and it basically translates everything so the camera is always at the origin. In the double precision space the camera is translated around the space, but in Unity single precision the world gets translated around the camera, so it's always at 0,0,0. So I suppose one absolute space is used and then one relative space (relative to the camera)
Maintaining object positions in two spaces brings up its own challenges though, especially since none of the libraries out of the box will work with double precision vectors etc. but it seems to work OK for now.
Diablohead
Member
FormallyMegadrive
Member
Man, writing games in JavaScript is so shit. Probably doesn't help I'm using TypeScript and an API that literally only just came out (Phaser). Should I try something else? I'm developing on Linux currently so Unity and Flash are out of the question and I've never gotten anything working with g++.
Man I'm so burnt out. I have this feeling of helplessness because I literally cant even get started. :-(
Man I'm so burnt out. I have this feeling of helplessness because I literally cant even get started. :-(
So I presented my game, 100% on it keeping my 100% for the course.
Now we've been given our final projects and I'm really wanting to make something special. Is there any tips and tricks for Unity (free license) to create some interesting looking objects? Dunno what I'm thinking of, but I need some zest in this thing to really blow my classmates out of the water.
Yeah, the 3rd guy who contributed nothing emailed me late last night saying send the game to him when I was done because he wanted to play it. Fuck off, I ain't sending shit.
Edit:
If it's an option for you and you want to try out Unity, I have an unused Windows 8 license from my MSDNAA through my University that I am not using. I would give the license to you if you wanted to install Windows and try out Unity.
Now we've been given our final projects and I'm really wanting to make something special. Is there any tips and tricks for Unity (free license) to create some interesting looking objects? Dunno what I'm thinking of, but I need some zest in this thing to really blow my classmates out of the water.
Yeah, the 3rd guy who contributed nothing emailed me late last night saying send the game to him when I was done because he wanted to play it. Fuck off, I ain't sending shit.
Edit:
If it's an option for you and you want to try out Unity, I have an unused Windows 8 license from my MSDNAA through my University that I am not using. I would give the license to you if you wanted to install Windows and try out Unity.
When you're feeling frustrated, the last thing you want is to be battling against immature tools, that's for sure.
I'm not sure what to recommend for Linux. I know you can publish to Linux with Unity but I am not sure the editor is available yet. Wish they'd hurry up with that. Hope someone can recommend something for you though meantime.
Look into mono and a OpenGL wrapper lib.
I think I am going to make a little game and instead of doing it in c++ and porting it I am going to use mono and .net.
I am surprised at how fast I got the basic engine up and running in mono. Sure there is a little overhead with .net but nothing massively performance hindering.... Yet. The time it saves from not having to do a lot of type heavy coding in c++ makes it a much mor enjoyable hobby project.
Today's galaxy work. I'm kind of live-journaling here, but I find it's good motivation to share!
Stars are no longer all the same...they randomly adopt from a subclass of the spectral types, class O to M, with various resulting colour tints and luminosity etc. Though size is still the same.
The distribution is not based on any physical numbers, is purely random, so that'll have to be fixed. The colours are also a bit off, but this is far from the final render treatment anyway...just coloured squares
Having different star classes also means the stars now have varying magnitudes. Which means we can calculate the apparent magnitude of each star from the camera's location...and thus, which stars would actually be visible to a human eye.
Views like the shot above are great for visualisation. And using the same settings gives a view like this from a point inside the galaxy, looking out to one of the arms:
But it's not very realistic to see so much. When you factor in apparent magnitude the view looks more like:
Which looks a bit more like the kind of starfield your eye would see naturally from a given point, I guess.
Although every sci fi movie or TV show seems to have its own idea about how much the human eye would see, for different styles of starfield. Maybe I'll let the user customise...or maybe there'll be things like telescopes or other toys that let the user see more deeply into space.
Stars are no longer all the same...they randomly adopt from a subclass of the spectral types, class O to M, with various resulting colour tints and luminosity etc. Though size is still the same.
The distribution is not based on any physical numbers, is purely random, so that'll have to be fixed. The colours are also a bit off, but this is far from the final render treatment anyway...just coloured squares
Having different star classes also means the stars now have varying magnitudes. Which means we can calculate the apparent magnitude of each star from the camera's location...and thus, which stars would actually be visible to a human eye.
Views like the shot above are great for visualisation. And using the same settings gives a view like this from a point inside the galaxy, looking out to one of the arms:
But it's not very realistic to see so much. When you factor in apparent magnitude the view looks more like:
Which looks a bit more like the kind of starfield your eye would see naturally from a given point, I guess.
Although every sci fi movie or TV show seems to have its own idea about how much the human eye would see, for different styles of starfield. Maybe I'll let the user customise...or maybe there'll be things like telescopes or other toys that let the user see more deeply into space.
Just bought my pass but I figured I'd let people in this thread know about PAX Dev if they don't already.
http://dev.paxsite.com/
It's only two days (the Wednesday and Thursday before PAX Prime).
It was a ton of fun last year. It reminded me of what GDC was like years ago before they started letting in the press and large talks started blending with marketing. (GDC is still cool but it's clearly got bills to pay).
Everyone that spoke last year was really open and transparent with information in their talks so it was very cool.
Student badges are also discounted, so that helps.
http://dev.paxsite.com/
It's only two days (the Wednesday and Thursday before PAX Prime).
It was a ton of fun last year. It reminded me of what GDC was like years ago before they started letting in the press and large talks started blending with marketing. (GDC is still cool but it's clearly got bills to pay).
Everyone that spoke last year was really open and transparent with information in their talks so it was very cool.
Student badges are also discounted, so that helps.
I'd say next time, 100% you!
In other news, I have nearly completed all the components for the BMX
Edit: BMX is complete! Now to make it move..
You can rig it with a motor.
FormallyMegadrive
Member
Yeah, I had a similar experience with my 6 month - end of year game with two others. One got evicted in the beginning and the other ended up literally telling me he wouldn't do what I asked him to do (I was dev lead) so I told him to just pack up and move to a different group cos I was sick of his shit. I ended up basically coding the shitty game myself.
My laptop's harddrive screwed up gigantically and I installed a temporary Linux partition to troubleshoot and potentially diagnose the issue, which is why I'm using Linux currently. I have a Windows 7 machine but I wanted to do my development on my laptop (aside from art) because it's more mobile and I can move around. Your offer is seriously very generous but I can't take it because I won't take proper advantage. Thank you so much for offering though.
To your first point, absolutely. I mucked around in NodeJS for a few months and really enjoyed it and got the hang of JavaScript very well. It's just I have zero clue how to handle JS for games. I know how to work it for Apps and jQuery stuff though.
This is actually a pretty good idea. I may just do this, thanks for the suggestion! I've been using Unity quite a bit and I really love the suite but there's something else about coding the game entirely yourself.
Cheers.Heh well, If we're sharing team project stories... by the time final project came round everybody knew who worked hard and who didn't in my class, so teams formed super fast and all of a sudden it was just 3 of the least productive left so they had no choice but to team up. Called themselves team misfits... they knew what happened and just kinda rolled with it which I thought was cool of them. I think it actually forced them to start putting in some effort because there wasn't anyone else they could rely on. Funny thing was that if you asked each of the 3 about their teammates, they would each talk about how they were doing all the work and carrying the team. And so they actually ended up with something decentish to show.
As to project advice, I don't know Unity, but in general what I've noticed after seeing countless projects is that all the ones that stick are the ones that focused on polishing just 1 simple clever mechanic. For example a 2d game where the world can be rotated in 90 degree increments, or a game where you can temporarily phase objects in/out of existence, or one of those that records your characters movements and can play them back, or one that transforms the environment, or one that lets you create portals, etcetcetc. If you try to do more than 1 standout mechanic, I find you just don't have the time. Having 1 awesome mechanic is so much more memorable than 3 half assed ones.
Making stuff look good depends on several things:
- Programming: Making interesting shaders
- Art direction: Making sure the colors blend well together, that different elements stand out without sticking out like sore thumbs, and basically designing the objects that have to be modeled and textured
- 3D art: Making the models look good while optimizing the UV maps and without using too many polygons, bones, etc, and giving the models interesting and characteristic animations.
- 2D art: Making decent textures and baking some visual effects that would be too expensive or impossible to draw in-engine
Overall, I'd say you could do something that looks good using only shaders, but that'd probably require shooting for a simpler aesthetic, while making good stuff and using the default shaders could result in the game looking too same-y compared to other Unity projects.
I'd say it applies more to having a single central theme as the base from which everything is built from, and that you do have to take care not to fill your game with mechanics, but I think that applies more to mechanics that clash with the main theme or take too long to develop, draining the team's enthusiasm in the meantime.
You could make a good an interesting game simply by creating a tightly-packed experience where everything works towards transmitting the feelings/experience/message you intend it to, even if you didn't have an amazing main mechanic; but yeah, it's simpler to build games based on a single, good mechanic.
I've never dealt with writing any shader code, is there any place I can get a quick crash course, or perhaps some repository of shaders free to implement?
UnifyCommunity's wiki has some shaders you could use, as well as the source code to Unity's built-in shaders.
As far as tutorials go, I'm not so sure what you should use. Unity's reference manual has a whole section dedicated to explaining the basics, but it's a bit complicated, and MonoDevelop won't be giving you any autocomplete hints or stuff like that (plus, it's case-insensitive). The "ShaderLab Syntax" section explains the rendering pipeline, and some basic stuff you can do without having to write your own vertex, pixel or surface shaders with CGProgram.
Once you've learned a bit about doing stuff such as shader properties, SetTexture, texture combines, blending, z-testing and render queues you could delve into the world of CGProgram, which is where the real magic begins, since you're given full control of everything related to rendering the object, so you can do toon outlines, sisplace vertexes for trippy effects, change texture coordinates and many more interesting things. You should read up on that link Blizzard posted a while ago (http://duriansoftware.com/joe/An-intro-to-modern-OpenGL.-Chapter-2.2:-Shaders.html).
It took me quite a while to learn how to think in shader logic, and I'd say you've got to be interested in programming if you want to make your own, but shaders let you do amazing things as long as you know the tools you can use and what you're doing.
Thanks for those resources. I also just started messing around with creating normal maps, and hell even these alone are very interesting to me.
Hey Indie gaf, could you guys give me some advice? I'm interested in making a game for Android, but I'm not sure which development tools I should use. So far I'm leaning towards andengine or unity. I'm thinking of making a 2d game, but I'm not sure about the specifics yet other than that I want to start out with something that's not too complex.
A little about my background is that I'll be graduating soon with a b.sc in comp sci/ math, and I'm pretty proficient in java, javascript, c, and c++. I've also made a simple 2d platformer Mario clone from scratch using c++ and opengl.
Any suggestions on which dev tool I should pick?
A little about my background is that I'll be graduating soon with a b.sc in comp sci/ math, and I'm pretty proficient in java, javascript, c, and c++. I've also made a simple 2d platformer Mario clone from scratch using c++ and opengl.
Any suggestions on which dev tool I should pick?
If you want to make a 2D game, perhaps you could do better than using Unity, since it's, for the most part, geared towards 3D stuff. You could make something interesting in 3D, like a Katamari or a basic first-person something (probably a platformer, since shooters often require some AI work to play well) as long as you're okat with using a Terrain mesh or Portal-esque architecture (ie: textured cubes everywhere!).
If you want to take it as a challenge, you could try to make a 2D game in Unity. A 2.5D game could work really well in there if you're interested in that, as well.
Sanic
Member
Hey guys, I was wondering if there are any good solutions to deploying an HTML5/JS game as a native app for windows and mac (and I suppose mobile as well)?
I'm not talking about programs like Construct that will do it for you, but something that will just package my from-scratch application.
edit: Perhaps it would be best to explain why i'd like to go this route. I'm interested in making something over the next few months, and I would like it to support as many platforms as possible, but i'm not interested in doing much 'porting' (so for example, if I chose C as my language, there would be some work involved in supporting Android/iOS, and I don't want that). I'm also not interested in using any game development platforms like GM/Construct, i'm building from scratch.
I'm not talking about programs like Construct that will do it for you, but something that will just package my from-scratch application.
edit: Perhaps it would be best to explain why i'd like to go this route. I'm interested in making something over the next few months, and I would like it to support as many platforms as possible, but i'm not interested in doing much 'porting' (so for example, if I chose C as my language, there would be some work involved in supporting Android/iOS, and I don't want that). I'm also not interested in using any game development platforms like GM/Construct, i'm building from scratch.
Dynamite Shikoku
Congratulations, you really deserve it!
If you want to do 2d, what about cocos2d-x?
Decided to go with a game similar to VVVVVV for my last game. I just whipped up a demo in an hour or so, switching gravity is pretty easy. Just need to fine tune that so momentum is kept when switching. I also want it so the gravity can be switched in all 4 directions, which I have working as well.
Big thing is momentum. And then finding a way to create great art and/or effects for this.
Big thing is momentum. And then finding a way to create great art and/or effects for this.
A 2.5D game would be pretty cool. I think that I just want to start off small though. Maybe work my way up to something like that once I have a bit more experience. I plan on just doing this as a side project, so I'm hoping I can come up with something that won't take too long to make.
I didn't realize that cocos was also for android. I'll look into it.
Dynamite Shikoku
Congratulations, you really deserve it!
Yeah, you can do cross platforms in c++ with x (although I have never used it myself)
Dead Prince
Banned
So i was browsing the Construct2 forum. and saw this game being made by one of the member of the team. Anet posted the game on their blog. https://www.guildwars2.com/en/news/presenting-the-rytlocks-critter-rampage-game/
the forum post is here. http://www.scirra.com/FORUM/topic66631.html
Game can be found here: https://d3b4yo2b5lbfy.cloudfront.net/rampage/index.html
the forum post is here. http://www.scirra.com/FORUM/topic66631.html
Game can be found here: https://d3b4yo2b5lbfy.cloudfront.net/rampage/index.html
Ok, I added a few new things to my game. The first thing, is the ability to change the sound level. The second, is specials. These will sometimes show up instead of normal enemy bots. If you get one (Money, Points, Atomic, Health), they will help you out.
Money and Points will double the amount you get for the rest of that level. So for example, instead of getting $5, you may get $10 or $20 and so on. Doubles each time you use it for that level and then resets. Atomic will kill all enemies that happen to be on screen. This is for enemies that have spawned. And Health will restore up to 50% of your health.
LEt me know what you guys think. Game will still be just as hard, but will enable you to get to more advanced levels and higher scores. I also reset the leaderboard do to the change.
http://games.gregthegamer.com/towerdefense/
Edit: Oh, seems I also just became a member. Sweet.
Money and Points will double the amount you get for the rest of that level. So for example, instead of getting $5, you may get $10 or $20 and so on. Doubles each time you use it for that level and then resets. Atomic will kill all enemies that happen to be on screen. This is for enemies that have spawned. And Health will restore up to 50% of your health.
LEt me know what you guys think. Game will still be just as hard, but will enable you to get to more advanced levels and higher scores. I also reset the leaderboard do to the change.
http://games.gregthegamer.com/towerdefense/
Edit: Oh, seems I also just became a member. Sweet.
Very fun! Got to lvl 10. I like the new additions, although it's pretty confusing as to what each powerup symbol does (besides the $).
Yeah, I'm working on replacing them with actual sprites to make it much easier. I presented them in the order they appear ($, #, @, &) - (Money, Points, Atomic, Health). I have a few more ideas on the game currently that I will implement soon, first being the last remaining weapon type.
I'd do something with the bombers. As is, using them almost assures you won't make it very far. They don't have very good damage and range, and their speed is too low to be of value. All sentries seems to be the way to go.
Though I did just get a new highscore I like.
RelentlessRolento
Member
so I asked before but no response, but this time I'll be more specific:
I'll be looking at RPG Maker and Adventure Game Studio (own both) and wondering which to jump into... i've never done a narrative game but want to try my hand at it. I'd prefer it to be an adventure game, but not sure how deep I have to go to understand AGS, where RPG Maker seems easier, but I'd probably not use the RPG functions of it and just use it for interaction and cutscenes, etc...
I don't write code but I do alot of visual coding (GameMaker, Unity + Playmaker)plus I can read code pretty easily... so now I don't know what I'm asking... I guess:
Anyone use Adventure Game Studio and have input on its ease for a visual script person? also RPG Maker?
I'll be looking at RPG Maker and Adventure Game Studio (own both) and wondering which to jump into... i've never done a narrative game but want to try my hand at it. I'd prefer it to be an adventure game, but not sure how deep I have to go to understand AGS, where RPG Maker seems easier, but I'd probably not use the RPG functions of it and just use it for interaction and cutscenes, etc...
I don't write code but I do alot of visual coding (GameMaker, Unity + Playmaker)plus I can read code pretty easily... so now I don't know what I'm asking... I guess:
Anyone use Adventure Game Studio and have input on its ease for a visual script person? also RPG Maker?
Thanks man. Yeah, but then, I've never marketed anything ever before, so it was a complete unknown!
Any tips on some good beginner Unity tutorials? The issue I have with a lot of the Youtube video tutorials out there is that they are so damn slow. I'd rather follow a book and focus on my own screen instead of slavishly staring at someone else's screen clicking around for an hour.
And with the official tutorials, they always show what needs to be done, but they don't seem to explain the actual function of anything or why it needs to be done that way. I find that pretty frustrating.
And with the official tutorials, they always show what needs to be done, but they don't seem to explain the actual function of anything or why it needs to be done that way. I find that pretty frustrating.
Forget tutorials. Aim to do something simple, and use the internet to find out how. Learning by doing is a million percent more effective. Just pick simple things to implement, and you'll start absorbing knowledge at a ridiculous rate.
Crits please.
Looks good to me.
This is how I learned Unity. Didn't touch a tutorial. Just dove right in and googled any time I ran into a roadblock.
Yeah. I noticed the bombers were useless. I changed them yesterday to be more useful. The upgrade power actually makes them useful now.
Originally I gave the bombers a power of "25" and added 5 every time someone upgraded them. But obviously that made them useless after 3 levels. So I changed it so it adds 25% more or so every upgrade. Meaning 25 becomes 31, then 39 and then 48 and so on (rounded). So they are now much more useful then before.
Also nice job on that score!
Edit: I'll have to rethink the "Helper", as there are times where it can freeze enemies continuously without them ever moving again. Certainly doing it's job, but problem occurs if the other bots can't even hit it. lol. The simplest solution would be to do small damage to them, that way even if stuck, can allow game to continue. (I added a temp small level power to ensure game never gets stuck, but I need to go back to the drawing board on this one)
I'm playing a round right now, the levels go extremely slow, but rocking 4 clouds and 1 helper, 415k and counting.
Edit: Also, I noticed that powerups aren't resetting on new game. So theoretically you could spend hours collecting them, then in one game unless 100 +$'s and be RICH!
2nd Edit: Levels are taking about 45 minutes now lol. Currently on level 25.
shaowebb
Member
I started prototyping in Construct 2 for my runner game and for a guy who isn't a coder its pretty perfect for what I'm needing. Lots of behavior and event stuff loaded for all the platforming elements I need with just some simple spreadsheets to edit variables on. I could honestly have a bare bones runner in a day if I used this and had my graphics, but I'm adding a lot of coin collecting and money growing mechanics to mine so it'll take a bit longer. Still feeling my way through how to spawn platforms at a decent width and distance based off of my player's screen position, but its doing okay so far.
My goal is to get it playing with jump, double jump, dash, and thats it before I jump back in on asset creation so I can at least release a simple game if I get stumped on code trying to add my other mechanics later on. I'll work on this code this week and I should have it figured out and doing what I need soon. After that its back to making more art before I go back into code to add mechanics to the core game I build this week.
My goal is to get it playing with jump, double jump, dash, and thats it before I jump back in on asset creation so I can at least release a simple game if I get stumped on code trying to add my other mechanics later on. I'll work on this code this week and I should have it figured out and doing what I need soon. After that its back to making more art before I go back into code to add mechanics to the core game I build this week.
Dunno if this helps at all or is even related, but selecting objects in your hierarchy then hitting 'F' while hovering over the scene window will snap to that object. Also, you can edit snap settings in the Edit menu, if I'm not mistaken. Will allow you to adjust how an object moves while holding the Control button.
I wish I could just die now :/ Level 26 started about 20 minutes ago, level 25 took like an hour and a half.
Edit: On firefox sadly :/ Chrome crashes my computer constantly to blue screen.
2nd Edit: Works on FF too! Woo.
I'll be taking a look at some Blender tutorials, since I learned 3D Studio Max at college, but I won't be using it for the time being since it's out of my budget. If I had the money, I'd rather purchase Unity Pro licenses before this.
So far I've bookmarked Blender's 2-minute tutorials (except for the last one since it covers lights and rendering, which don't matter in games), as well as "Making a Mecanim-compatible rig in Blender" and "ENGLISH --- Blender Unity BASIC tutorial: the tank (parenting, rotating objects)".
Does anyone have any suggestions for video and/or text-based tutorials I should watch? I mostly want to learn some basic modeling (extruding objects and drawing with vertexes), UV mapping, rigging, animating and exporting the files to be used in Unity.
So far I've bookmarked Blender's 2-minute tutorials (except for the last one since it covers lights and rendering, which don't matter in games), as well as "Making a Mecanim-compatible rig in Blender" and "ENGLISH --- Blender Unity BASIC tutorial: the tank (parenting, rotating objects)".
Does anyone have any suggestions for video and/or text-based tutorials I should watch? I mostly want to learn some basic modeling (extruding objects and drawing with vertexes), UV mapping, rigging, animating and exporting the files to be used in Unity.
- Status