In a previous thread here, I discovered that there were several people who were dabbling in game development on the side, just like me. It’s been a couple weeks since, how is everyone doing?
I’ve been chronicling my experience in my blog, the tabbynat devblog. Managed to get Snake done so far, now I’m working on a infinite runner - seems all the rage these days. Mechanically getting things to work is a lot easier now, since I’ve gotten more experience with the Unity documentation, but I’m running into actual game design roadblocks now - turns out, it’s hard to actually design games! You never think about good level design until you actually have to do it.
Learning about Unity, I realized that I’d have to use another program for modelling, texturing, animation and such. So I got started on Blender. It was confusing at first but I’m getting better. I can make things which are actually recognizable as what I meant to make like this Beepboom:
I’m still working on my engine, but it’s not going as quickly as I’d hoped. The main problem is that I think I’m overengineering the engine, I keep trying to solve problems I’m not having yet. It’s really a fine line, I have experience making games and one rule I know is “your first implementation is going to be your only implementation”. The second game I wrote had a terrible hacked together brute force collision system, with the triumphant comment that it would be swapped out for spatial hashing or AABBs or something “eventually” which never happened, despite it being a real problem. So now I think I’m overcompensating and going the other extreme, which is not aided by the fact that without any deadlines I trend towards the “perfectionist” side of the curve.
Some of my slow progress can be excused, I suppose, by my work on open source libraries that I’m going to use in production. Still, it’s a bit disheartening when hours of work end with no visible progress on the game.
I think my other constraint is what type of game I want to make. I should really just make some silly game jam/ludum dare scale game, but my thoughts always trend towards the type of game I “really want to” make. It doesn’t help that I started this engine with an eye towards it aiding me in research in grad school when I start, so I feel compelled to do original research AND make a game at the same time, despite this being a silly, silly idea. Not in principle, just as a first game in this language/engine.
I do have a square that moves when I move the scroll wheel over it and disappears when I click on it, though, so that’s something! I’m working on getting audio working right now, and really wishing that OpenAL was better documented. My biggest fear is content generation, I have no art experience, and every time I open Blender my brain threatens to explode.
I’m working towards doing a Ludum Dare myself, the next one is in April I think? So I’m working myself up to that level by then. Hopefully I can take a day or so off to really work on it!
Anyway, I totally found that working on silly little things really exposed just how deep the rabbit hole goes. You think jumping is really simple, but nooooo… execution is crazy hard.
It’s called blender because your that’s what it does to your brain. Good luck with it! I didn’t manage to get my head around it when I used it last, I hope you do better!
I did find it nigh impossible to get my head around to at first but by following tutorials, I gradually picked up tricks and got it, at least enough that I can make basic objects like the one I linked to.
Ehh… how about all of it? Undeciperable shortcut keys to do things, menus upon menus, and woeful documentation… All I remember was not being able to do anything, and not being able to find anything, and once I did find it, it did what I wanted to do in the most roundabout way possible.
Anyway, I’m leaving 3d to the side for now, 2d is much easier to get my head around. I like 2d better anyway.
Just checking in as a fellow hobby game coder. I actually work in a fairly niche part of the games industry, but as a backend developer. So my screwing around in Unity gives me the “ooh pretty colours” that I need to keep me happy
Nothing to show as of yet, I had planned to release something for the Candy Jam but then some kind of lung plague knocked me out for a couple of weeks. Oh and I’m attempting to use 4.3’s built-in 2D stuff, which brings with it its own challenges. One “brilliant” one I found so far is that the Box2D b2_velocityThreshold constant is a compile time option, which means Unity has a fixed value that was chosen at compile time and cannot be changed later. What I saw (and apparently others have) is that collisions at shallow angles and slow speeds don’t result in a bounce. Apparently they have an alpha version that has exposed some of these constants, but as that isn’t released I had to roll my own bouncing physics, which kind of ruins the idea of using an engine with built-in physics …
Yeah, 2D still seems kinda poorly supported. On the shallow end of the pool, there is no equivalent to the player controller for 2D. So you gotta roll your own. Luckily I found a third party script that works for my purposes, but really now. You couldn’t have implemented something for 2D?
Yeah, it does seem half-done. Is this the controller you use? Everyone seems to recommend it. I haven’t looked at it as what I am currently working on doesn’t have a player character:
Yes, that’s the one! I made a minor extension so that I can also test if the character hits his head on the ceiling from a jump, but otherwise it seems to be working fine.
So any of you guys interested in collaborating on a small project for funsies with Unity?
I’ve had some decent luck working with github and Unity project, no major issues at least so far, and with a distributed version control system it’ll be easy to collaborate at our own pace and wherever we want to apply our talents.
If this sounds like a good idea to you let me know! We can brainstorm a little project we can work together on! If you don’t have a github account it’s free over at github and the windows and mac clients make it easy to use.
I’ve personally just been getting acquainted with the IDE myself. Going through the tutorials on the Unity site one by one, so I may not be a good game dev yet, but I’ma decent programmer!
Hmm. I’d like to get a solo project done before I even think about collaborating. Besides, I really only have 1 hour of dev time a day (lunch) so I doubt I’d be very useful in any event.
Have been working my way through the online Unity videos and creating a small little play world with random objects doing random things with random code. It’s very Katamari-esque.
Generally have been super impressed with Unity and learned a lot about what it takes to be a developer. I don’t think I’ll ever complete an actual game, not because I can’t, but because it’s enough for me to know roughly how it would be done once I see the mechanics in my mind (and life gets in the way too:)). It’s really interesting to think about different types of games and how they would actually be created and coded in different ways.
Also, it’s way easier than the straight Java coding I was trying before, which was tough because I am not a programmer by training. Unity really bridges the gap nicely from straight code to game objects.
Hello fellow hobby developers! How are your projects going?
It’s slow going for me - still working on the camerafor my platformer, and it’s slow going. I’m not even completely happy with the camera as it is now - it’s basically impossible to have it track my character closely without having enormous screen jitter, either that, or the camera lags the character waay too much.
I’m thinking of possible solutions right now, but I really have no idea what to do, and I haven’t been able to find a good camera algorithm online.
I’m using a similar method right now - most of the scripts use Lerp, but I’m manually taking the difference of the x positions and multiplying it by time.Deltatime myself - the effect is the same, finding a fractional point between the camera position and the target position.
The problem with this is that it’s very slow. If you look at the Mario camera, it’s really snappy - Mario is always in the centre of the screen, whereas with Lerp type implementations, the camera will lag behind the target somewhat. I see that some scripts multiply by 5, presumably to reduce the lag - I haven’t tried such a high multiplier yet, but the last time I increased the multiplier I got jitter again. =/
