December 23, 2017

2017 In Review

This year has been a busy year. While it feels significantly less productive than the previous year about the same amount of effort has been put into my ongoing project. I want to keep this post short but realistically there isn't a ton to talk about and a heck of a lot more to get done.

My first part of the last quarter was focused on reworking how terrain generation is going to work and get it capable of handling faster generation that the previous version I had. It helped my point out a few bugs I had in the previous version so it was very much worth the time. I've been working on my physics engine for most of the latter part of the last quarter with the odd piece being done for the actual project itself.

I have come up with a robust and efficient solution to both physics and graphics that I am currently in the process of developing out the implementation. Nothing to really show as everything I've been doing is extensively unit and performance tested to make sure I'm not missing any details.

At first, I thought my solution was going to allow for extremely fast physics, but the simple truth is that I have to compromise between accuracy and robustness. On the upside, I have a moderately efficient solution to solving complex convex polyhedral collisions. One of the challenges has been making it robust without comprising too much on performance. The solution I use is an approximation and by putting some limits on the physics engine I can make it stable and fast. The cost, however, is that I need to be careful with edge cases because it is only an approximate solution. I can very easily get myself into a situation where I get completely garbage solutions which would ruin the experience. The exciting part of what I have done with the physics can also be used by the graphics part of the engine which should allow me to scale up some of the neat lighting effects I plan to make use of.

The plan for the next quarter is to finish off the physics engine and get all the parts of it working smoothly and integrated into the game engine. Then I'll be able to focus on getting the graphics engine switched over to the new system.

All in all, I'm a little disappointed with where the project is at after a year of deciding to swap out the physics and graphics parts of the game. It certainly feels like a few steps backward, but each step forward I take now seems to be on a more solid foundation than before. However, that's about 10s of thousands of lines of code later.

October 12, 2017

The Rewrite

There has been a very clear lack of posts this year, but not because I haven't been busting my butt. In my last post, I outlined what I wanted to accomplish and by what sort of time frame I was hoping to make it all happen. For the fun of it, I haven't looked at that post since I posted it in December 2016 and I feel like I can take a guess at where I wanted to be by this point.

I believe the main thing I was hoping for was getting a custom physics engine up and running, and I had debated reworking the graphics engine. I think I was aiming to get a prototype going by the end of summer. (Something I certainly haven't been able to do). Now I'm going to take a look.

Well, I certainly promised a few other things that I haven't finished yet. I also dramatically deviated from my outline. I think maybe the best approach is to outline what I have done in the last almost 10 months.

This one was a big one, I knew going into it that I would be going deep into a very technically complex topic. There is a huge difference between knowing physics engines and actually writing one. It was a challenge, which I took on and am very happy with my result. I have a very unique physics engine that should be able to do everything I want. It is extremely robust and has absolutely no issues handling high-speed collisions flawlessly. I also have a nice automated test suite to make sure all the little parts tick correctly. I still haven't really built out the part that I originally intended for my game, but it isn't a high priority because have solved most of the difficult technical challenges.

As I had expected, I wasn't going to be doing a lot with client-side graphics. However, what I ended up doing was completely ripping out my old graphics and rewriting it and an engine from scratch. I was able to come up with something I believe can do high-quality real-time shadows. Time will tell if that truly ends up being the case, but it would be wonderful if it works out. I suspect it will be based on a few prototypes but we will see. I'm using a completely different rendering technique that I haven't seen used by anyone else as well.

The motivating factor for doing my own graphics engine is because of the art style I've decided to go with: primitive object graphics. That doesn't mean low-polygon but it might have the look of it at times. I basically want to mix high-quality lighting with basic shapes to give it a very unique feel.

I was planning to upgrade my graphics, what I wasn't planning on doing was reworking my integration with CEF. I ended up reworking my build system and dramatically simplifying it. I am now also able to make use of bleeding edge web technologies and have the ability to easily write automated tests anywhere in my software stack. I also have opened up the ability to debug every layer, which is going to be extremely useful for development. Effectively there is no guess and run, everything is testable and debuggable. The build system is straightforward and very easy to work with. Most of my frustrations with having a complex build process was having to figure out what went wrong and how to diagnose issues. Now everything is easy to debug and test.

A while ago, I was using a number of libraries for networking, but they lacked customization and seemed to carry a decent amount of overhead. Seeing how I basically reworked most of the engine/game I decided to rip out the network libraries and roll my own solution. Instead of going all low-level I settled on using Netty as the backbone for my implementations. I was able to whip up a TCP client/server, an HTTP server, and a WebSocket server in a few days. Each customized to match my needs for the engine. I now have a proper authentication layer and security around the client/server as well as considerable control over how everything operates. The best part of it was I basically tossed four libraries and now use just one which does everything I need and more.

Slightly unrelated, but I ended up extending my own linear algebra library because I was sick of dealing with terrible API and not being able to distinguish between immutable and mutable data structures. I even included a few useful types to deal with special cases and basically make the type system do correctness checks for me when performing calculations. All that dramatically simplified my work on the physics engine and made writing unit tests a breeze. I also wrote an arbitrary precision fixed-point library. This allows me to work at whatever precision I need, and I can use fixed-point for consistent number handling where it makes sense. I'm still using floating point for internal parts of the physics engine because it is just faster. I'm also using floating point for the graphics engine because that is just faster when dealing with it. However, I can now store and operate on numbers of any scale when I need to, something kind of important when working at solar system scales. Unfortunately, it takes up to 3 times longer to perform some calculations compared to a floating point but at other times is just as fast as floating point. Enough I won't use it for core calculations but I certainly can use it for intermediate values when transitioning between coordinate systems. It was something I was attempting to figure out previously, but now that I can scale to whatever precision I want and still run at fairly high performance there isn't any concern now.

After all the previous work I've finally been able to get back to working on the game itself. I spent a ton of time working out engine design and am not able to crank out features related to the game. It is extremely easy and I've come up with an architecture that looks to be robust and extensible.

I now am also using my own wiki for keeping track of architectural and development information (such as how to build the application). I wasn't sure what I wanted to do to keep track of it but figured it is better to use a real wiki with markdown over text files in the repo. I'm also recording and planning out my designs and thoughts in OneNote. It seems to work well, but I'm split on if google drive would simply work better. I'm also using a proper issue tracker now and making cards/organizing my backlog for cards. It really is just a glorified TODO list at this point, but it really helps me stay focused and provides a nice way to show progress.

Based on my outline from before, I look like I'm roughly on track. Behind on a few things, but also pulled in a lot more than I was expecting. I will be continuing to work towards a playable demo (technically I have one but its nothing special) and as I work on it I'll be continually improving the physics and graphics engine. To the end of the backlog I go, one card at a time.

December 27, 2016

Looking back at 2016

I figure with the end of 2016 coming closer it is time to look back at the year. One of the major differences this year has been my ongoing project. Not only has it taken up a large portion of my time, but it has also been a huge learning experience and refresher on things I had forgotten from my undergraduate engineering courses. I also have come to release that frequent updates to this blog are just not going to happen. It isn't because I don't want to post updates, but I tend to prefer working on my project than posting about it.

Near the end of 2015, I wanted to get back into working on my own games. Shortly after finishing my master's program I was just mentally and psychologically exhausted. I had spent a significant amount of time playing games with a few friends and started to get an itch for making my own game. I was originally thinking to make a mobile game that I would want to play while taking transit. Something I could easily pause but also be complicated and requiring a good amount of problem-solving. At the time I had spent many hours playing Factorio, and so I was wanting to create a tower defense which required you to collect resources, manage power/communications, and fend off attacks from a moderately intelligent adversary.

I got a very basic prototype done after a few weeks, but I started to envision a much bigger project. I was no longer taking transit simply because driving took about 1/5 the time. Together both of those factors made me change my mind. I began looking into technologies and libraries I could use to make it happen. Most of my development had focused on 2D but I had done some basic 3D work. In the end, I decided it would be worth my time to look into Unreal 4 vs Unity 5 as my experience with both engines had shown they had extremely robust rendering engines that would easily outdo anything I could create. They also had the nice feature of being able to support multiple platforms. The end result was my decision to go with Unreal 4, as the performance as my rendering engine exceeded what Unity 5 could deliver.

After a few weeks of hacking together the general concept for what I wanted to create I through together a basic prototype in April 2016. Which you can take a look at below:

It was very simple but gave me the foundation for what I was going to build. As time went on I slowly started to build up a robust engine which delegated the rendering to Unreal. I still use the approach today and surprisingly it allows for rapid iterations and excellent utilization of the hardware.

I had a short and bitter fight with Unreal's GUI frameworks in May of 2016 and decided to not bother fighting with them. Instead, I decided to stick a web overlay on top of the game canvas and use web technologies for my interface. On the surface, it seems more complicated, but it turns out to be significantly simpler and more productive. It also means I can create really good and solid looking user interfaces.

Everything in my game is built to be networked and scalable. Over the summer I worked on pushing towards planetary sized worlds. Optimizing and reworking terrain generation, and pushing the limits of what can be done with today's technology. Eventually, I got terrain generation to a point where I could very quickly create terrain on the fly and travel at high speeds with fairly minimal lag. This is when I started to notice the limitations of my physics engine, more on that later.

I then started working on making the terrain modifiable in September of 2016. I was originally planning to make use of an embedded database engine but I quickly came to realize that a database just isn't fast enough. That isn't to say the databases I tested were not high quality, or that they couldn't scale. It was mostly the fact I wanted to make them store and retrieve data with extremely low latency. I also didn't need to do a significant amount of querying so most of the advantages of a database were not helpful. I ended up writing my own storage system which allowed me to handle high levels of concurrency with low latency by keeping fragments of data in memory and using an asynchronous event-based approach. It worked really well and was straightforward when it came to implementing replication across clients. Something I only partially implemented, enough to prove it would work but then something else caught my attention.

The physics engine was causing most of my headaches. Again, not the fault of the physics engine. It was doing a great job for what it was designed for. It just wasn't scaling well and caused a massive amount of headaches and complexity in my engine because I was having to delegate the game structures to something the physics engine could work with. I was also suffering from issues with high speeds and the occasional terrible collision response handling. I also had to write my own 'fix' for issues with tunneling because they solutions provided by the physics engine only worked most of the time. I could consistently break it.

I had written my own physics engine for a 2D game many years ago. I was able to handle 17000 objects colliding at 60 fps using a few strategies I developed. I discovered the names of some of the strategies a few years later. It sort of hit me, there are no 2D physics engines that could handle that many objects and I have doubts any will be able to in the near future. Part of the reason is they are generalized solutions and can't make certain classes of optimizations in the problem space. Their generalizability makes them inefficient, which isn't a bad thing, it just means they are only limited to micro-optimizations. Such as using SIMD instructions, GPU acceleration, newer hierarchy structures, or newer algorithms. Whereas if I wrote my own I could tailor it to my application and only make the physics engine do the work that it actually needs to. I would also be able to deal with high-speed collisions and tunneling without having to fight against the physics engine.

After reading a number of books on the subject, a few research papers, a thesis, and looking at various open source physics engines, I started thinking about how I could integrate a custom physics engine into my game engine. It was now December of 2016, but I wasn't ready to hook it up and swap out my existing physics engine interface. I wanted to start building a prototype physics engine. Very quickly I also realized that I would need a visual debugger to help me develop the engine and come up with test cases which I could use to verify correctness as I switch from naive implementations to heavily optimized solutions. I slapped together a quick visual debugger and was able to see my simulations. I also made sure I could easily step forward through time to make sure everything was behaving correctly.

It sort of bothers me. I originally wanted to use a bunch of frameworks and libraries to make my life easier and allow me to develop the game faster. Instead, I kept having to deal with limitations and find workarounds for them and noticed my productivity decreased. I do have the fortunate benefit of not having a deadline and so I can take the hit of writing my own solutions where appropriate.

This year has been a huge learning process, a year ago I would have never dreamed of writing my own physics engine or writing something instead of using a database for storage. I certainly would have thought by this point I'd have something playable instead of a sort of early tech demo. I guess that is the problem when you have to build your own technologies to solve problems others have yet to solve.

Well enough about the past, where am I going next?
Short-term (Q1 2017)

  • [Physics] Expand capabilities of the physics engine, such as supporting more shapes
  • [Physics] Deal with high-speed interactions and tunnelling
  • [Engine] Switch from existing physics engine to custom physics engine
  • [Graphics] Finish implementing client-side terrain updates

Medium-term (Q2-Q3 2017)

  • [Graphics] Determine what type of art style to use.
  • [Physics] Optimize custom physics engine including special handling for game
  • [Gameplay] Start working on core gameplay
  • [Interface] Revisit UI with updated technologies

Long-term (Q4 2017+)

  • [Gameplay] Get playable demo
  • [Graphics] Polish rough areas