I’m personally investing time to level up my proficiency with digital content creation tools — Houdini, Unreal Engine, Unity. They mostly only apply if you work in games or visual effects, but they’re a blast to play with even if your day job is in a completely different industry.
I similarly have difficulty learning a concept when it's presented using real-world metaphors that aren't often actually expressed via code. This is how object oriented programming was taught to me: imagine a door object, which contains methods open and close, and it contains a knob property, which is its own object with functions to turn left and right, and so forth. Once you're familiar with OOP this might seem like a good way to explain the concept, but without a specific application to apply it to, it's difficult to see the purpose.
OOP examples such as these and their instruction in traditional schooling is arguably one of the biggest failures of computer science education of our time. This is literally the last thing object oriented patterns are ever used for (depicting real world objects as members of categories). They are primarily designed for describing systems that are managing the user's interactions in the software rather than representations of real-world physical objects.
The only time these overlap is in video games, which is usually not used as the context for teaching software development when OOP is taught.
It does a great job of explaining the underlying ideas behind different databases and data processing systems. Even if your needs are met by, say, a traditional RDBMS like PostgreSQL, it’s helpful to know about the alternatives and what problems they solve. It’s the sort of book I wish I had read years ago; I’m sure you’ll be happy you picked it up.
It’ll be a shame to lose gems like Escape Velocity. I know several people (myself included) have asked about acquiring the rights to the games but received no reply. It appears Ambrosia is indeed defunct.
Build more long-lasting artifacts. As a software developer that has primarily worked in startups, most of which no longer exist, it can be frustrating to have nothing to show for the hours I’ve worked. I envy a friend of mine who works as a 3D animator on feature films who can point out the exact shots he’s responsible for and always has a permanent record of his efforts.
I recognize that software inherently lacks the shelf life that mediums like film enjoy, but I plan to at least partially achieve this goal by writing more often, contributing to open source more frequently, shipping and maintaining side projects, and striving to make the company I work for a success — steps that will give me something concrete that others find useful / enjoyable.
There is an additional issue to be faced here. I did a lot of open source in my teens and early 20s that's still sitting on SourceForge, but the last decade+ of my work is locked behind proprietary walls as well.. So while I have concrete things to point at, I wrote them all 10-15k hours of experience ago and they're all terrible and use long obsolete technologies (Borland Delphi, oh how I miss you..).
Unreal Engine and C++. I've long worked with Unity, but as part of a new job I'm tasked with developing our Unreal plugin. Previously I only touched C++ on occasion, so I had a lot to learn — and have a lot to learn yet — of best practices, new features available in C++11, dealing with exceptions (Unreal disables them by default), and so forth. Likewise for Unreal. Like C++ itself, it's wonderfully powerful but sometimes painfully complex.
I also continued to deepen my understanding of databases and distributed systems. My favourite read this year was Designing Data-Intensive Applications which made me more familiar with the pros and cons of the various datastores and provided a better sense of the tradeoffs that each makes. It also gave me an appreciation for the guarantees that the battle-tested relational databases provide. One of my goals for 2019 is to improve my SQL knowledge — thus far any extra effort to understand it better has payed dividends.
On the verge of choosing Unity for a project with middling performance requirements. Can you tell me which platform you'd choose if you had the luxury?
Almost without question I'd choose Unity for most projects. Its rich ecosystem, large community, and decent documentation make it a solid choice. Many of its warts like only supporting an ancient version of C# have been remedied in recent releases (as of Unity 2018.3 you can write C# 7.3, which is a nice language indeed). Unreal's Blueprints are slick for gameplay scripting, but I'd choose C# over C++ for game code any day.
In the end they're both excellent engines, and I suspect a AAA game would be better served by Unreal, but I find Unity far more flexible for arbitrary game designs whereas with Unreal you sometimes feel like you're fighting its roots as an FPS engine (though if you're making an FPS, UE4 might be just what you need).
Thanks for the links, those look promising. After some searching I often saw TypeRacer (https://play.typeracer.com) recommended also. I tried a few other typing tutors but found that few emphasize symbols like {}*&#$ that I use frequently and are responsible for most of my mistakes.
