Hacker News Reader

Comment by fake-name | Hacker News Reader

I've never really understood some of the zachtronics games.

Things like TIS-100 and Shenzen I/O confuse me. Why would I spend time learning some esoteric assembler that exists no where else, when you could take that same time and learn an actual assembler?

I really, REALLY wish they had done something really clever, like targetted AVR opcodes or something for their vCPU. That'd make it so not only do you get to have fun, but you learn a marketable skill in the process!

That's one of my major complaints about most of the "programming" games I've seen, in that most of the time, if I'm going to spend time writing code, I'd rather just hack on one of my projects then play a game with no real end-product. At least with personal projects, you wind up with an actual app at the end.

I did enjoy Spacechem and Infinifactory, though I didn't finish each before becoming bored.

------

View on HN

> Things like TIS-100 and Shenzen I/O confuse me. Why would I spend time learning some esoteric assembler that exists no where else, when you could take that same time and learn an actual assembler?

TIS-100 and Shenzen I/O are theme park training grounds where a solution is guaranteed to exist. For education or for people lacking confidence, that has appeal. Also, I _wish_ I had hardware like the TIS-100.

TIS and Shenzen superficially look like assembly programming, but in reality they're very sophisticated puzzle games. You basically create these networks of routines, each of which is limited to something like 9 lines of code (and labels cost a line!). All the tasks are structured so that when you try to code them the straightforward way you're always short a line or two, and then you have to sit there racking your brain to figure out the 'trick'. Then you incorporate that new knowledge in your solution for the next level, and so on.

So it's really nothing like real coding. Highly recommended though! Especially Shenzen.

Labels and code can share a line, like this:

  ST: MOV X1 ACC
  TEQ ACC 1
  + JMP ST
  MOV X1 X2
  SLP 1

Now go back and optimize your solutions! :)

My fav part of Zachtronics games is that they show an alternative vision of massively multiplayer PVP where the metric of success is not, "Who is the best murderer?"

> Why would I spend time learning some esoteric assembler that exists no where else, when you could take that same time and learn an actual assembler?

Fully agreed. I own TIS-100 but never saw the appeal. I can write actual programs in real-world programming languages that accomplish something, thank you!

This is why I found Factorio infinitely more engaging: theme matters! The game is still an assembler/automaton, but disguised as a sci-fi RTS game. By the time you realize you're actually programming, it's too late and you're having lots of fun conquering an alien planet.

The games aren't really about the language--which have maybe 12 instructions at most--they're about the challenge of working with an extremely limited instruction set. They're puzzle games disguised as games about assembly languages. A game using a real instruction set might be fun for some people, but it would feel a lot different, and probably have to be a lot more challenging to make up for all of the convenient instructions. At that point you're basically just programming.

I am a university CS student and my microprocessor course has us learning assembly. The concepts I learned in TIS-100 are definitely making it easier to grasp how things work at the most fundamental level. It might not have taught me x86 assembly, but it taught the basics like moving data between registers and working with only the most basic instructions. Still found it very useful.

>Things like TIS-100 and Shenzen I/O confuse me. Why would I spend time learning some esoteric assembler that exists no where else, when you could take that same time and learn an actual assembler?

I think it's really more for people who don't realize they're learning an assembler. Also, it's less daunting than a real assembler because it's so much simpler.

Haven't played TIS-100, but Shenzen I/O is just a really good puzzle game. Nothing more to it than that.

I'd also add that a large segment of their player base likely has programming experience, yet no experience in assembly.

Not to mention the huge loss of a learning opportunity to have players learn programming language basics for specific real world code

This smells of when industry shows up at a college and complains they aren't teaching them "what we use in the real world". After enough practice learning multiple fictional languages, learning the n+1th one should be a walk in the park! The important part is the game is fun and encourages learning. That the developer didn't choose to use an "actual" language to do so shouldn't ultimately matter if the concepts are sound, right?

Let me prove your axiom wrong with a counterexample.

I love these games, and I can program (and have shipped to prod) in over a dozen languages including Erlang, Haskell, C++, C, Common Lisp, Java, Clojure, Scala.

What you're saying is nonsense. Practice in programming is practice in programming. If we wanted to, we could make the TIS-100 system real for embedded computing and I bet it'd be popular with some instruction extensions.

I did an annex course at a local middle school with TIS-100 to teach kids that programming is fun. It went over well.

Reply on news.ycombinator.com