Hacker News Reader

Comment by ThrowawayR2 | Hacker News Reader

Every time I think about playing those games, I can't help but be dissuaded by the likelihood that learning actual machine language for ARM or x86-64 would be considerably less difficult and, in the long run, more beneficial to me.

View on HN

Learning an actual machine language is definitely harder, all of these games use extremely limited instruction sets and/or simplified operating conditions.

As for more beneficial, I'm sure learning a real practical machine language is more beneficial. However you can definitely draw some of the same lessons out of these games and depending on you, might have a lot more fun and chances of actually going through with playing these games.

Harder indeed (in some ways) but the raw satisfaction of booting your first toy OS on bare metal is unmatched.

I enjoyed writing a DCPU-16 emulator very much though.

https://github.com/lloeki/python-dcpu_16

You just reminded me that I wrote a stack language for DCPU16 back in the frenzy. It gave me an excuse to learn the IronMeta parser generator, which is an implementation of OMeta. Unfortunately I kept it on Codeplex and never bothered to move it since the entire idea was dead by that time.

https://en.wikipedia.org/wiki/OMeta

https://github.com/kulibali/ironmeta

https://archive.codeplex.com/?p=stackr

I got about halfway through TIS-100, realized that what I actually wanted to do was to finally learn assembly language, and actually went ahead and wrote a working Fizzbuzz program in x86 assembler! Really the best outcome for not finishing a game

Just keep in mind, it's very likely you will never use ARM or x86-64 either and, even if you do, you'll probably have less fun using it.

This is nonsense, if you write natively-executed software you'll definitely make use of being able to read assembly in a debugger.

I totally get what you're saying. For x86-64 I recommend Ray Seyfarth's book: http://rayseyfarth.com/asm/

But I can't help suggesting you try TIS-100. It's genuinely enjoyable and not like regular assembly at all.

Why not both? Unless you use ASM every day, you can learn a very limited subset (basically enough to reverse engineer a fizzbuzz) and get some benefits from that. Maybe look through what sse and other extensions provide. That doesn't require a lot of time.

Along the same lines, why spend all that time on this game when you are a programmer in real life! I look at games as a change of pace / relaxing not more programming. ;-)

I can play the real guitar but I still enjoy Guitar Hero.

In the case of TIS-100, I don't think the challenge of the game is 'write assembly' as much as it is 'deal with the somewhat crazy architecture and its self-imposed restrictions'. I've programmed some 6502 and x86 assembly in the past, but it didn't look anything like a TIS-100 program, which is more like a ridulously simplified transputer network with memory-less nodes.

for the same reason that airline pilots build their own experimental aircraft, long-haul drivers jump on a Harley and cruise around town, and machinists who make neat things at their home workshop -- freedom, excitement, the joy of exploration, the joy of working with the stripped-down essence of their expertise. TIS-100 really is different enough to be challenging and enjoyable, plus there's a neat story within.

an interesting thing about exapunks, is to get the best scores you need to leverage parallelism well. So you will at least learn a lot of fundamental things that are useful there, and that is a very relevant skill set on modern hardware.

Probably not less difficult but definitely similar. I tried TIS-100 but it just feels like it exercises the part of my mind that does programming. It’s not relaxing or unwinding (for me).

Ahh... so Human Resource Machine and 7 Billion Humans are closer to your likings?

Reply on news.ycombinator.com