- It was relatively late. The heyday of coding games in assembler was years prior (maybe there were some exceptions in portable platforms?). Was there any other smash hit PC game in 1999 coded in x86 assembler?
- It's a pretty substantially complex and large-scale game, at least relatively speaking. It's one thing to write a game like Tetris in assembler, RCT is magnitudes more complicated than the vast majority of games on e.g. the SNES. Doesn't mean the games are bad or anything, and there are probably exceptions. (I know the SNES in particular has a Sim City port, though it's pretty slow.)
- It's not just about assembler really, it's about the whole mindset. RCT is very well-optimized. For example, gameplay mechanics are adapted (e.g. stretching the length of months, shaping algorithms for calculating scores and ratings, etc.) around reducing the number of multiplications, and even on fairly crummy computers of the early 2000s it was possible to have huge parks with a lot of guests running quite well. Contrasting RCT2 with RCT3 paints a pretty good picture, because if you ran both on contemporary computers for their respective releases RCT3 with its fancy 3D graphics and modern development practices couldn't handle a fraction the size of parks without becoming a laggy unplayable mess.
I admit that I think people focus on it a bit much, especially since I'm not sure most people who repeat this fact actually understand what it means. But honestly, I'm willing to be arrogant enough to say I understand, and I salute. Writing scalable and complex code that actually works in macro assembler is not at all impossible, but it's certainly not easy. It requires a discipline that is not to be taken for granted.
That said, I watched the video, and while it did talk on this point, it was largely about the death of hit games from small teams and the bedroom coder.
- It was relatively late. The heyday of coding games in assembler was years prior (maybe there were some exceptions in portable platforms?). Was there any other smash hit PC game in 1999 coded in x86 assembler?
- It's a pretty substantially complex and large-scale game, at least relatively speaking. It's one thing to write a game like Tetris in assembler, RCT is magnitudes more complicated than the vast majority of games on e.g. the SNES. Doesn't mean the games are bad or anything, and there are probably exceptions. (I know the SNES in particular has a Sim City port, though it's pretty slow.)
- It's not just about assembler really, it's about the whole mindset. RCT is very well-optimized. For example, gameplay mechanics are adapted (e.g. stretching the length of months, shaping algorithms for calculating scores and ratings, etc.) around reducing the number of multiplications, and even on fairly crummy computers of the early 2000s it was possible to have huge parks with a lot of guests running quite well. Contrasting RCT2 with RCT3 paints a pretty good picture, because if you ran both on contemporary computers for their respective releases RCT3 with its fancy 3D graphics and modern development practices couldn't handle a fraction the size of parks without becoming a laggy unplayable mess.
I admit that I think people focus on it a bit much, especially since I'm not sure most people who repeat this fact actually understand what it means. But honestly, I'm willing to be arrogant enough to say I understand, and I salute. Writing scalable and complex code that actually works in macro assembler is not at all impossible, but it's certainly not easy. It requires a discipline that is not to be taken for granted.
That said, I watched the video, and while it did talk on this point, it was largely about the death of hit games from small teams and the bedroom coder.