Saturated Fat Controversy

[GrandizerII]

One of the first things we were told in our Digital Electronics classes was that AND, OR, NOT and XOR were considered basic ("osnovni") gates because they are implemented with 6 transistors, except for NOT which is implemented with 2 transistors. And, frankly, I am not interested in going much deeper than that, it's quite unlikely I'll have to do something that low-level.
That's not really the same thing. If you mess something up with UX, you will probably not immediately notice that, because UX is a softer part of computer science. Making a compiler or a PacMan game playable on smartphones is a more rigorous part of computer science: what you want to accomplish is quite well-defined and you can't reject it because you subjectively don't like the way it looks. Fine, maybe you can say you don't like GCC and that you like TCC more because GCC does many optimisations which slow down the compilation and that don't make the Assembly it produces significantly faster (hardly any programmer would agree with you, but it's not outright false), but you can't deny that GCC works just because you don't like something about it. And you can't deny that my compiler works, since there is evidence that it works (it can compile QuickSort written in my own programming language). Is it low-quality? Sure it is, the Assembly it produces is a few times slower than one produced by GCC for an equivalent C code, and the compiler itself runs slowly (running on top of the Duktape framework). But, frankly, I don't think you know enough about that to even evaluate those statements. Show me a compiler you've made, and then we can talk about your criticism of my compiler. Otherwise, you have no credibility.

When one can write machine code directly in native binary on multiple platforms, one learns that compilers are for the weak.

Do you still do this manually? Sounds like a lot of effort to do. That's like someone still doing trig tables manually to this day.