At first I had a C64 and basic and only the manual. I could not do much with just a listing of command and their use. You had to guess how to construct a progra

At first I had a C64 and basic and only the manual. I could not do much with just a listing of command and their use. You had to guess how to construct a program from commands, as if you try to learn a language by using a dictionnary. Not very efficient. But, I did repeated the begin and stop experience with «learning the language from a dictionary» with a casio calculator (422bytes code) and a HP48SX. I dare say that I went way further with the HP, it was an awesome start at learning programming, because it had all these data structures that where close to being functional without the burden of being functional crap. My first try at ASM was in the 1980's a «geek» told me it's easy : he gave a floppy with an assembler and said to deal with it, that's all you need. Second time 5 years later a coder gave me a floppy with pascal and nothing more. And I was demotivated. Then, I went to university, I had to achieve results in programming with less than useful training (just C with the difference between pointers and integers), and it went better. It was code or you'll not be able to graduate so f*cking find a way fast. Finally, in my master degree in physics I had an obscure option for interface programming that was in fact 68000 asm programming. And I had a COMPETENT teacher that explained us HOW computer worked, and HOW the code was working. It unlocked so many keys in my brain I understood stuff ranging from OS, to C, to ASM to even matlab mechanism. Learning like in sport is made of steps and stagnation. And people often think it is a bad thing. I think it is human brain normal behaviour, like strats of learning putting each others on top of one an another. I also would like to spot in the era of «online is all you need to self learn anything» that it is bullshit crap: knowledge does not jump from internet to your brain. A good «craftman» teaching you his craft with the intuition that comes with is like a catalyser. Long story short my 15 first years of learning are made of 4 distinct steps of «stopping to learning code» and I think they were necessary, like both a trial of faith and a necessary evil.