Ask HN: I'm 40 and feel my mental ability declining. Programming seems harder.

Instead of porting it as-is, rebuild it with the actual functionality required and throw away the overly abstracted parts.

Oh, and I echo the other guy: make sure you have a test harness that validates everything, so you know if you've done it correctly.

maybe it's just because i listened to that podcast last week, but it sounds like you don't have the theory of the api you're porting. by theory i mean an intuition of what the software is, what motivates its existence, what it ought to do. to define the term "theory" by an example from the paper (or possibly from "the concept of mind" by gilbert ryle, referenced by "programming as theory building"): you don't have newton's theory when you can answer a question that he answered the way he answered it, you have newton's theory when your answer to a question he wasn't ever asked is how he would've answered.

so this notion of theory means, first, you can never record a theory. you can't write it all down. to have the theory means you can apply it to any new circumstance that arises, so you can't just write it all down, because you'd be writing forever. a theory is knowing how to respond. the impliction for software is that the code and documentation can't communicate the theory of the software. a theory can't be communicated at all, it has to be acquired.

so how does one acquire a theory? how'd you learn to read? how'd you learn to write? how'd you learn arithmetic? theory transmission is a social function. people with the theory show you, teach you, guide you in forming the theory for yourself.

the conclusion is that software firms should put a lot of effort into retaining people who have the theory of their software, and put a lot of effort into transmitting that theory to new people. that's not the world we live in, is it?

actually, i think that conclusion applies to any business, yet onboarding is terrible everywhere, and workers are treated as interchangeable.

what i'd be asking myself if i were you is who has the theory of this api you're porting, and how much of their time can i get?

This is so true. I'm 37 now, and I'm struggling with that - I'll come up with a project idea, only to second-guess myself whether it's ambitious or innovative enough, and really worth spending the limited time I have available. I end up questioning whether it would really count, or just serve as frivolous self-entertainment.

When I was younger, it was good enough that ideas were interesting or would help me achieve mastery. Once you have a bunch of well-worn skills under your belt and have plenty of options and the ability to make commitments, everything becomes just a function of effort/time and choosing what to do next becomes the biggest challenge.

I'm trying to cut myself some slack and allow myself some amount of time to pursue frivolous ideas. At least you can intentionally combine them with e.g. new implementation technologies (e.g. a new tech stack or new tools) to get important first-hand experience and maintain relevant skills (and therefore decision freedom).

I'm also hoping that some day, having kids might help me with some of these anxieties :-). If I revisit/retread older ground then perhaps to make it accessible to them at home or teach them by doing little projects together and give them a boost and allow the next generation to take it further. I.e., even doing non-innovative stuff could count, as a fundamentals teaching moment for someone else. Hope they take an interest!