I'm slower because I think more about the code I produce. I have a wider knowledge of the domain I work in, cross cutting concerns, and technical choices that I'm taking into account when producing code. So I arrive at a more-correct solution faster.
I would say the tradeoff is worth it, but of course I would because I'm not in my 20s. If faster while missing more security issues, performance problems, and corner cases is what you want, my 20 year old self is probably better.
So it can look slower up front, but over the span of the entire task, it's faster.
I guess what I'm getting at is it's sometimes hard to untangle what changes are due to getting older, and which ones are due to the world changing. For me it's boredom, I used to have long boring summers, and get bored on weekends in the 90's and early 2000's. I haven't been bored in years. Is this because of the internet? But I'm not bored when I don't have access to the internet now, so maybe I just enjoy some peace and quiet now I'm older.
I wrote a small program[1] to help me maintain context between different projects, and I'm now more productive than I was in my 30s.
[1] I tried all the personal task management and task tracker tools, even wrote a few of my own. It turned out that my brain works differently to how the tools want you to work. My new tool which works well to keep me on track matches how I work.
IOW, now I don't have to adjust my brain to the tool, the tool is already adjusted to my brain's process.
Believe it or not, I actually named the executable 'frame'. However I'm unwilling to share it[1] until I have a good 5m explanation (it's all shell-based).
[1] Well, announce it, anyway. It opensource anyway.
I'm conflicted; on the one hand I would like to let the world know about it, but on the other I can see how it would be easily dismissed as useless when there's literally no docs about it.
I think I shall make a small example usage (terminal only) to demonstrate how it helps me, then post a show HN tomorrow or Saturday (depending).
I'll reply to your post again once I post the show HN.
It feels like a shift slightly away from "build the thing right" and toward "build the right thing". There's no sense racing to a destination that we don't want to arrive at in the first place.
But I know I can force myself through hard stuff (prolog, category theory, ML research papers) but only if I can kick out instant gratification for a while (reddit, hn, videogames).
I remember this week more than 10 years ago where my graphics card died, and had to go with this ancient card with zero acceleration, and no videogames. I was surprisingly productive that week.