On Wheel Making
I just found this article and I think it's pretty amazing. It sums up my ideas of wheel making well and touches up some new points well. With that in mind, here are some of my own points on wheel making and other DIY things.
>> Wheel make for fun. And/or profit. And/or learning. But usually not for a specific use-case. Generally, if you need to reinvent a wheel for a specific use-case and there's no other purpose besides the specific use-case, I would recommend getting someone else to adapt/reinvent the wheel rather than doing it yourself. In other terms, if you need an artisan wheel for your artisan cart, just get a wheel-maker.
>> As stated in the article, scope out your project, or be deliberate. Don't go for a big project all at once, go step by step. Don't try to build a forest, just build a shrub. Don't try to build a car -- try to build a wheel first. And if you really, are going to build a car, remember that building things is hard and takes a long time; and thus go prudently and deliberately step by step instead of all at once.
As an example from my life, last summer, I remember I wanted to build a NN (neural net) chess bot that would be hosted on lichess (github). And that turned out badly -- partly because I was motivated by motivation and not discipline, and partly because I just scoped too damn hard. The finish line was 3 miles away and I stopped after like 300 yards. Trying again, I descoped heavily and just tried to make tetris work in the terminal (github) -- no algorithms, no lichess, nothing fancy, just being able to make a chessboard work in python in the terminal. And with that descoped, simpler, more concrete project, I was able to mostly1 finish the project in a week.
>> Document your progress. After making your wheel, you'll probably remember the process for life. But if not, documentation can help you make that next custom wheel more easily. Documenting also helps others (if you open source it) look/learn/improve on your wheel too.
>> Start over; there's no need to salvage lost wheels. Say for example you're making a physical wheel. And it's going poorly. You've got the wrong thickness rubber, a bent up rim, and so forth. Well, since you're making the wheel from scratch anyways, why not start over?
It's not like your knowledge diminishes -- starting over merely moves the starting point backwards, not the progress (and the learned knowledge). Thus, if you find yourself debating whether or not you should abandon or salvage a (figurative) wheel, I would always lean on abandoning it and constructing a new one2.
Enough reading. Go make your own wheel.
In the future, I'll hopefully take smaller steps and eventually finish this up.