5 ms·
I like the "put it aside until the next wave comes" approach. I have many hobby projects, ideas, partly-built things. I have some that are very near to comple
by __d 4y ago
I like the "put it aside until the next wave comes" approach.
I have many hobby projects, ideas, partly-built things. I have some that are very near to completion, and others that are probably a lifetime's work.
I tend to work on them in short stints: a few nights, sometimes like a month or two. And then set them aside as I burn out on that idea, or come up with a compelling extension/revision/use-case/etc for a different one.
It's really crucial to document where you're at in this model: you need to be able to pick things up again without too much time to figure out where you were. I have two things that help with that:
Firstly, a README.md file for each project which includes my roadmap, with stuff crossed off as I get it done. So I can pull it up and see exactly what's next and what I just did at a glance.
And secondly, I keep a (large) journal.md file, divided by date, where I write down what I did. Especially stuff that I needed to research to figure out. That includes literally copying command lines out of my shell and into the journal. It's just a massive bulleted list, but when I restart something, I can quickly search back to the date of the my last commit, and see exactly what (and how) I did the things I did in that burst.
Combining my roadmap/README, journal notes, and Git history generally means I'm productive again in about 15 minutes.
- mronetwo 4y agoThis is how I go about my projects. I can't say I'm at peace with this approach though. I feel like context switching makes it that I won't finish anything during my life time. Anyways, documenting projects for yourself is very important. Treat your future self as best as you can. Think about the problems your future self my encounter.