2026-11-14
Growth: The Willow I Never Meant to Plant
At the end of last month, I used AI to stitch two existing open-source projects together into a classical poetry reader with pinyin, which I called Yinshi. You search for a poem, every character gets its pinyin, and you can print it out, which makes it handy for reading with kids. From idea to running on my machine took less than an hour.
One evening a week later, it suddenly started picking up stars. Forty by the next morning, sixty-eight by that night, and over eighty now. Before this I'd built plenty of projects, some reworked over several weeks, and not one of them passed ten stars.
That afternoon a line slipped out of me, an old saying: the willow you stick in the ground without a thought grows into shade, while the tree you plant with care never takes. This week I wanted to think that line through.
Effort is something you tell yourself
The projects nobody looked at really were built with care. The features were thought through, the docs were long, the details tuned again and again. But all that effort measured how hard I worked, not how much anyone else needed it.
Someone opening a project doesn't know how many rounds of revisions went into it, and doesn't care. They ask one thing: is this useful to me? Yinshi answers that in a second. Anyone with a kid memorizing classical poems can glance at a screenshot and know what it does. Most of my earlier projects needed a paragraph of explanation before you understood what problem they solved, and the problem was often one that very few people have.
There's nothing wrong with effort itself. It just makes it easy to assume that what feels important to you will feel important to others.
A low barrier beats polish
Looking back, the difference wasn't how well things were made but how hard they were to use.
Yinshi is a web page: click and it works, with nothing to install, no sign-up, no technical knowledge needed. Many of my earlier projects meant setting up an environment and configuring options first, sometimes even knowing what a particular tool was. Every extra step filtered out another group of people. The ones who made it all the way to actually using it were people much like me to begin with.
I used to spend my energy making things deeper, when I should have lowered the barrier first. An ordinary tool anyone can use often goes further than a refined one only a few people understand.
Exposure can't be arranged
The bigger surprise was where the traffic came from.
Almost none of my own promotion worked. I submitted it to Ruan Yifeng's weekly newsletter and it wasn't picked up; I posted about it on Xiaohongshu and got next to no views. What actually took off was GitHubDaily sharing it one evening: the post got nearly ten thousand views, and the stars started climbing at eight o'clock that night. I didn't know it was coming, and I had never contacted them.
It made me admit something: whether a thing gets seen is largely not up to the person who made it. What you can arrange is making it and putting it out there. When it gets seen, and by whom, you can only wait.
Planting with care isn't wasted
Still, the reason the willow could go in so fast was all the time I'd spent planting trees.
Because I'd built so many projects before, I knew how to connect two open-source projects, how to handle the licenses, how to deploy, and how to write a README other people can follow. Without all that seemingly wasted groundwork, Yinshi wouldn't have taken just an hour. The trees didn't grow, but the skill in my hands stayed.
So I'm not concluding that effort is useless. The more accurate version is that effort decides how fast and how solidly I can build, but not which thing will be needed.
What I'll do next
Once I'd thought it through, I set myself two rules.
First, plant more willows. Since I can't know in advance which one will be needed, keep each attempt cheap: make a version an ordinary person can understand in one sentence and use with one click, put it out and see how people react, rather than spending weeks polishing in one direction with my head down.
Second, water whichever one takes root. Yinshi has already shown that people want it, so the next step is to help it grow: fix the issues users raise, watch how many people use it each day and where they come from, instead of turning straight around to plant the next tree.
Which willow took root this time wasn't my choice. What I do with it now that it has is.