Fulfillment in Gemini
2025-07-01
---
Christian of AuraGem, who has developed many excellent services for Gemini, shared today that he is feeling burned out with all of his projects.^ He lamented that Gemini has largely stagnated; very few developers are active in the community, and even when they develop new tools for Gemini, very few people use them, if anyone at all. This leads to boredom, frustration, and overall exhaustion.
I can't say much on the current state of the protocol. It is still fairly active, but it hasn't grown much in the last 2 to 3 years. The community consensus seems to be that after the COVID-19 pandemic subsided in 2022 and people returned to the outside world, Gemini lost its momentum. I broadly agree with that. However, I do think that many people who came in 2022 did expect Gemini to supplant the Web in some way, and when the protocol failed to do so, they left.
In late 2021 I launched a game on my capsule: Chess Over Gemini. My goal was to provide a game for Geminauts, something that was lacking in Gemini at the time (this was before the days of Spellbinding, Where in the World, and Farkle). I first wrote the service using shell scripts and flat files, before rewriting everything in Python and implementing SQLite as the backend. Designing the service and getting it into a working state was a lot of fun.
There was only one problem. I'm not good at chess, and I have almost no interest in playing it. This meant that once my chess service was built and launched, I didn't interact with it at all, because I didn't want to use it myself. I could have enticed users to join by playing games with them as they signed up. But I don't play chess, so I didn't bother; players languished while patiently waiting for other players to sign up and begin matches with them.
Because I didn't use the service, I didn't catch when bugs occurred--and they were frequent. The service ran or about a year and a half, and at least forty percent o the time, there existed major bugs that largely rendered the service unusable. Even worse, because I had no interest in chess, I quickly lost interest in trying to fix the bugs--it became a chore that I dreaded when I received an e-mail that something wasn't working. The bugginess of my chess service was even immortalized in a video shared by Techrights.^^
Meanwhile, I started building other services on Gemini. I created a CGI that provides daily weather forecasts for a given US ZIP code. I built out a todo.txt service for managing todo lists. I made scripts for playing with twisty puzzles. I even implemented a mirror of the Online Encyclopedia of Integer Sequences.
I didn't run into the same motivation issues I did with chess. When I find bugs, I'm eager to fix them. I like building new versions of them and trying out different features. What's the difference?
I think the difference is that I built these services for myself first. Before I leave in the morning, I pull up the weather service on my capsule and check the day's outlook. When I need to remember a task, I add it to my todo.txt service. If I'm reading about a math problem and it references the OEIS, I check it over Gemini, not the Web. I hope these tools will be useful to others, but primarily, I'm the one who directly benefits from their existence.
Building services for myself has another implication: it doesn't really matter to me if other people use it or not. If I'm the only person who ever uses my todo.txt service in the entire history of our species, it will still be worth having built it, because I got use out of it. It's there for Geminauts to explore, and I hope they find utility in it as well, but if I use it and I like it, that alone justifies its existence.
Chess, on the other hand, was written for others first. That's a noble goal, but if I have no personal interest in it, it's hard to motivate myself to support and develop it. I'd like for other people to get use and enjoyment out of it, but I won't want to spend time on it. It also means that my satisfaction with the service is based on how much other people use it--and on a platform as small as Gemini, I can't count on that.
Gemini is somewhat stagnant these days. I don't deny that. And it's true that if there were more practical services on Gemini, it would probably be much more popular as a protocol. The same could be said for Gopher too. But as a capsule operator, if I build a service and expect people to use it, I have to be invested in what I build. If I'm not, then everyone's experience will suffer.
I got a few inquiries about the chess service when I decided to close it, and a few people even posted publicly that they were sad to see it go. But it just wasn't in me to continue operating it. Two developers reached out to me asking if I'd be willing to share my source code with them. If they'd like to develop a chess service of their own, I have no qualms about it. Gemini always benefits from having more to do, and if others are more motivated to host chess than I am, all the better.
In the meantime, I try to remind myself every day that Gemini is a small, transient place. I could spend days, weeks or even months working on a Gemini project, and no-one might use it at all. If I personally like it and get utility out of it, that's all that matters.
^ Bored and Exhausted
^^ 01.14.2023
---
Up One Level
Home
[Last updated: 2025-07-01]