Why do even Gemini enthusiasts argue about Gemini?
Gemini space is usually a quiet, orderly place but, every so often, a rash of arguments will break out about whether the Gemini protocol is all it could be, or whether it ought to be modified or extended.
But a key, deliberate feature of the Gemini protocol is that it can't readily be extended -- not without breaking all the existing software, anyway. This non-extensibility is one of the ways Gemini distinguishes itself from the regular web. I doubt that anybody who uses Gemini wants to see it turned into an advertising and telemetry platform, which is what the mainstream web has become.
This month seems to have been a particularly fractious time for debates about the future of Gemini. Here's a sample.
What's significant here, I think, is that these discussions are taking place using Gemini, so it's safe to assume that all the participants are Gemini enthusiasts to some degree. Some are adamantly opposed to the slightest change to Gemini. Some would like to see radical change. Others, I guess, fall somewhere in between.
The Gemini protocol and Gemtext document formats have been defined for six years, and scarcely changed in that time. So why do we keep having these arguments? Why aren't folks just accepting that Gemini is what it is, and either living with it or adopting something else?
This is not a technical question, and I don't think we should be looking for a technical answer. I think the true answer comes down to ambition.
When I first saw Gemini six years ago, my opinion was that it was a hopeful development, but its (intentional) limitations would hold it back from mainstream adoption. I believed at that time -- and still believe -- that, with some changes, the maintainers of personal and academic websites would be encouraged to move to Gemini. I thought that both the core protocol and the Gemtext format would need to be changed for that to happen and -- this is the point I'm trying to make: that they would change.
In the early days I genuinely thought the Gemtext format would either be replaced or supplemented by something that offered genuine expressive possibilities: Markdown, perhaps. The first client I wrote supported Markdown natively, because the tenor of the discussions at that time suggested opinion was moving that way. These days I'm fairly sure it isn't.
My JGemini client -- inadequate in many ways, but at least it had Markdown support.
Similarly, I saw limitations in the protocol that could easily be remedied, that would allow for things like client-side caching -- essential, I think, if there is ever going to be substantial traffic.
In short, I thought that Gemini was a good starting-point for something that could, with hard work and good will, be developed into a real rival to mainstream web publishing. Not for commercial websites, but for the tens of thousands of low-volume academic and personal sites and blogs.
Change, of course, is a problem. I think it's fair to say that almost any change to the existing Gemini specifications opens the door a crack to the nightmare forces of the commercial web. "If we allow in-line formatting today," so the argument runs, "we'll be allowing Javascript tomorrow."
Maybe. I suspect not, but I fully understand the reluctance to open that door even a millimetre.
The result, though, is that Gemini remains essentially a hobby project. It's hard to gauge the number of active Gemini users, but I suspect it's no larger than a few hundred world-wide. Many among that number are perfectly happy with the situation as it is.
And that, I think, is why we keep having arguments. Some of us would like to see Gemini developed into something that offers a real alternative to the mainstream web for "ordinary" bloggers and site owners. We accept that such a development carries risks. Others are happy just to have found a quiet corner of the Internet, where they can interact with like-minded people.
I think it would be nice if there were some sort of compromise position, but I rather suspect there isn't, and that the arguments will continue.
Published 2026-03-26, updated 2026-03-26
Categories
gemini gemlogConverted from my Gemini capsule.