IRC vanished after many years to favor texting, but IRC was a big part of life for me in the past because you could simply hop on without even using an email address, and conversations were free form without much moderation. The main rule is nothing was taken seriously back then, except net splits.
Not every community needs tracking and user accounts if you ask me, but they do if you ask the government. The increasing call for every community and even IOT products requiring a user profile is going to eat itself with an endless amount of passwords, or even a chain linked authentication scheme, it defeats the utility and benefit of getting quick answers to questions, and having simple human interactions minute-to-minute throughout the day which is what many people really want without knowing it. The key to keeping that manageable will be to keep each community relatively small and specialized if you ask me. Focus on the functionality and purpose thoroughly -- The language used will change over time, so whatever I can build quickly and efficiently in is best. Also, I don't have to worry about security as much as FaceBook (etc.) if I'm not storing user accounts and identifiable data, I just have to put failover, high availability, and data integrity from that perspective.
I've been building on a few ideas that have this in mind in addition to running my own web development company since Feb 2021. I know the struggle too, keep grinding! Cheers.
Good luck with your company!
On your second, I immediately thought of Fabulous, a package that uses the Elm Architecture to build Xamarin cross platform apps in F#. [1] Have you seen it before?
I had not seen fabulous before. It looks really cool! Gives me a sense of validation that others have thought of this.
I tried writing an iOS app in the past (hence the inspiration for the project). If I ever try again, fabulous will be my first choice.
One note about Discord though: why do you find it exclusionary? We use Discord for our customer ops, and we bring on a lot of not very tech savy people. It's really easy: they click a link and it opens in their browser. Not sure how you could make something more accessible in this regard.
What do you mean by "dynamic"? I think the reason it can be hard to make friends on reddit (and other social media) is because friendships require repeated meetings. The population of reddit is massive and the mechanics of the site make repeatedly meeting the same individual unlikely. Also, the design of reddit can make it difficult to recognize people (they'd need to have a distinctive and memorable name).
Discord fares better because there are many small communities and hanging around often results in talking to the same people repeatedly.
The language is still in its early stage, but looks like it covers your use case
First idea, a website that combines Reddit (or HN) and Discord. The goal is to create ephemeral, real-time communities (like this thread). I find discord too exclusionary to bring strangers together. And reddit is not dynamic enough to encourage relationship formation. With this site, you post a thread, people start chatting, there's a simple interface for groups of people to break off the public chat and go into a private chat or into a webrtc call. Once the thread loses momentum the chat room sort of dies. And people move onto the next thread.
Backend is written in Go. Frontend will likely be Elm. Redis for pubsub. Postgres for crud.
Second idea, an Elm-like language for building linux applications (and crucially linux-phone applications). I'd really like to de-throne apple and google from the smartphone market. I think a dead simple language like Elm + a super simple, standardized dev environment + no weird configuration or new conventions to learn would be a killer feature for the linux platform. Something like Elm-UI would attract a wave of new developers to Linux.
Unfortunately, I've never made a programming language before. So I'm working through "Crafting Interpreters" to get started.
Third, possibly finding a SWE in Miami and starting a software consultancy partnership. Could be fun and lucrative but I've never been on the business side before. So likely this idea will stay half-baked for a lot longer.