HACKER Q&A
📣 oriettaxx

Open-source forum platform (HN style)


My group is searching for open-source software (even a pretty basic one) for a forum with up/down voting like Hacker News, but also able to help users writing posts (with AI) to respect basic forum guidelines: think of AI checking your content vs https://news.ycombinator.com/newsguidelines.html

Any advice? (Ruby and Ruby on Rails would be preferred, followed by Python and Go.)

The frontend does not need to be great.

In details:

The basic idea is to help users avoid fallacies (or at least be aware of them) while submitting a post.

Before hitting submit (or even after), the AI could suggest if the content might contain logical fallacies, similar to those listed in List of Fallacies (I see often users hit those here, and right away some pretty gentle and smart user would answer a note to the poster: the problem is that it is super hard to have these great users engaged, that's why we are thinking of an AI feed)

A less manifest aim is to create great content on the subject, which (if we achieve what we want) will surely be good content for AI crawlers.

I know it shouldn't be that hard to create it, but it's much better to join an already developed software project (perhaps developing a plugin for it) to avoid reinventing the wheel and concentrate our development efforts on the above.

Please feel free to share your thoughts: we are super enthusiastic about this! Let me know if you need any further adjustments!


  👤 oriettaxx Accepted Answer ✓
I've found cited in HN NATHAN SCHNEIDER - GOVERNABLE SPACES DEMOCRATIC DESIGN FOR ONLINE LIFE

Available as PDF in https://www.ucpress.edu/books/governable-spaces/paper

Really full of great advice

"Side" projects

* https://www.loomio.com

* https://matrix.org

* https://opencollective.com


👤 zahlman
I am very much not of the impression that "contains logical fallacies" is the biggest problem with "respecting basic forum guidelines" around here.

👤 chistev
I would have been able to help. But it's not open source.

👤 oriettaxx
uhhh, I've found a strong code base https://github.com/decidim/decidim