[ihinsdale]

You've described a lot of my vision for Sequiturs (https://sequiturs.com). Check it out! I'd be very interested to get your feedback.

[Chris2048]

Now we just need to go through HN an incorporate comment-knowledge into this format :-)

Is there good search across questions?

[ihinsdale]

That's an idea! Motivating people to use the argument format has proven a challenge, since it often takes more time to break an argument down into this format than to string sentences together into a paragraph, but I'm confident that is a solvable problem.

You can search argument content at https://sequiturs.com/search?q=&page=0&limit=10.

The search indexing is pretty decent considering how easy it was to implement--I'm using Postgres's full text search capability. It's high on my to-do list to make comment content searchable as well, and at some point I'd like to switch over to ElasticSearch.

[Chris2048]

Here's another idea - have a index from news-source article to a page that breaks down it's assertions, then links each one to a discussion of it :-)

[ihinsdale]

Yeah, I think that's a great idea. What I'm imagining along those lines is an interface for marking up a source text--basically, for easily turning a highlighted portion of the text into a proposition.

I'm currently working on giving every proposition its own page that contains links to all the arguments where the proposition is used as a premise and all the arguments where the proposition is used as a conclusion. Once that's in place, I can work on the functionality for actually highlighting source text and creating propositions from it. The data model behind Sequiturs already features a notion of 'source text', in the form of a Prompt, which consists of a url or some actual text.