Exactly, HN is really good for this unlike pretty much every other source these days.
The mainstream news/ tech sites all cover the same things and social media destroyed discovery with their algorithm bubbles. Their algorithms are always tuned to 'more of the same' which is the opposite of what I want. Apparently most users are mindless zombies that just want more cheap cat memes because the approach does seem to work for them. Or maybe the users are just stuck in this because they don't know any better?
As an expat (multiple countries even) I used Facebook to stay in touch with friends but I just can't stomach that some stupid algorithm gets to decide what updates from friends I might want to see or not. Especially because it's mostly wrong. When they dropped the feed for the timeline Facebook basically became useless.
It's so bad now that I don't use any social media anymore at all (except a presence on LinkedIn I maintain in case I need a job change :P )
But HN still shows me stuff I find interesting and I didn't know almost every day. Having raw unfiltered access is a feature, not a bug. Of course top is filtered by popularity but I often check new as well. Yes it takes some effort going through it but it's worth it.
However good that there are options for those that do want it.
I agree with your sentiment and I also use HN because I like the wide range of topics that I may not normally come across. Perhaps some users (not all ofc) aren't even aware of the possibilities of the differences in aggregators and sources, I'm not saying advertising is an answer but discoverability for novel information is likely pretty low.
Is it possible to have it ingest the Upvoted Submissions page on a user's profile, instead of having to manually rate stories? Or does it already do this, and I'm not understanding correctly?
The linked repo requires manual ratings. There's a way to grab your upvoted posts programatically, but it requires scraping, which can get your IP banned if done incorrectly. There doesn't seem to be an endpoint for access to private HN data.
Meta: maybe this rates a (2012) in the title, the initial commit was 10 years ago in February 2012.
Not saying it negatively affects the usability of the code, but perhaps things have moved in the field and it might appear odd or old-fashioned, so a heads-up can be useful.
