My vision:
Free to browse, subscription to interact. You choose your subscription amount. I take $1 to run the server, anything beyond that gets split evenly between everything you upvote that month.
For the $1, I give 30% of the cut to any 3rd party app that user is using.
With this model, creators make money, and the site hopefully is self-sufficient without ads.
Right now I have a working website, backend, payments, and payouts. What I don't have is a mobile app. If anyone is interested and wants to partner, my email is in my profile.
Site: https://non.io (feel free to play around with login hackernews, pw: helloworld)
Github: https://github.com/jjcm/nonio
API: https://api.non.io (slightly outdated, I need to do an update pass here)
One idea I’d love to see somebody implement: weight content that a user spent longer typing + editing higher as a proxy for thoughtfulness. Unfortunately this specific client-side metric can realistically only work on a native-only service with device attestation and no API, so that the metric can be trusted, but would be cool to think about other solutions to surfacing higher quality content (like TikTok did), as well as solving the inevitable authenticity problem as more and more online content will be astroturfed by LLM’s.
Part of the reason why I want engagement to be paid only, is because when you factor in the "anything beyond [the $1 server fee taken from your subscription] gets split evenly between everything you upvote that month" aspect, you have to be more choosy with what you upvote.
Take a scenario where I subscribe for $10/mo. That means $9 gets split between everything I upvote. If the first thing I upvote in a month is an extremely engaging and insightful article that clearly had hours poured into it, as of that time the author of that article will be scheduled to get all $9 of my subscription pool at the end of the month. If I stumble upon a low effort meme that made me chuckle, will I upvote it knowing that it will reduce the amount of the pool that is going to the first article? If I do they'll each get $4.50 at the end of the month, but are the two of them equal? On reddit votes are free and require nothing, but now that there's an aspect of "voting on this takes a little away from my previous votes", there's another factor at play.
My hypothesis is that this will lead to more conscientious voting. No clue if it'll actually work, but again since you're thinking along the same lines I'd love your take.
The problem with Reddit is that the interests of the users, mods, shareholders, and advertisers are not aligned. Their interests are being pitted against one another in order to generate profit through enshittification.
The way you make a better Reddit is by creating a cooperative model where the users, advertisers, and mods interests are aligned, and there isn't a profit motive for a shareholder group that takes precedence over the quality and governance of the site.
The web site and free app are ad-sponsored. Paid subscriptions first give you an ad-free experience, and secondly, give you and API key you can use in any 3rd party app. 3rd party apps make revenue from the subscriptions, giving them an incentive to create an ever-better user experience.
Mods earn a fee based on traffic and interactions with their subreddit. Not enough to call it a job, but enough for the community to show appreciation.
As a cooperative model, there are no investment shareholders, the site sells bonds to fund it that pay a deferred, fixed yield to get it started. Bondholders, have a vested interest in the success of the site, but have no say in its operation. Without shareholders demanding profits and growth, the interests of the users, the mods, the advertisers, the bondholders and the non-profit that administers the site are aligned.
The best things in this world aren't for profit, the hug of a child, a national park, a sunset and the stars at night. If you want a Reddit-like experience to stay pure, it needs to upset the apple cart and be a tech entity of a completely different model in order to avoid enshittification.
A lot of Reddit (like HN) is linking to 3rd party content, and not much of it is actually user-generated OC.
Nice site though - it looks really nice.
Allowing external links would be fairly trivial, but I'd have to think through the monetization aspects of that. Not sure if I have the right model, but my hope is that since it's open source even if I go wrong someone else can find the right one.
Also thank you for the kind words about the design - it's been a fun labor of love.
Avoid the hypernormalizing of commentary and costly moderation by forcing scrollers to comment elsewhere; cut and paste a link is not hard. Not every social media product needs the normalized set of features.
Social by default is exhausting. It’s the biggest downside to the internet and society these days. Yes we’re social creatures but we also evolved over centuries of social interaction being difficult to come by given slow travel across fast distances. We did not evolve around 24/7 consumption of human inane and repetitive first world gibberish.
Make sure the community has proper and sane moderators/admins and stamp that down before it flairs up. Because it's the internet, it will always flare up if unchecked.
But I get it, that involves trust in judgement, and the internet attracts some very zealous personalities that don't have that. You'd need someone with a proven record to champion that cause. I'm not sure who that is.
Building a single subreddit (or handful, if they're interrelated) alternative seems way more doable. See eg thedonald which transitioned to an external forum until a mod shut it down because of (not really sure, I didn't follow, but it was a working site for a couple years.)
Most of my tech subs are covered pretty well here, but not media.
Possibly use some amount of web scraping too
That’s a small needle to thread.
If the whole thing is lean, with minimal costs, no account reps, and essentially non-profit, it's a lot easier.
Of course, that means it goes one of two ways:
1. Never gets traction, dies after a bunch of work
2. Takes off and the people controlling it (rightfully) want to be compensated for their hard work, so they start adding advertising and focusing on monetization, which requires hiring more people, which requires a more mature business model, which heads towards IPO.
Coincidentally, the forum we're commenting on right now uses a fork of the original Reddit codebase, and shares a small amount of DNA with old.reddit.com. I'd love to see something like HN for all of my favorite hobbies. But I'd also settle for a return to a healthy ecosystem of web forums. What is old is new again!
The current codebase was developed from scratch - zero forking - from what I can see.
HN is written in ARC http://www.paulgraham.com/arc.html which is a language designed by Paul Graham.
Here’s a thread talking about the early days of HN https://news.ycombinator.com/item?id=27452276 and it mentions “The earliest version of HN I can find; 166 LOC” https://news.ycombinator.com/item?id=723767
Here’s Paul talking about the release of HN: https://news.ycombinator.com/hackernews.html
AFAIK this is the latest public copy of the HN code, but it is significantly outdated: https://github.com/wting/hackernews/blob/master/news.arc
The hard part would probably be the anti-spam and other anti-abuse stuff that lives entirely server side. Maybe all it is IS the API endpoint, at least to start, so only the mobile clients can access? Doesn't solve abuse but mitigates it.
It's important that it's not an alternative, or else we'll be in this situation again a few years from now. https://join-lemmy.org will allow us to create a network of sites (or instances) that are still able speak to one another.
The problem apps are shutting down is because they don't want to loose money. If they make a new site, they will likely loose even more money. And they likely can't get profitable without resorting to same scummy behaviour as reddit.
Given the quotes I heard, I'd be very surprised if you lost more money in say, 2 years of building and hosting a new site than trying to even support a month of API calls from reddit c. July 2023.
You can certainly make a decent front end for little cost (even if we're talking about paying competitive rates to a few talented mobile devs) and a small backend won't cost much more. It's always about the community to gather to make all that work. That's the hard part.
Is the reddit API complex?
Of course, "just add a backend" is easier said than done!
But not unachievable.