Goldfish Social – Free open source alternative for Vine / TikTok (alpha) | Hacker News Reader

Goldfish Social – Free open source alternative for Vine / TikTok (alpha)

github.com

85 points_b6mc25 comments

I’m going to go out on a limb here and wish the project the best, on grounds of the hacker mentality that doing something just for the sake of it is the entire point. Which is great.

At the same time, in the real world, I expect it to go nowhere at all. For that to happen, the technical implementation matters not a bit. You’d have to provide _anything_ new, different and “better” as compared to what is. To be an alternative, it has to provide similar functionality but improve on it in some regard. Instead, this is a worse experience in every regard.

Maybe it’s just way too early to release anything. But I believe the entire idea is deeply flawed, if targeted for general public use.

> the technical implementation matters not a bit

This was the hardest lesson I learnt, it doesn't matter how polished and amazing your code is, or what technologies you use, end users really don't care at all.

It's definitely true that they don't care directly about it, since they aren't exposed to it at all, but I would say they care transitively about it, as the decisions made to build systems impact the reliability & performance of the product, and its ability to be maintained and improved. I've seen too many products that start of usable but due to maintenance issues end up becoming bloated to the point where a sleeker competitor comes along and solves those problems for me.

Yeah I was going to say something along those lines.

End users care about features and code quality, but not in the way developers think.

Users deeply care when their application keeps crashing (but they don't care at all when it doesn't crash)

Likewise, users care when the app is too slow, too.

And, they care about the app having all the features they want or like (even if they don't know it and can't articulate to you exactly what they want).

Users don't care how maintainable the code base is. Whether a new feature takes 2 hours or 2 weeks to make is completely irrelevant for them. Now, they care about the second orders consequences of this: an unmaintainable mess will probably never have all the features they want, or will be too buggy, etc.

But, code quality is just a lever to achieve the things that users want. It's a means to an end. If you don't pull the lever correctly, it's wasted

It’s not surprising that people who care about the tech would care about its execution.

At the same time, if you‘re looking to create a solution to anything beyond scratching your own itch, users are key.

The job essentially becomes dealing with irrational human wants and desires when it’s quite possible that one of the reasons you went into the profession was to precisely not having to do that.

I hope this project goes somewhere, but I feel like it may be too early for a public launch. I also think it may hurt the project to position itself so clearly as specifically a TikTok alternative given how difficult it is to replicate TikTok's defining characteristics - a powerful video editing app, ability to duo with other people's videos, and a famously good recommendation algorithm.

Thanks for your input! The live 'demo' is more as a way to get input from the people who are going to use it in the future :-)

For example the UI is still a WIP, I do not want to make a simple clone of TikTok itself but rather develop it along the way, the way the user wants it.

About the 'TikTok alternative', well that was my intention and the plan is it will get its own name and leaving the TikTok part all behind but at this stage it's a good way to get the right people to the project I think :-)

Love seeing Laravel in the wild. We've had a lot of success with it so far. Good luck with the project!

Completely aside, but I used to write PHP 10-ish years ago during the PHP 3/4/5 era. My career launched me into Node.js and JavaScript but I've since come back to PHP and it's been really lovely to write and maintain.

PHP is the goat!

What did you like about nodejs and js?

The operational challenge of hosting this is similar to the challenge of hosting a youtube alternative. Bandwidth, bandwidth and bandwidth.

Maybe a naive question: how about cdns like cloudflare? Too costly?

Edit: and also P2P style caching? I would expect a big number of users can act as relays?

And, IPFS, why not.

The way Youtube made it work was that they own their CDN, globally.

Yes I think using someone elses CDN would eventually cut into your profits, which btw, wtf are your profits? Self hosting a video streaming platform is a black hole for money. The only way it's possible is if you use it to create an ad platform. You must have a revenue stream to keep it going and allowing kids to share videos a million times an hour is not it.

And regarding p2p, it's a deep technical discussion but I think you can simplify it if you look at this from the end user perspective. They're swiping down like once a second sometimes and every time they swipe down a video is ready to play. That's highly on-demand video streaming, for thousands, or millions, of users.

So I don't care what p2p solution you got, you can't bypass the basic laws of internet tubes. Didn't you hear Al Gore? There is only so much on-demand video streaming you can fit into one tube. ;) I say that tongue in cheek but what I mean is at the end of the day someone has to pay for the bandwidth of constant video streamin, even if it comes from p2p it's still a lot of bandwidth.

What peertube is doing would not work for a tiktok/vine-style app because it's too slow at establishing a p2p connection.

I'm curious about this line from the README:

> We started out with something different, but after some great advice restarted the project in SPA form (Single Page Application).

How did the project start?

At first I made a 'normal' web app with reloading on each page etc. Currently I'm working on an API and web front end.

So at someo point it will be easy to create mobile apps and connect them via the API

"…later add ActivityPub." Famous last words. Anyway why do people try to repeatedly reinvent the 1980s BBS?

Because BBS were awesome, but as they werent really monetized, a lot of people think they can reinvent it in a way to monetize it.

I just think it's an easy to understand model, topics and replies, but its well past its sell-by. It leads to sloppy meandering discussions, and the best way to monetize it is to mine the users without encouraging them to do better.

It seems that the only way such social media can survive is through ads. Have you put any thought into how to monetize and stay afloat?

Everything is slick and fast. And the design has some personality like a dark glue.im I look forward to see how this grows.

Working on it still! :)

Cool, looks pretty clean! By the way, the "Popular" hashtags on the right are not changing the page when I click on them. Nothing happens.

tldr

- AGPL

- No recommendation system

- Videos don't autoplay

- Videos are not fullscreen and have different aspect ratios

It's quite different from TikTok.

Also I got 4/5 posts about guns…

3D printed guns..

Reply on news.ycombinator.com