4 ms·

Here's a thought I've been turning over for a little while now. It occurs to me that all extant social media apps have, at a high level, exactly the same requi

• TulliusCicero 9y ago
  You're kidding yourself if you think people actually want to run their own server for the purposes of social networking.
  • mundo 9y ago
    You're implicitly assuming that Amazon does a bad job of implementing the front end. If the goal is to get even a tenth of FB users to be paying EC2 customers, they can probably spare a couple of UX developers to give it a easy-as-a-ios-app interface.
  • lifeformed 9y ago
    It could be pre-made and run on AWS automatically be default. More advanced users could tinker with it if they want. You would pay money for the costs of running it, plus an extra something.
    • TulliusCicero 9y ago
      > You would pay money for the costs of running it Nobody except hardcore tech nerds would want to do this.
  • ellius 9y ago
    I do wonder though if you could create a compelling product that would keep the customer from ever knowing they "own" a server. I've thought about a service where I effectively would run one big "load balancer" front end that would route a customer's traffic entirely to prefab cloud servers that I would spin up & configure via tools like Terraform and Ansible on the back end. I wanted to load those mainly with privacy tools--things like Pi-Hole, encrypted DNS, etc. I don't see why I couldn't say "hey in addition to this you get an app portal" that is effectively providing a standardized social media app. So for a regular price you get privacy, ad blocking, and a social media / cloud storage app that doesn't spy on you. It seems doable.
    • mundo 9y ago
      There is a chicken-and-egg thing too - it's not normal for Joe Internetuser to have a $5/mo cloud server in addition to his Netflix/Hulu/etc because there is no easy-to-use-for-regular-people software to run on one, and no one builds such software because such users don't have $5/mo cloud servers to run it on. This is sort of the fatal flaw of app portals like Sandstorm, but it doesn't need to be that way forever.
      • ocdtrekkie 9y ago
        Well, you can use a managed version, both Sandstorm and Cloudron can be subscribed to as online services. I also think there's a lot of opportunity for families and small groups. Most people I am close with would not run their own server, but perhaps would use mine if I offered it.
    • warkdarrior 9y ago
      If users do not see a visible distinction between your app portal and Facebook, why would they pay you instead of using FB for free? Is it enough to claim privacy in order to make a sale? I really doubt it.
      • mundo 9y ago
        You could be right, but it isn't that far-fetched. Consider: * A lot of users would be in the pennies-per-month level, charged by standard Amazon usage prices * Owning your own data is a big deal - it's not uncommon for people to pay for backups, which would essentially be an included service * It's not unreasonable to imagine a company of Amazon's size and competence doing a very good job of this, and including novel features that FB/IG/etc can't match * It could also be a loss-leader to get people to own servers for other purposes, e.g. hosting a vanity domain
• rhcom2 9y ago
  You might want to read about urbit [https://www.urbit.org/ https://www.urbit.org/]. I think it might be similar to your thinking.
  • mundo 9y ago
    I'm familiar with Urbit, and intrigued by it, and a big company deciding to try a project inside it is inarguably the best thing that could happen to Urbit; but I'm not very bullish on that happening, for a variety of reasons.
    • rhcom2 9y ago
      I think that's an excellent way to view that project. Maybe something down the line can use some of Urbit's ideas.
    • justinmk 9y ago
      Urbit is designed in theory so that proportionate incentives exist at each stage of adoption. Initially it will be curios only, like the interwebs in 1995. If it gains value then applications will be built on it. Even if urbit fails, the theory is sound. It turns out economics is really effective. And advocacy (99% of the other solutions proposed in this thread) is as limp as it was in 2000.
• s73ver 9y ago
  "It occurs to me that all extant social media apps have, at a high level, exactly the same requirements: 1. Allow users to upload some data to cloud storage 2. Make that data discoverable to certain other users 3. Show everyone ads" Identification should be on that list. Allow me to uniquely identify myself, and prevent others from attempting to impersonate me. "a. "ads" - No one actually wants them, so get rid of them" Nobody wants them, but you need to keep the lights on somehow, and people are even less inclined to pay for things. "i. Most people don't have a server on which to host it ii. Most people won't pay for it, so someone would have to write it and make it really easy to use, for free ...and by a lucky cooincidence, both of those objections have the same answer: Amazon. Most people don't have a server? Amazon will rent you one. Who would develop a self-hosted FB clone for free? Amazon, to get people to rent servers." So now you're trading Google for Amazon? That's no better. And you just said that people don't want to pay for things, but now you want people to pay Amazon for servers?"
• the_common_man 9y ago
  Highly recommend cloudron.io. They are trying work a SaaS style deployment angle for people running their own server.