The Small Web isn’t about going back to the days of GeoCities. It’s about going forward differently, using modern tech in a non-colonial manner.
It’s not about building clones of Twitter, YouTube, etc. There’s no way anyone can self-host a dozen different services. Instead, it’s about having a single-tenant place on the Web that you own and control without technical knowhow; a place you can add Twitter, YouTube, etc., *features* to.
Please stop saying that things like "There is no way anyone can self-host a doze. Different services"
This is simply not true. I am currently self-hosting 19 services.
@SigmaOne @aral Don't get me wrong, I do appreciate the sentiment of moving from choosing between companies that own and control your data (Google, AWS, et al) to moving to everyone self hosting everything, but speaking in absolutes is not helpful in getting there
As for saying that easy-to-use single tenant servers is not self hosting nor being controlled by the hosting provider (e.g., AWS)... well I'm not sure what to say to that. I must be misunderstanding something
@adam @aral I think the point is that non-enthusiasts don't have the capacity or willingness to run and maintain a bunch of services just to avoid the big companies, but there should instead be a single service they can run which has the features they need without the hassle of maintaining separate servers
So essentially things like peertube, a fedi instance, etc. rolled into one easy to maintain and use modular bundle
@SigmaOne @aral Sounds like just improving existing self-hosting solutions. Yunohost and Nextcloud both come to mind. However, I only hear @aral talk about building new things that will compete with the existing solutions.
If the community's efforts are going to be further divided, there should be a compelling reason. What requirement do existing solutions not meet?
@adam @SigmaOne They’re created to host from 1-1,000,000 people. So they’re designed for 1,000,000 people not one. That’s a huge amount of additional complexity.
It’s like telling someone, you don’t have to fly commercially, you can fly your own 747. Now it’s certainly possible; John Travolta does it. But the folks who want more people to be able to fly themselves are designing solutions that look more like drones.
@adam @SigmaOne In the same way, Small Tech and Small Web are about designing solutions for 1. Not 2, not 100,000, not 1,000,000. Single tenant apps, as we call them. That’s what reduces complexity and will make it possible for everyday people to use them.
So, for example, compare the effort of setting up PeerTube (which is designed to support thousands of people) with Owncast (which is designed to support just one). The latter can be up and running in under a minute on a tiny server.
@aral @adam @SigmaOne the different software architectual tools exist to allow building a non-poweruser friendly solution that scales up at the turn of a knob. we're not there, yet but the lego pieces exist. you got lovely ubuntu ux; they allow easily installing minikube; you got autoscaling objects in k8s, plethora of nice monitoring tools. i dont think complexity for 100k+ is infinite. it should just be a matter of available hardware. _should_
@aral @adam @SigmaOne It sounds kind of like what Drupal 6 offered: An easy to use system that mostly worked out of the box and could be enhanced with all kinds of services.
Though that offered cooperation out of the box. Just add more users as needed.