Awesome-Selfhosted
github.com
github.com
These days I feel like there is more value in curation than an exhaustive list..
There is just too much on here and a lot of it not of super great quality for me to want to take the time to try a lot of things..
I would probably get more value out of a list that had the best overall 1 or 2 from each category on this list, and even a "best standouts from all categories" would be cool.
Perhaps more useful would be a feature matrix for each section of this repo so you can determine the "best" ones at-a-glance (since best is subjective based on your particular needs).
Still a great start!
Could I make that work? Probably, if I cared enough. Do I want to? Not unless I can't find something newer and similar enough.
One of the CLI tools I use every day (minor filesystem management tool, written by me for me) hasn't had a commit since 1991. But I use it every single day, it does what it needs to do.
Most of these projects are works in progress. It would not be fair to exclude them in favor of projects with larger budgets and internal marketing.
Some of the projects haven't seen any significant work for 5+ years. In some areas that might be fine (projects can be "done") so it's not automatically an indicator of being abandoned, but level of activity and whether the maintainer is still merging PRs is usually the first thing I look at when I'm deciding which of a few options on the list I'm going to dig into further.
I wrote a competitor to OwnCloud and NextCloud that took years of working by myself to make. With no budget by myself. Obviously the quality isn't quite as polished as my competition.
By your logic only OwnCloud and NextCloud should be allowed on here. Despite both of them having their own marketing budget they promote with. They should get to come into aggregate lists and boot out the little guy.
Projects that fail to attracts users tend to do even worse at attracting developers because there are always going to be far more users than developers.
Some products prioritize different things. For example, all of my products internalize fonts and scripts so they are truly self hosted. My competition doesn't do that. They host scripts on CDNs and utilize Google fonts. On some networks, particularly air gapped intranets or offline locahosts, that won't work.
By passing objective judgement on the contents of the list, you're dictating to the audience the factors that you think should be important to them rather than showing them all factors and letting them choose what they believe is important.
Now you search one to three from category you're looking for to see what smart people are blogging about/what people are using who are accomplishing what you want to accomplish with that tool.
I would probably be more helpful if i actually tried to do something about it.
0 results
In terms of the intersection of "awesome" and "self-hosted" I am surprised to not see Cloudron[0] mentioned here or in Github Awesome list.
I'm not affiliated in any sense but a lazy+loyal lover of their service. https://cloudron.io
Dealing with collaboration and HN/Reddit level of traffic coming to your server is best left to the big collaboration platforms.
Imagine if app installs on the server were as simple and easy as installing a desktop app. Having a ridiculously easy web app installation process for servers would unlock countless opportunities for developers to reach more users or customers.
There appears to be no appetite or interest in the industry to tackle this problem. And just to pre-empt the suggestions that Docker, Cloudtron, Sandstorm, command-line scripts are possible options: none of these are easy or simple for non-technical users. Even one-click app marketplaces (Linode, Digital Ocean etc) have their complications.
Without easy deployment, particularly for non-technical users, self-hosting will never be a viable option for most people, only for a tiny minority of technical users. If you disagree, take a look at the unstoppable juggernaut that is SaaS.
- Have some "global self-hosted sass installer service" pre-installed and running
- Visit someselfhostedsaas.com - it detects that you don't have their software installed yet
- It automatically initiates a download and installation. User is shown a loading screen. Doesn't even need to be different than your standard SPA spinner.
- Once installation is complete, the app/local-site automatically opens. User is off to the races
And it will liberate your email, the most vital of internet services (think of every password recovery) from the control of arbitrary shutdowns by google et.al. Own your own destiny.
You'll also get less spam and way less (none) legit emails going to spam and you'll discover how terribly bad gmail spam filtering actually is in comparison.
If you're completely nontechnical, sure, don't go there.
But if you're even passingly familiar with Linux administration and honestly willing to spend a week or so learning (assuming you're starting from knowing nothing about email hosting, only a couple days if you already know about SMTP and postfix), you can do it.
Ongoing maintenance is nearly none. The majority of years I do nothing on my email server other than apt-get upgrade.
Why the heck we need javascript to navigate within a page is beyond me.
Or was this always the case?