@Akshay https://codema.in/d/o7ytjFpg/future-of-git-fosscommunity-in-service/2 five people volunteered. Only if new people manage to setup the new instance, the migration will succeed. 500 Euro is enough to cover 2 years with 20 Euro per month server - we might be able to move to a smaller server too. We could ask Infomaniak again or will have to find other funding once we run out of this credit balance.

@Akshay yes, we should shut down services where we lack volunteers and funding (we did this with poddery.com diaspora already), in case of git, we have secured both. Only if the new volunteers fail to take over the service and complete migration, we need to think about shutting down.

We also optimized cost and effort for xmpp by moving poddery.com xmpp to durare.org server. poddery.com xmpp is a virtual host in durare.org - one extra configuration option in prosody. So cost and maintenance is shared between two communities (durare pays for both).

We should go through this process for each of our services - videos and git, we are going through that currently and next in my list is poddery.com matrix (others are welcome to beat me).

We are also targeting to setup forgejo instance before end of October. Does this address your concerns?

As for general strategy for fsci services, I think different teams taking ownership of different services are better as some people may like to maintain some specific services based on interest rather than providing a collection of services.

There may also be some people who might be motivated by providing many services too, so they can focus on providing a broader strategy, messaging and support to services that need help.

So we need both approaches and people can choose whichever they prefer. For example I was interested in diaspora, matrix, xmpp and git earlier, but lost interest in diaspora, matrix and git. I want to focus on xmpp (prav and durare) so finding a new team for git.

Lets target to finish setting it up before end of October.

I just found out about this https://recipes.coopcloud.tech/forgejo, @FbIN (VB) , @Athul @Hari Govind S @sudo_medark @Badri Sunderarajan what do you think? I think collaborating on configuration management is good. We can reuse common setups like email, anubis etc easily.

Documentation link https://docs.coopcloud.tech/operators/tutorial/#dns-setup

@Pirate Praveen I don't like this one specifically because it depends on Docker and I prefer to run things directly where possible. But if other people feel comfortable with this, I'm okay to use it.

Forgejo specifically does not need much in the way of setup because it's just a single binary; the "upgrade process" is usually to replace it with a new binary and restart the server. But I agree that in general it is good to be able to collaborate on configuration management, and I'm open to experimenting with something here so that we can learn from it and use the same method for other FSCI services that have more complex setup.

YunoHost is another option, which runs programs directly on Debian, but it is opinionated in the sense that it assumes we want one account with a common login for multiple services and integrates that wherever possible. I think there is another initiative called FreedomBox or something; I am not sure how that works. We'll probably find a few NixOS based options under the NGI Fediversity initiative, and the Guix community could have recipes too, though we will have to check how mature they are. More information on these or other suggestions are welcome.