Mastodon intends to move from Discourse forum to Github Discussions

As of yesterday, you can now sign gitea commits with your SSH key

Gitea already do federated profiles which have your ssh key in them

It would be really useful if gitea federated more, so that I could add a pull request to a self-hosted gitea repo without having to register there

In essence a federated replacement for github, powered by your profile and your SSH keys, I think there would be decent demand for that

1 Like

Very nice that you are a Gitea contributor as well.

I think that Gitea has an absolutely unique opportunity right now, and I don’t know if they fully realize what federation could bring to the project. The social features that could be built might set Gitea apart from Github, which is more of a ‘toolset for techies’ rather than an inclusive free software community facilitator. In forgefriends community we are discussing on campaigning to increase that awareness. (PS. Just saw your GH comment. For others: It looks like they did not get a NLNet grant for federation support).

Update: CC @anon20068248


Awesome! I’ll be working on federating my own self-hosted instance of gitea, starting with the profile, and will compare notes

It’s slightly unfortunate that git users (ie as in git config user) are keyed off name and email. I asked about 8 years ago if another field could be added for a url/profile page, but I dont think it was taken up

Otherwise you could today beautifully link together, commits, pull requests, profiles, followers, inboxes etc.

Our biggest setback as a federation was keying identity off email addresses, which creates a bottleneck with the big email providers. That’s not to be confused with logging in with a user@host type identity, or following someone with a @user@host type nick. Nicks can be tied to a profile, and used for login etc.

Nevertheless I agree that git federation would bring enormous utility. I’ll leave it there to avoid going off-topic from thread title :slight_smile:


See this thread for the latest updates I have on the NLnet grant (dated November 21st).

1 Like

I’m cautiously optimistic that there will be good news on this front

I think we should really get behind this, as a community, especially input into the design and early prototyping

Similarly support for forgefed, and I was pointed to this project too:

Let’s make federated git happen in 2022. I’m in!


Can we please stop to tell “alternative facts” here?
Are there any proofs that discourse has anything to do with the fediverse?

Also 2 days ago (apart from personal talks with nlnet people) :

Nearly all smaller successfull Fedi things were rejected by NGI0 …
So, can we please stop to pump up hot air ???

On the other hand, it is a remarkable moment to do it now regarding

That would be great, thank you.

Read your original sources. It is Pavilion the plugin developer that intends to write federation support in them. First mentioned by @how and I followed up on the Discourse forum yesterday to ask what the grant rejection would do for their plans:

Well, then continue to ignore official votings; for example that this official forum should be federated.
The group did this 3 years ago (!)
And if it will be continued to be ignored probably all will leave this walled garden.
No worries.

The EU parliament voted for mandatory interoperability anyway.

It is entirely unclear what your beef is. Groups and projects express an interest for the fediverse. We try to help them by giving feedback and show that we would love them become part of fedi to encourage them to make that step. It is that simple.

Ah. I see.
That is not “Discourse will federate”, this is “Discourse got rejected again”.
And everybody was told when asking nlnet.
We wanted to do the work with 8 people and got a “no” response with various strange reasons like

Obviously, it was not a priority to them if there is external money on the
table and it still doesn’t happen. A scalable forum application with the
modern features people expect would take a long time to build…

Otherwise we would have done immediately

Anyway - all I am saying is:
If the fact derives
Person X applied for Y
then let us not write it as
Software Z will do magic …


Well, that’s only if you never use git with a .mailmap.
The way I use git for my projects, I always just put a bogus email address into commits so it will never be current one day and then outdated another day. Then I put my “real” email address in the .mailmap.

So if I wanted to connect my git identities to the fediverse, I would have a couple options:

  1. start consistently entering in my git identity as or something directly based on a fediverse instance
  2. using some weird improved counterpart of .mailmap, connect whatever git identities I’m using locally (those might look like valenoern@pinebook, valenoern@server, etc) into a particular fediverse identity like

I may be misunderstanding what the problem is here, but if we assumed collected user email addresses during registration and could forward emails to its users if absolutely necessary, then wouldn’t fediverse handles on it like become about the same as a ‘real’ email address for the purposes of git?

The process of emailing patches might even be a bit improved, because you could email a patch to out of git and have it show up as a pull request in my inbox.
(Hopefully with the patch file annotated with what project it’s for so it would be grouped with the correct project. Maybe the most simple and future-proof way would be something like, you email to to send a pull request for Valenoern/bopwiki?)

Or, who knows, maybe upgrading the emailing capabilities gitea servers already have would be too much and we’d be extending git to be able to directly send ActivityPub.
I guess it depends on whether we’re counting on most contributors being able to install things like git, or assuming entire software communities are going to shift further toward platform computing where does all the technical work for a big category of completely “non-technical” users that just sign on and advise all the platforms they use on what direction they should go.

(I have my own biases against some forms of platform computing, but I also don’t want to make the assumption it couldn’t lead to interesting new forms of governance where, say, a bunch of non-programmers elect a new leader to a stagnating project that keeps wilfully ignoring issues, and this eventually leads to a real plan to get them fixed.)


Thanks @Valenoern cc’ing this to @anon20068248 as possibly relevant to forgefriends community.

1 Like

@Sebastian I would like to express my discomfort when I read

and I did not want to participate in the conversation anymore. I enjoy discussions where the tone is friendly.

I see. I do not like alternative facts, cause I fight misinformation.
For a reason.
It is explained here Mastodon intends to move from Discourse forum to Github Discussions - #11 by Sebastian and because I was told “forum software is to complicated to fund” and if we pretend that such projects are funded, then things like I fight for on an EU level, might be considered as not needed just cause of false facts. The same maybe for Sovereign Tech Fund and the new stipends my public broadcaster is creating.
We need to be honest about the state of funding, I think …
Especially since I see the nice rejected projects in meetings frequently and I know and see how hard some people are coding.

I will leave this forum for the reason explained above. Could someone in @staff please anonymize my account?

Being offensive and abrasive is no good journalistic quality.

And due dilligence on the other hand is…

Argh - repeating false facts, I get crazy !!! waybackmachine is your friend.

Person X applied for Y is NOT Discourse will be joining the Fediverse
It is not.
It is not !!! (!)

Apart from that:
We wanted to pick it up with 8 people. It was just one round before. nlnet simply does not seem to want to fund it. One person withdraw, but 3 different applicants were then rejected 3 times.
How often should I repeat cites?

We had several group decisions and I do not think, you can change decisions by 60 people.
When I payed the events in 2017-2020, when nlnet did not give a penny, I co-organised both conferences:
And >60 people travelled to Prague and agreed to a federated forum.
And I respect the majority, cause democracy.
And I am really fed up with the complaints about not federating, cause I can’t change it.

Several teams tried to get funding over the last two years to implement some kind of ActivityPub support in Discourse. As you well know, all such attempts failed so far.

You can do two things:

It you need a culprit you’re looking at things the wrong way.

No worries. From now on, I will clearly respond to all that I have tried and can’t change the walled garden thing. Although I am missing some people and so began to talk with all from the conf.
I do not understand “follow-up” cause not getting funding for redaktor and two other partner projects, also getting rejected by nlnet again, I am working 8 hours for ActivityPub at Gruner+Jahr and afterwards 8 hours for free for ActivityPub at GitHub - redaktor/widgets-preview – each day and that changed life.
Please cross fingers for (Steffen had a Lightning Talk at the conf …)

The NGI Zero Discovery Fund is fully allocated, and no more new projects
are accepted - the fund is wrapping up.