The #gitea developers concluded after 10 years that it is not able to host its own development and that they depend on #GitHub.

[nik@toot.teckids.org]

@davidak @forgejo @Codeberg

It's good for any project to be well-known for something. Even if it's being the most embarassing example for not eating your own dogfood, I guess.

[cinebox@masto.hackers.town]

@davidak @forgejo “we can’t use Gitea because no AI. Also our server keeps going down from the AI crawlers” bruh

[ratsnakegames@mastodon.social]

@dan @davidak @forgejo @Codeberg my strategy of never updating anything keeps paying dividends

[avghelper@gts.average.name]

@davidak @forgejo @Codeberg

"The gap between GitHub and Gitea is larger and larger (including various AI powers). ... Another challenge is the server stability ... due to AI crawlers"

interesting.

[claudius@darmstadt.social]

@nik @davidak @forgejo @Codeberg god you said it so much better than whatever I had in my mind. Why would I trust a project-team that apparently does not even trust its own creation enough to use it?

[blahajj@mastodon.social]

@davidak lol

[moepman@social.troll.academy]

@davidak I would swap gitea for @forgejo right now if there was a documented way to migrate that is not ever user has to move every repo on its own.

[abekonge@venner.network]

@davidak this is ironic

[compfu@mograph.social]

@davidak @forgejo We're using gitea at work. I didn't know about forgejo back when I picked it.
Do you know if there is a way to import gitea's database of issues and pull requests into forgejo?

(If this is something I can easily look up myself, please let me know.)

[feld@friedcheese.us]

@davidak @forgejo @Codeberg disingenuous. They want some Github-specific features and to not have to deal with hosting infrastructure because they're developers not sysadmins/network engineers.

meanwhile Codeberg has had several DDoS/AI-crawler related outages. 4 months ago they were only hitting a 90% uptime... so I guess they're comparable to Github in that respect

but also it's hard to take their hosting infra seriously when they do stupid stuff like make their status page a subdomain of codeberg.org. This is a big nono, you always host your status page on a completely separate domain and ideally on a different Autonomous System

[mahlzahn@nerdculture.de]

RE: https://floss.social/@forgejo/116441502237665413

@compfu @davidak @forgejo

[mahlzahn@nerdculture.de]

RE: https://floss.social/@forgejo/116441502237665413

@MoepMan @davidak @forgejo

[gedankenstuecke@scholar.social]

@cinebox @davidak @forgejo lol, stumbled over that part ass well, don't even seem to see the irony in that!

[feld@friedcheese.us]

@mark22k ahh yes thx

codeberg.org and codeberg.eu are AS29670

status.codeberg.org and status.codeberg.eu are AS24940

so they put it on a separate AS which mitigates that problem but not a different domain

codeberg.org uses dnsowl.org

codeberg.eu uses gandi.net


so if DNS for one of those goes down, the status page goes down too...

[gusted@social.linux.pizza]

@feld@friedcheese.us @mark22k I'll propose internally to setup a mesh network of HAM radio's around the world that can broadcast the status of Codeberg's infrastructure.

If DNS goes down of codeberg.org and codeberg.eu, HAM radios are clearly the last reliable resort and there's clearly no bigger issues that Codeberg (or the world) would be facing at that time.