it is curious how many people feel they need to ask me why #curl is not moving off #GitHub

The cold and boring answer is money. GitHub sponsors us with a crapload of CI infra that there is no other company even close to doing.

So while they may be an AI-first delusional company, they help our project so much more than all other hosting sites combined.

It would be reckless and irresponsible of us to ignore this.

in reply to Wolf480pl

I don't find that a particularly interesting question because it's an unlikely extreme. Everything in life is a balance of pros and cons, and so is running an Open Source project. We make decisions every single day that impact how the project looks like tomorrow.

We will continue to make such decisions until the project dies. A service we use going rogue is just one thing that can happen, and one I consider is reasonably unlikely.

We react. We overcome. We struggle on.

This entry was edited (3 weeks ago)
in reply to daniel:// stenberg://

This statement appears to be a direct and honest response from someone involved in the cURL project (likely Daniel Stenberg, its founder), explaining why cURL continues to be hosted on GitHub, despite possible philosophical or technical concerns about the platform — particularly around GitHub's AI initiatives (like Copilot and broader Microsoft integrations). peryourhealth.cc
in reply to daniel:// stenberg://

'Github' the corporate personhood barely exists - there is always a lot of good people in companies ... with much fewer having less noble intentions - continuous engagement helps steer the whole to better outcomes. FWIW I think we always think in terms of 'forever' when we should always think in terms of 'all good things come to an end' ... knowing the 'when' is more art then anything.
in reply to daniel:// stenberg://

Sysadmin/DevOps here, who worked on CI/CD for a while specifically..

Serious question: exactly how much CI infra (which, for those unfamiliar, actually means Linux computers, Free Software, network gear and bandwidth) are you guys using, exactly? And just how big is your team, that you need so much infrastructure?

If they were paying salaries for full-time devs, I could understand, but CI Infrastructure is a pretty small problem, all things considered, and solvable with a modest budget, if you have a competent DevOps person available. Is it that you have not found any DevOps people to volunteer? Cause I'd volunteer.

Perhaps the team is dispersed around the world, and Microsoft's presumed access to fast global pipes makes it possible for them to deliver, where a self-managed solution would be cost-prohibitive for connectivity reasons?

I'm going to assume I just don't understand the complexity of your CI needs, and I do not demand you explain anything to me specifically. But I would also urge you to consider that corporate largesse is not a dependable support stream; you should assume that it will get yanked at some inopportune moment by ignorant bean counters who do not, and do not care to, understand, and have plans in place for what you will do when all that infrastructure becomes unavailable to you overnight.