[surdu]

It was pretty easy before October 2018, when Microsoft bought them:

https://www.githubstatus.com/uptime?page=31

[13hunteo]

Scaling to human use vs automated/agentic use is a very different thing.

[gbear605]

Github has had horrible uptime for years, before agentic use was a thing. The killer was Microsoft.

[surajrmal]

It had uptime issues before Microsoft. They just got better at tracking and reporting after Microsoft.

[surdu]

So, before October, they were lousy at tracking downtime issues for 2 years (no downtime from 2016 to 2018), but in November, Microsoft came and gave them the technology to correctly track downtime, and they had their first downtime logged in November.

Is this your argument?

[gbear605]

My subjective experience was that it got a lot worse post-Microsoft. I could be misremembering though!

[Barbing]

Do we know this is the predominant reason?

[discordianfish]

Not saying they are doing a particular good job but its not as simple as "paying more for bigger machines" and be done with it.

[AlienRobot]

To be fair, that was 8 years ago. Github now has all that days and users + 8 years of data and users.