I get you're trying to be a white knight of sort and protect the world from HN's american-centric view, but of the 215 world capitals it's currently Sunday in 195 of them and Monday morning in the rest, so...
Most people in Auckland are very probably still asleep. That they be the responsible maintainers - of a service which I understand is in Germany - is not impossible, just pretty far from the main bets.
Jokes aside, I think you’ll be surprised to learn that the DevOps effort for most official programming language resources are surprisingly alike, regardless of their popularity.
OTOH static sites are an obvious solution for high resilience, low effort devops, and low cost hosting.
They could decouple the comments section to a dynamic API + minimal client-side code. Even if the API and/or DB go down, all the pages of the site will keep working.
I would think reliability and uptime are high priorities for a language that could almost be considered critical infrastructure at this point... but what do I know.
1xx is the response code you try to ignore because it's really annoying to deal with properly (like when you receive a 100 response to a file upload, or even worse, a 101 with a protocol that isn't exactly what you expected, or a 103 which means more content comes later).
1xx response codes are crucial for things like WebSockets and HTTP/3 but they make the HTTP state machine more complex and many people writing manual HTTP requests are ill-prepared to deal with them.
Seemingly https://www.elektronauts.com/ (community for Elektron devices) is also down, coincident? AFAIK, Elektronauts is running Discourse, so Ruby, not PHP, so unlikely to be related that way.
But, I'd wager a bet on some DNS misconfiguration at some CDN or likewise, that both are using :)
I haven't used PHP in years, but I always felt the user examples section made their docs better than the Python docs which are still somewhat wanting for common use case based examples.
I understand posting about outages of critical infrastructure (like us-east-1) or interactive websites. But this is a non-interactive website, whose only purpose is delivering text (PHP docs) and maybe some binaries. The few people working on PHP projects on a Sunday may not have access to the official docs for a while. But they can just hop onto the Wayback Machine and carry on with their work.
For the manual (of which, I insist, coders should have an offline copy - they have been available since forever and prudence dictates), there is a backup site, apparently, at
When latency and pageload was in seconds instead of milliseconds, downloading the documentation as a Windows Help file improved my efficiency.
This is a "kids these days" post. Maybe in a few years "The connection to the data center hosting the 1000 GPUs that run the LLM I ask how to do stuff is down, I can't get any work done!"
There's absolutely nothing preventing anyone from downloading the docs. However, search engines won't return results from your local .hlp files and there are plenty of cases where the comments underneath a manual page give context for edge cases that wasn't necessarily described in the docs.
I personally rely on the local documentation in my IDE because I can't be bothered to continuously update my docs with the latest errata and additions.
reply