TLDR: Curl has been around for 10,000 days and they’re collecting user stories.

Back in the 00s we had a misbehaving application distributed on about 20 servers that would go haywire from time to time. I built a server monitoring solution consisting of a cgi bin shell script that cat’d a bunch of /proc info together, then curled them from a central server and formatted into an html page that one could refresh every few seconds to monitor server health. Then I could go in and restart the node before the business users started screaming. They were too cheap to buy an existing solution to do this.

Don’t miss those days at all.

Prometheus way before Prometehus became popular!

Prometheus a decade before Prometheus existed :-)

c10k was already a (different) thing.

https://en.m.wikipedia.org/wiki/C10k_problem

I had the same thought, "oh damn, I haven't heard about 'the c10k problem' in ages" :D (it was "solved", right?)

Given the author, his history in the field, and the project, though, I'm confident the repetition is intentional.

Like a contemporary rapper dropping Doom bars on a track, I read it as an homage / callback / shout-out, even if the meaning is changed.

Yep, and that problem is history now. Just like c10k from curl will be history.

The ability to submit a story using a curl would have been fun.

My personal same is that since wget url did what I wanted, I took too long to learn how to curl a url to a file.

I still don't know the exact command. I just use wcurl if I need to curl a file, but muscle memory still dictates wget most of the time.

Curl has been in my toolbelt for perhaps 9000 days. Can’t remember when I discovered the retry/timeout options, but that allowed me to remove a large chunk of buggy bash code, from a semi important cron job.

It's in the second sentence of the article, so not exactly buried?