Retry link checking with backoff

[grst]

Found myself restarting a job too often because a random site didn't respond on time

[flying-sheep]

Why are you adding all this code instead of using something pre-made?

https://will-ockmore.github.io/httpx-retries/

also we should just reuse a session instead of always touching every single place we use httpx every time we change something, right?

[grst]

It was simple enough that I didn't bother searching.
But this looks neat, I'll use that instead.

[flying-sheep]

You have a stronger stomach than I do if you consider anything HTTP related simple.

There are always so many edge cases that I just want to play lego with blocks built by people who really understand all the layers and pieces instead of rolling my own.

[grst]

That's fair, but the worst thing that can happen here is a CI job failing

[flying-sheep]

Of course! Still, less complexity to maintain is good, right?

[flying-sheep]

Great, thank you! I just moved all the repeated timeoff and follow_redirect parameters into the client creation, and now this looks better than before with more functionality!