[arp242]

GitHub PRs don't solve anything about that, but I wouldn't have to spend (waste) time figuring out the contribution process. At least I learned a few things writing the patches. I learned nothing of value dealing with git email or Phabricator. It's just work of the boring and tedious kind.

[TheDong]

Dealing with github is the boring and tedious thing, you have to run huge amount of proprietary javascript, keep up with their weird UX changes, start X11 to open a browser to render their html, overclock your CPU for a large PR review conversation to scroll without locking up your computer for minutes, constantly click "load more" since their webpage keeps hiding comments (while still lagging massively)...

Email is simple. It's just text, there's no weird javascript or html or lag. I don't have to open X11. I can just open mutt and read or write. I can type "git send-email". It's all open source, so I can read the code to understand it, and write scripting around it. It runs on any computer with ease. Even on a slow connection, it's quite speedy.

I totally agree with you about Phabricator though.

[einsteinx2]

So you can compile and test your changes to Firefox without starting X11 or “overclocking your CPU” but you can’t use a simple website?

[twic]

It's not true that you need to start X11. GitHub's UI renders pretty well under Wayland.

[arp242]

"Boo hoo I need to start X11"? Seriously?

I have some unconventional workflows. And I try not to bother anyone else with it, especially in a volunteer driven open source context. It would be selfish to do otherwise.

To be honest based on what you've written here, keeping you out of my projects sounds like a good thing. What a bunch of piss and vinegar over how other people are choosing to work in a way that works for them.

[elteto]

Starting X takes forever on his PDP11. Only real way to run Unix.

[Osiris]

Use the GitHub CLI.

You can do nearly everything the website does entirely in the terminal.

[elric]

Many projects have rules about what kinds of pull requests they accept. You would still have had to familiarise yourself with those rules, as well as the usual things like coding style, testing policies, etc.

[andybak]

Surely the claim being made is that the overall effort was increased in this case. That makes sense to me. I guess you can debate "but by how much?" but it seems fairly clear that there is more friction than there would have been via Github PRs