I've never made a PR or issue on anyone else's project. Don't want to annoy them.
Discussion
Have you?
I do enough for the both of us.
Girls mostly just clone or fork and change locally, I think. To be less-intrusive.
Then how do you learn what it takes to bring something upstream? We for example encourage our interns to propose changes to open source projects. There is a lot to learn there about how to do things properly. All assuming you don‘t treat GitHub as if it was a GDoc or CMS.
I ask male developers I know to submit for me.
That's what she said?
Well, I like knowing that they also think it's worth submitting and when questions arise, they respond and sound all official and mansplain-y, and I can just lurk.
It's different in my own projects or with some particular developers I have a good rapport with because they're used to me and take me seriously.
Yes. It’s difficult to get developers on other teams to even look at unsolicited pull requests. (The open source world is a different story.) I’ve written extracurricular but unambiguous improvements to my own team’s code, and only the tiny ones have been reviewed.
I've almost done it a few times; write it up and then don't submit.
These days I try to resist even doing the work, because it’s almost always wasted effort.
These can be disabled in the repo. If they're not, assume that the project is glad for you to annoy it
Even to update a readme? 😅
Nope. I don't do anything unless I already have a working relationship with them and we're collaborators or chatting on Slack or Nostr, or whatever.
Last time I filed an issue (or a comment to the issue), no one understood why this was an issue, and I decided to keep a .whl snapshot of the library with my proposed patch and redistribute it myself. Instead of even trying to understand, the repo owner "locked the issue and limited conversation to collaborators". I don't want to file issues anymore since.
Yes, I think this is not unusual.
I've seen the same behavior on OS repos and corporate bug boards.
Gatekeeping behavior.
I was employed as a software tester once and a woman who didn't like me sat on the bug board and everything I submitted was immediately closed at the next meeting as "not reproduceable". Within a couple of weeks a customer would inevitably find and report it and it'd get immediately corrected.
One of the developers thought I was being paranoid and sat down to have me write up a report of a bug he'd found and when it was closed, he threw a fit, but she just kept doing it until I willingly changed departments.
Corporate software is one thing, but an open-source VoIP library is quite another. If the maintainers don't understand why using different UDP sockets for sending and receiving won't work if the client is behind several NATs and they don't want to accept the proposal of, like, three people to fix it with the patch being just two edits within one file...
Just checked - they finally did what was necessary in v2.0 though. I wonder who (or what) persuaded them to make that extremely difficult two-line change.
It just had to be the right person suggesting the change. Most repos are like that. They exist to facilitate gatekeeping.
They're a quasi-corporate structure. A code central committee.
