OneSignal/onesignal-ruby-api

[Bug]: Crashes when used in a default rails/puma server on macOS

Opened this issue ยท 1 comments

What happened?

When trying to send a notification, I get this:

objc[65540]: +[__NSCFConstantString initialize] may have been in progress in another thread when fork() was called.
objc[65540]: +[__NSCFConstantString initialize] may have been in progress in another thread when fork() was called. We cannot safely call it or ignore it in the fork() child process. Crashing instead. Set a breakpoint on objc_initializeAfterForkError to debug.

And no notifications are delivered

Steps to reproduce?

Trigger a notification from within puma, I guess

Root cause is:
https://github.com/typhoeus/typhoeus/issues/687
and
https://github.com/typhoeus/ethon/issues/206

But these are longstanding open issues. Perhaps Onesignal should reconsider using a library with a longstanding crash?

What did you expect to happen?

obvious

Relevant log output

No response

Code of Conduct

  • I agree to follow this project's Code of Conduct

Agree 200% ๐Ÿ™ please consider using some http lib w/o libcurl native extensions. Unfortunatelly they are all prone to crashing with ruby libs - ethon, curb, practically for many years.

๐Ÿ™ default net::http should be good-enough as a gem dependency, thanks.