Closed zspencer closed 9 months ago
We were getting this and we tracked it down to a gem version issue.
We had (for various old reasons) gem "rails", "= 7.1.0"
. Once we loosened that up to gem "rails", "~> 7.1.0"
it properly upgraded our dependencies and this issue went away.
@zspencer YMMV but that worked for us.
Thanks for taking time to work on this. I took an alternate approach in #492 and just made ReplayLogger
extend Logger
so it has all the functionality of a real logger.
That sounds much wiser than my attempt! Thank you!
I'm not entirely sure if this is the "right" way to fix this (in fact, I'm pretty sure it's not...) I'm presuming that setting the
ReplayLogger#level
to log the most things will ensure it doesn't drop messages.I noticed on a project that since the 3.0 release of
dotenv-rails
, the application would fail to start due to theReplayLogger
responding to#level
as if it were a command to log a statement, rather than a query to determine the level theLogger
is configured for.You can see the exception here: https://github.com/zinc-collective/convene/actions/runs/7923719579/job/21633994956?pr=2210
The application in question is a Rails 7.1 app, running Ruby 3.2.