Here’s a simple fix for a minor annoyance in your Ruby environment.
I have some tests that rely on an external gem, and that gem has some deprecation warnings that have been fixed on master but not released in a new version yet. Since my app is in production, I don’t want to use an unreleased (and possibly buggy) version of a gem just to get rid of some deprecation warnings.
Even so, I get tired of seeing 50+ lines of warnings every time I run my test suite.
Niftily enough, there’s an envar I can set on each spec run that will silence these warnings:
(Tested working on my machine running Ruby 2.2.x, and according to this StackOverflow post, it also works on 2.1.x, 2.14.x, and 2.3.x.)
Great, it works! But… how?
In my Googling, I found little info about
RUBYOPT. One documentation page defined it thus:
Command-line options passed to Ruby interpreter. Ignored in taint mode (Where $SAFE is greater than 0).
Okay… Kind of helpful. So, it seems to me that whatever is stored in
RUBYOPT gets run as a flag on the Ruby program (in my case,
More digging found some official Ruby documentation about flags, but did not include
-W0. However, this helpful StackOverflow post tells the story. It appears that
-W works the same as
-w in that it enables verbose mode. The difference is that the capital
-W can then be further configured for exactly how verbose we want the logs to be.
-W0don’t show any warnings.
-W1show some warnings, but not Kernel warnings.
-W2show all warnings (
-W3and upward seem to do the same thing)
Enjoy some quieter logs! (Or louder, if that’s your thing.)