Hi there! I recently ran into some trouble installing ruby on a fedora vm, and in the process of resolving the build errors I ran asdf install several dozen times.
Every time I attempted to install ruby with asdf, it downloaded a new collection of artifacts into a timestamped folder. This slowed down my debugging process, and a handful of times in the process I ran into "no space on device" errors and found that the disk backing /tmp was full. Manually clearing out the cache is easy enough, but seems like it should be unnecessary.
I posted on ruby-build about the process and the maintainer there mentioned RUBY_BUILD_CACHE_PATH. A quick search here didn't turn anything up -- perhaps this would be an easy way to decrease the suffering of someone who is already fighting with the computer to get ruby installed?
Hi there! I recently ran into some trouble installing ruby on a fedora vm, and in the process of resolving the build errors I ran
asdf install
several dozen times.Every time I attempted to install ruby with asdf, it downloaded a new collection of artifacts into a timestamped folder. This slowed down my debugging process, and a handful of times in the process I ran into "no space on device" errors and found that the disk backing /tmp was full. Manually clearing out the cache is easy enough, but seems like it should be unnecessary.
I posted on ruby-build about the process and the maintainer there mentioned RUBY_BUILD_CACHE_PATH. A quick search here didn't turn anything up -- perhaps this would be an easy way to decrease the suffering of someone who is already fighting with the computer to get ruby installed?