As I've gone and added more and more steps to the rworkflows actions, I haven't really paid much attention to the R package caching steps. These were originally adapted from biocthis, and use the actions/cache action.
I'm now reevaluating these steps in their current context and trying to rearrange them to improve overall speed.
Something I noticed in biocthis is that caching doesn't occur on Linux. I'm assuming there's a reason for this, but I'm not sure what it is. Would you mind elaborating on this @lcolladotor ?
I'd also like to explore caching software other than just R packages (eg system deps, latex packages). I'm not certain, but I don't this is being done currently.
As I've gone and added more and more steps to the
rworkflows
actions, I haven't really paid much attention to the R package caching steps. These were originally adapted frombiocthis
, and use theactions/cache
action.I'm now reevaluating these steps in their current context and trying to rearrange them to improve overall speed.
master branch (before): https://github.com/neurogenomics/rworkflows/blob/master/action.yml dev branch (latest tests): https://github.com/neurogenomics/rworkflows/blob/dev/action.yml
Something I noticed in
biocthis
is that caching doesn't occur on Linux. I'm assuming there's a reason for this, but I'm not sure what it is. Would you mind elaborating on this @lcolladotor ?https://github.com/lcolladotor/biocthis/blob/d86021c7e642a684240bcdfd67dd371198988f48/.github/workflows/check-bioc.yml#L104