Gábor Csárdi

I don't think this is a common issue, though. FWIW a good practice is to create user-library with ```r dir.create(Sys.getenv("R_LIBS_USER"), recursive=TRUE) ``` (and then a restart), and then using your...

@dragosmg pak could print some advice if it sees that it is installing into the system library? (Maybe not on Linux, because containers often use the single library.) IDK if...

> What is dir.create(Sys.getenv("R_LIBS_USER"), recursive=TRUE) meant to do? Create a directory for the current user's package library. (Don't do this as the root user, obviously.) If that directory exists then...

> I don't see the Sys.getenv("R_LIBS_USER") directory added to the .LibPaths(). Am I doing this correctly? You need to restart R to take effect, because R checks its existence at...

Yeah, it would have to be optional in general. OTOH, when installing a windows-only package on Unix fails, that should just work by default, maybe with a message.

The plan is to use the `?` notation to tell pak to ignore a failure in the dependencies, e.g. ```r pak::pak(c(".", "badpackage?ignorefailure")) ``` `?ignorefailure` ignores that package if it fails...

I thought about that, but where is the harm in doing it for non-interactive sessions? pkgman will already print messages about the packages to be installed, etc.

Technically it would not be a warning, just a message. Sorry, the issue title is confusing. We would also add an env var to suppress it. pkgman will probably have...

Yeah, we can check it every time the metadata is updated, and `alert()`.

There is no way to know the size of GitHub packages.