Hervé Pagès

Thanks for your patience @gkanogiannis. Package looks good. Only minor cosmetic issue is this: ``` > library(metabinR) To cite metabinR in publications use: A BibTeX entry for LaTeX users is...

All looks good. Thanks!

Would be nice to have 2 env. variables for this e.g. `BIOCPARALLEL_DEFAULT_WORKER_NUMBER` and `BIOCPARALLEL_MAX_WORKER_NUMBER`. `BIOCPARALLEL_DEFAULT_WORKER_NUMBER` would just be a rename of `BIOCPARALLEL_WORKER_NUMBER`, and it would keep doing what the latter...

I just noticed that too earlier today after doing `library(BiocStyle)` on one of the build machines. Looks like the `.onAttach` hook was replaced by an `.onLoad` hook (see commit 1fed5289f00871b8b4d42736c7c329378da21627)....

Maybe best practice should be to explicitly `library()` the package before using it interactively? That's kind of the standard thing to do, for any package.

Thanks Marcel. If `BiocManager::install()` is not going to reinstall those packages by default then the command suggested by `BiocManager::valid()` would need to have `force=TRUE`.

Also the documentation for the `force` argument should clarify what happens in that case. Right now it kind of suggests that using `force=TRUE` is only needed for reinstalling packages that...

Isn't it that all the packages that `BiocManager::valid()` report as being either out-of-date or too new can be installed with `force=TRUE`? I understand that `force=TRUE` is not needed to install...

Great! Thanks Marcel. Just a note about the use of `character(0)` vs `""` in the context of `paste()`/`paste0()`: ``` paste0("X", character(0)) paste0("X", "") ``` Both work and do the same...

I dunnow. I doubt this is how you would teach a 12-year old how to represent an empty string in R :grin: