Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

header only and dependency #13

Closed
romainfrancois opened this issue Mar 1, 2014 · 3 comments
Closed

header only and dependency #13

romainfrancois opened this issue Mar 1, 2014 · 3 comments

Comments

@romainfrancois
Copy link
Contributor

It would be nice if RcppArmadillo was header only and did not depend on Rcpp. The dependency can be resolved in client packages, e.g. a package can choose to use the Rcpp implementation in Rcpp11 plus armadillo from RcppArmadillo.

The alternative is for me to fork RcppArmadillo.

@romainfrancois
Copy link
Contributor Author

This means letting go of fastLm and other things in src/. Not sure this fits with the maintenance policy. If it does not, then close the issue and expect a fork.

One of the plus is that the combo Rcpp11 + RcppArmadillo could leverage C++11 capabilities of armadillo.

@eddelbuettel
Copy link
Member

R(-devel) has supported C++11 for a while. With R-devel today, which becomes R 3.1.0 in six weeks on April 10, you only need the USE_CXX1X define in src/Makevars which I built upon to turn on Armadillo's C++11 support. (It's commented out now in the repo, but just remove the one comment, build a tarball and use it via R-devel locally or via win-builder).

I also do not see R Core moving away from having the toggle in src/ which makes the quest of removing all files from src/ a little less compelling. I am in no hurry here and suggest that we wait til R 3.1.0 is out, let the dust settle a little and review then -- no need to rush things.

As for removing src/ from RcppArmadillo and RcppEigen: we could, of course, move fastLm() to add-on example packages. But as @dmbates said in RcppCore/RcppEigen#3, we do have these documented in published papers and the book, so user-facing changes of documented interfaces and functionality must be merited. I don't see that quite yet.

@romainfrancois
Copy link
Contributor Author

Ok. I have my answer. Closing this now

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants