[Rcpp-devel] A beginners guide to Rcpp

[Walter Mascarenhas]

   I don't mean to come down hard on you personally, but consider that
listening to criticism is a good way to keep a balanced view of things.

  It is natural for you to defend rcpp and your views on R and C++,
However I do believe Gabor points are also relevant, in the same
way that Knuth, Linus and Wirth may have some reason in
their disdain for C++.

  Rcpp is a worthwhile effort and your work is admirable, but
it is not perfect. In order for it to evolve you should also
consider carefuly the problems it will face in order to
be used by many, many peope with different backgrounds
and levels of skill (if you care about that, of course).

             walter.


On Thu, Nov 29, 2012 at 10:43 PM, Dirk Eddelbuettel <edd at debian.org> wrote:

>
> On 29 November 2012 at 19:08, Gabor Grothendieck wrote:
> | That is only the case if you don't know C or C++ .  Many people who
> | are using these interfaces already know C and in that case the  .C
> | interface is trivial (its just a simple interface) whereas Rcpp is a
> | new environment you have to learn to program in.
> |
> | The trivial amount of time to learn .C for anyone who knows C is
> | hardly comparable to the effort to learn a new environment even if
> | that environment is a nice one.
>
> I don't mean to come down hard on you personally but consider that
>
>   a) All this is irrelevant in the context of this list. We do Rcpp here,
> it
>      requires .Call. We do what we do here because we want to send _whole R
>      objects around_ and not "K & R 1987 style" double vectors with a
>      counter.
>
>   b) There was recent thread on r-devel, and I would kind ask you to send
> you
>      defence of .C() there.  As I read the tea leaves, .Call() is it with
>      changes for R 2.16.0 and beyond and larger indices.  .C() is legacy
> and
>      (per that thread) discouraged for new projects.
>
> I suspect that I personally have not used .C in my programming in a decade.
>
> Dirk
>
> --
> Dirk Eddelbuettel | edd at debian.org | http://dirk.eddelbuettel.com
> _______________________________________________
> Rcpp-devel mailing list
> Rcpp-devel at lists.r-forge.r-project.org
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rcpp-devel/attachments/20121130/76729aa0/attachment.html>