Re: [R-pkg-devel] Weird MAKEFLAGS bug

2017-01-11 Thread Kasper Daniel Hansen
For a package, you should not mess with the optimization flag unless you are disabling high optimization because it produces wrong code. If people want to use higher optimization flags, they'll compile with higher flags. On Tue, Jan 10, 2017 at 3:39 PM, Costas Tsirogiannis wrote: > Thanks a lot

Re: [R-pkg-devel] Weird MAKEFLAGS bug

2017-01-10 Thread William Dunlap
MAKEFLAGS='CXXFLAGS=-std=c++0x -O2' R CMD SHLIB passes two items, 'CXXFLAGS=-std=c++0x' and '-O2', not the single item 'CXXFLAG=-std=c++0x -O2', to make. Suitably recent versions of GNU make take the make argument -O2 to control the 'output-sync'. Bill Dunlap TIBCO Software wdunlap tibco.com

Re: [R-pkg-devel] Weird MAKEFLAGS bug

2017-01-10 Thread Costas Tsirogiannis
Thanks a lot, indeed it worked using the Makevars file. I still do not understand why the command-line input fails but so be it. One more quick question: if I include the Makevars file in my package code then the requested flags are used every time somebody installs my package in his system. Howev

Re: [R-pkg-devel] Weird MAKEFLAGS bug

2017-01-10 Thread Dirk Eddelbuettel
On 10 January 2017 at 11:55, Costas Tsirogiannis wrote: | Greetings and thanks for previously offered solutions, | | I am developing a package that makes use of C++ code, and I want to compile | this code for testing reasons with the "R CMD SHLIB" command. Using Ubuntu | and gcc 5.4, and because

[R-pkg-devel] Weird MAKEFLAGS bug

2017-01-10 Thread Costas Tsirogiannis
Greetings and thanks for previously offered solutions, I am developing a package that makes use of C++ code, and I want to compile this code for testing reasons with the "R CMD SHLIB" command. Using Ubuntu and gcc 5.4, and because for certain reasons I have to use the c++0x standard, I try the fol