| 1 |
Mark Knecht wrote: |
| 2 |
> For kicks I decided to give -fortran a try. I find it interesting that |
| 3 |
> with Fortran installed using either the fortran flag in package.use, |
| 4 |
> or just not setting the flag at all, I apparently do not need the c++ |
| 5 |
> libraries that I do need when I choose -fortran. However it seems that |
| 6 |
> -fortran doesn't eliminate the need for installing a Fortran compiler. |
| 7 |
> It simply shifted it from the GNU gcc Fortran package to the Intel |
| 8 |
> Fortran Compiler (dev-lang/ifc) |
| 9 |
> |
| 10 |
> c2stable ~ # emerge -pvDuN @world |
| 11 |
> |
| 12 |
> These are the packages that would be merged, in order: |
| 13 |
> |
| 14 |
> Calculating dependencies... done! |
| 15 |
> [ebuild N ] sys-libs/libstdc++-v3-3.3.6 USE="(multilib) nls" 23,459 kB |
| 16 |
> [ebuild N ] virtual/libstdc++-3.3 0 kB |
| 17 |
> [ebuild N ] app-emulation/emul-linux-x86-compat-20100611 |
| 18 |
> USE="(multilib)" 930 kB |
| 19 |
> [ebuild N ] dev-lang/ifc-10.0.026-r1 40,378 kB |
| 20 |
> [ebuild R ] sys-devel/gcc-4.4.5 USE="gtk mudflap (multilib) nls |
| 21 |
> nptl openmp (-altivec) -bootstrap -build -doc (-fixed-point) -fortran* |
| 22 |
> -gcj -graphite (-hardened) (-libffi) -multislot -nocxx -nopie -nossp |
| 23 |
> -objc -objc++ -objc-gc -test -vanilla (-n32%) (-n64%)" 61,647 kB |
| 24 |
> |
| 25 |
> Total: 5 packages (4 new, 1 reinstall), Size of downloads: 126,413 kB |
| 26 |
> c2stable ~ # |
| 27 |
> |
| 28 |
> I have a virtual/fortran driven AFACT through any number of dependencies: |
| 29 |
> |
| 30 |
> c2stable ~ # eix -I fortran |
| 31 |
> [I] virtual/fortran |
| 32 |
> Available versions: 0 {openmp} |
| 33 |
> Installed versions: 0(06:12:54 AM 06/22/2011)(openmp) |
| 34 |
> Description: Virtual for Fortran Compiler |
| 35 |
> |
| 36 |
> c2stable ~ # eix -Ic fortran |
| 37 |
> [I] virtual/fortran (0@06/22/2011): Virtual for Fortran Compiler |
| 38 |
> c2stable ~ # equery depends fortran |
| 39 |
> * These packages depend on fortran: |
| 40 |
> dev-python/numpy-1.6.0 (lapack ? virtual/fortran) |
| 41 |
> sci-libs/blas-reference-20070226 (virtual/fortran) |
| 42 |
> c2stable ~ # equery depends numpy |
| 43 |
> * These packages depend on numpy: |
| 44 |
> dev-python/pygtk-2.22.0-r1 (dev-python/numpy) |
| 45 |
> c2stable ~ # equery depends pygtk |
| 46 |
> * These packages depend on pygtk: |
| 47 |
> dev-python/pygtksourceview-2.10.1 (>=dev-python/pygtk-2.8:2) |
| 48 |
> dev-python/twisted-11.0.0 (gtk ?>=dev-python/pygtk-1.99) |
| 49 |
> dev-vcs/git-1.7.3.4-r1 (gtk ?>=dev-python/pygtk-2.8) |
| 50 |
> x11-misc/driconf-0.9.1-r1 (>=dev-python/pygtk-2.4:2) |
| 51 |
> c2stable ~ # equery depends blas-reference |
| 52 |
> * These packages depend on blas-reference: |
| 53 |
> virtual/blas-1.0 (sci-libs/blas-reference) |
| 54 |
> c2stable ~ # equery depends virtual/blas |
| 55 |
> * These packages depend on virtual/blas: |
| 56 |
> dev-lang/R-2.12.2 (virtual/blas) |
| 57 |
> c2stable ~ # |
| 58 |
> |
| 59 |
> In my case I'm sticking with the gcc fortran package as it's the devil I know... |
| 60 |
> |
| 61 |
> - Mark |
| 62 |
> |
| 63 |
> |
| 64 |
> |
| 65 |
|
| 66 |
I emerge -C ifc and revdep-rebuild rebuilt some stuff but it is gone |
| 67 |
now. So, you may have to work with it but it *might* be removable too. |
| 68 |
|
| 69 |
Dale |
| 70 |
|
| 71 |
:-) :-) |
Archives