[clug] Request for Ideas/Suggestions: Fedora Scientific Spin

[Robin Humble]

On Sun, Jul 03, 2011 at 09:16:27PM +1000, Brad Hards wrote:
>On Sunday 03 July 2011 13:31:45 Amit Saha wrote:
>> It would be of great help if the interested ones among you suggest some
>> generic tools for numerical computing/scientific research that you find
>> really useful which is not already included in the list of packages in
>> the above wiki page.
>I note gcc in the list, but a C++ compiler might also be useful. Maybe MPI.

openmpi definitely. more science == more cpu power, and cpus aren't
getting faster so parallel is the only option.

gfortran is necessary for all those 1970's fortran codes that are
(sadly) still going strong.

(p)netcdf, hdf5, and their python interfaces are widely used across
disciplines. pnmtools remains awesome, as does xv. some people use
imagemagick but it's very inefficient.

random others ->
boost, petsc, ffmpeg, mencoder, ipm, gmp, octave, valgrind, fftw

as you can see I'm not limiting myself to purely gpl codes, but these
are what always gets installed on science machines, so hey...

I guess you can't include the intel compilers and idl? :)

torque, maui, powerman, conman, pdsh/c3, modules, and maybe slurm for clusters.
not many clusters run fedora though - it obsoletes too fast. ISV codes and
cluster folks (eg. me) want stability for a few years so run rhel/centos/sl +
maybe epel. but as fedora flows on into rhel, I thought I'd mention it.

>several - git, bzr, hg) probably should be there too. Perhaps a GUI for the 
>DVCS.

a lot of folks seem to use cvs/svn still, while the new kids seem to
prefer github.

>Depending on how you see research, perhaps something for document management 
>(e.g. bibliography, perhaps some kind of tool to stash PDFs of reference 
>papers) might be useful.

bibtex, latex, and I forget what the latex gui is - never used it.

cheers,
robin