Questions during test of EasyBuild v1.2.0rc1

Questions_on_EB_v1.2.0rc1

OK, the log below came out of redirecting stdout of an automated eb run with various easyconfigs:

```
[...]
== COMPLETED: Installation ended successfully
== Results of the build can be found in the log file /opt/apps/HPCBIOS.20130226/software/Java/1.7.0_10/easybuild/easyb
uild-Java-1.7.0_10-20130226.221440.log
== processing EasyBuild easyconfig /opt/apps/default/software/easybuild/1.2.0rc1/easybuild-easyconfigs/easybuild/easyc
onfigs/r/R/R-2.15.2-ictce-4.0.6.eb
== building and installing R-2.15.2-ictce-4.0.6...
== fetching files...
== getting ready, creating build dir, resetting environment...
== unpacking...
== temporary log file in case of crash /tmp/easybuild-EhZasI.log
== temporary log file in case of crash /tmp/easybuild-4uozgZ.log
== resolving dependencies ...
== processing EasyBuild easyconfig /opt/apps/default/software/easybuild/1.2.0rc1/easybuild-easyconfigs/easybuild/easyc
onfigs/j/JasPer/JasPer-1.900.1-ictce-3.2.2.u3.eb
== building and installing JasPer-1.900.1-ictce-3.2.2.u3...
```

Questions:

1) Why is Java a dependency of R?

2) Is it the case above that the R build failed but no trace of the failure was visible in stdout?
   I think putting the tag "CRASHED", "EXITING" or "FAILED" would make it more clear for the random bystander...

```
== COMPLETED: Installation ended successfully
== Results of the build can be found in the log file /opt/apps/HPCBIOS.20130226/software/netCDF-Fortran/4.2-ictce-3.2.
2.u3/easybuild/easybuild-netCDF-Fortran-4.2-20130226.222154.log
== processing EasyBuild easyconfig /opt/apps/default/software/easybuild/1.2.0rc1/easybuild-easyconfigs/easybuild/easyc
onfigs/w/WRF/WRF-3.4-ictce-3.2.2.u3-dmpar.eb
== building and installing WRF-3.4-ictce-3.2.2.u3-dmpar...
== fetching files...
== getting ready, creating build dir, resetting environment...
== unpacking...
== temporary log file in case of crash /tmp/easybuild-gUr0wq.log
== resolving dependencies ...
== processing EasyBuild easyconfig /opt/apps/default/software/easybuild/1.2.0rc1/easybuild-easyconfigs/easybuild/easyc
onfigs/h/HDF5/HDF5-1.8.7-ictce-3.2.2.u3.eb
== building and installing HDF5-1.8.7-ictce-3.2.2.u3...
```

3) netCDF v4.2 is no longer available at the original website location; what are we going to do about this?

4) samewise, we need MTL4 - can UGent provide the open source tarball?

5) Is it the case above that the build of WRF-3.4-ictce-3.2.2 actually crashed? 
   (I guess due to osdependencies, is there an elegant way to give fuse just a little bit of error messaging in there?)

```
== Results of the build can be found in the log file /opt/apps/HPCBIOS.20130226/software/FFTW/3.3.1-ictce-4.0.6/easybuild/easybuild-FFTW-3.3.1-20130226.224019.log
== processing EasyBuild easyconfig /opt/apps/default/software/easybuild/1.2.0rc1/easybuild-easyconfigs/easybuild/easyconfigs/w/WIEN2k/WIEN2k-12.1-ictce-4.0.6.eb
== building and installing WIEN2k-12.1-ictce-4.0.6...
== fetching files...
== temporary log file in case of crash /tmp/easybuild-9jtaiw.log
```

6) Can we make it more clear that non open source software like Wien2K, 
   basically fails to build because the original bundles are behind paywalls/captchawalls?
   This will prevent someone thinking that EasyBuild has failed to do its job, when it is not the case.

7) Can we print the stack of dependencies somewhere, so that multi-build stdout is more understandable?

8) Does a "toolchain" build of, say, all bioinformatics tools, parallelize the task as desired? Ref. 
   https://github.com/fgeorgatos/easybuild.experimental/blob/master/users/fgeorgatos/HPCBIOS/HPCBIOS_2012-94-goalf-1.1.0.eb

9) R-2.15.2-ictce-4.0.6-bare still has problems, dammit, need to investigate.

10) EasyBuild does not build with EasyBuild for me.... sniff...

that's as of 20130227 @ 3 am

EOF!