Untar the binary files:
gzip -d adf2005.01.bin.tgz | tar xf -If you have a license that includes the source code and you have downloaded it:
gzip -d adf2005.01.src.tgz | tar xf -
We recommend that you define the following environment variables in your shell startup script (.cshrc, .login, .zshenv, ...):
# Environment variables with typical values (you may need different values) using C-shell syntax. # Add these to your .cshrc setenv ADFHOME $HOME/adf2005.01 # here $HOME is the directory # where you unpacked the downloaded files setenv ADFBIN $ADFHOME/bin # the directory for run scripts and program binaries setenv ADFRESOURCES $ADFHOME/atomicdata # the location of the ADF database setenv SCMLICENSE $ADFHOME/license # the location of your SCM license file (when you get it) setenv NSCM 16 # the default number of parallel processes to use # These environment variables are needed for PVM installations only # ================================================================= setenv SCM_TMPDIR /tmp # the scratch directory for parallel child processes setenv SCM_USETMPDIR yes # use a scratch directory for the master process too? setenv SCMSPAWNSCRIPT $ADFBIN/adfs # the script to spawn the parallel child processes setenv PVM_ARCH SGI6 # the PVM architecture name setenv PVM_ROOT /usr/local/bin/pvm3 # the master directory of your PVM installation
# generate license information $ADFBIN/adf << EOR EOR
and follow the instructions (send us the information just printed). In principle, you will need to send us the information from all machines on which you intend to use the programs from the ADF package.
After receiving your license file, copy it to the file referred to by the $SCMLICENSE environment variable.
That is the place where the programs will be looking for it.
If everything is working, you can now remove the tar files.
Read the Installation manual for more information. It contains advise on how to optimize your ADF installation for your particular situation. You will also need to read it to configure your installation properly for parallel execution.