A tar ``image'' of the source code will be either on tape, or is already availableG.2on disk. Extract it from tape (the name of the tapedrive in this example is /dev/rmt0):
% tar xf /dev/rmt0
or extract it from a disk tarfile (called nemo.tar here):
% tar xf nemo.tar or: % zcat nemo.tar.Z | tar xf - or: % gzip -dc nemo.tar.gz | tar xf - or: ftp> get nemo.tar.gz "|gzip -dc | tar xf -"
depending on the nature and location of the (compressed) tar file. It
may have been compressed (
.Z) with the standard UNIX compress(1) program, or the GNU utility
gzip. Instead of using uncompress(1), the
utility zcat(1) can be piped to tar(1) to
extract the tar file much more efficiently. You may even transfer the
file accross the (inter)net via the ftpG.3 utility and piping
it's output straight into tar to avoid having to store the tar
file on your local disk.
BE SURE to be in the $NEMO directory while doing this! On some UNIX systems it may happen that tar complains about not being able to create directories or such. You can probably ignore it, since SYSV tar wants to preserve ownership (stored by a user id, which may be unknown or even invalid on your target machine). The tar program may have a flag for this.
If you haven't set the NEMO environment variable (cf. Appendix ) now is really a good time to do it.
First add the following to your .cshrc file, somewhere before the PATH environment is defined:
setenv NEMO /usr/nemo (or whatever) source $NEMO/NEMORC
After having read the NEMORC startup file, a number of environment variables and alias will have been added to your current environment. Include the $NEMOBIN directory in your search path before any major system areas, e.g.:
set path = ( . $NEMOBIN $path)
and make it current:
Now that all source and documentation is in place, the installation can start. As said before, the top level $NEMO/Makefile can steer the whole installation process heirarchically, calling Makefile's in lower level directories and so on and so on.G.4