For RedHat you can install packages several ways.
For each of these
su -
<enter root pwd>
cd /to/wherever/the/files/are
(you can hit TAB to complete the pathname as you go or double-TAB for a list of possibles, e.g. typing cd /home/unfo then hitting TAB would complete the /home/unforgiven for you)
1) Ready-made RPM, choose one *built for your version of RedHat*, and either double-click the RPM file from the desktop or
rpm -ivh mynewapp.rpm
(= install, verbose, and show progress bar)
If you need other dependencies and download them as rpms you can install them all on the same line so you don't have to worry about which order to install them in e.g.
rpm -ivh mainapp.rpm somelibrary.rpm anotherone.rpm somethingelse.rpm
2) Source RPMs which are source code packaged up to build INTO an RPM package for your system. The benefit here is that seeing as it is built on YOUR system it should work.
rpmbuild --rebuild newapp.src.rpm
and NOTE where the RPM file is put!
Install as above from there.
3) Get source code and compile up. Will usually come as a .tar, .tar.bz2, .tar.gz file or something like this. Tar, gzip and bzip2 are often used to compress files and together, so a file may be a tar file which is THEN zipped too.
To uncompress use
bunzip2 filename.tar.bz2
or
gunzip filename.tar.gz
then
tar -xvf filename.tar (x-tract, verbose, file).
Note if your version supports it (usually does), tar can pass thru gzip or bzip2 files automatically like this:
tar -xvzf filename.tar.gz (gzip)
tar -xvjf filename.tar.bz2 (bzip2)
If you prefer GUI stuff use Ark which is available from the menus for unarchiving (if you double-click an archive file this is what runs).
When the tar file is extracted you will usually have a new directory to cd into. Look in there and there will usually be a configure file or MAYBE something like an install.sh file. The usual process is
cd install_dir
./configure
make
make install
the ./ part is because the shell looks for commands in the "command path" and nowhere else. You are in a directory that is NOT part of the command path so if you type a command from in there (e.g. configure) the shell will not "see it" automatically. A single "." in linux means "the current directory" so you are saying "run configure, which is in the current directory". You do not need this for "make", because the make command itself is stored somewhere that is already included in the command search path.
If there is something like install.sh try
sh install.sh
or something like install.bin do
./install.bin
however for source code it is usually done with configure and make as above.
Whatever, there should be a README file, so
cat README | more
will print this to the screen, hit space for more pages, this will tell you how to install.
Once whatever method is done, go back to normal user with exit, type the command name and see if it runs ok.
You can type
command_name &
to run in the background too to carry on typing in the shell.
If it runs ok you can add it to your menus, use the menu editor, add it as a new item, choose an icon for it, in the command field just enter the command that runs it, Apply and it is ready on your menu (now you can add it to your "task bar" equivalent as a button or whatever).
If you prefer pre-built RPMs to source "tarballs" (.tar.gz files etc.) you will find many kind souls build these up for different distros. Have a look on the net at places like
FreshRPMs or use
RPMfind I'm sure there are others too.
Hope this helps.
