Well 'make' is a command to build software from source code.
Generally you need to have these "-dev" packages installed for certain peices of software and you need to have things like gcc compiler installed and such.
The most generic method that most packages use would go something like this:
# wget
http://something.som/files/sourcecode.tar.gz
And that will download the item.
# tar zxf sourcecode.tar.gz
and that will unzip the item into sourcecode/ directory, then
# cd sourcode/
# ./configure
And that will run a script to configure the software your going to compile. Frequently it will stop and complain about missing items. Something like:
error: sdl-mixer can not be found..
Then you have to find the ubuntu package that contains the library header files that it needs..
# apt-cache search sdl | grep mixer
Then it will output all packages with sdl in it's name and then you grep through that output and find all lines with 'mixer' in them. The packages that contain the 'headers' will end with a -dev
You will then need to install those like this:
# apt-get install libsdl1.2-mixer-dev
(or whatever the package name is.
Then you try to re-run the ./configure script, and then rinse and repeat till you have all the development tools and header packages you need installed. Then you compile the packages...
# make
Then either you manually copy the binaries to the correct space or do:
# make install
And that will install the package.
That's the generic method for installing programs from source code under linux.
However there are many many variations. Sometimes they don't have a configure script and have a manually built Make file that the author made sure would work for everybody (to the best of their ability) and you just run 'make'. Sometimes they use a different build system and doesn't use 'make' at all. Sometimes they'll use something like scons or whatnot.
The only way you can figure this out is to read the README files and INSTALL files or any other documentation aviable on the software's website.
It's very difficult at first, but after doing it a few times then most times compiling software is easy... although obviously not as easily as installing from a package manual. You'll always run into software that is to difficult or just won't work. It happens.
For this paticular package it looks like the author just wants you to run 'make' to compile the software.
So just untar the source code, change to that source code's directory and run 'make'.
If you have a tarball that ends with 'tar.gz' or '.tgz', then that is a tar archive compressed with gzip. If you have a tarball that ends with tar.bz2 then that is compressed with bzip2.
So with filename.tar.gz
you go:
# tar zxf filename.tar.gz
With filename.tar.bz2 you go
# tar jxf filename.tar.bz2
For the second part...
In Linux there is no registery file. Generally the only directories you can access are your home directory and your temp directory.
~/ is shorthand for your /home/username directory, which is your home directory. That way you can do commands and such without resorting to typing out your home directory everytime or dealing with variables.
./ is shorthand for 'this directory' and
../ is shorthand for 'parent directory'.
So since there is no universal place to store configurations where everything is stored in 'hidden' files in your home directory.
These hidden files have a . before their name, that is what makes them 'hidden' in Unix/Linux. This doesn't actually make the files invisable or anything it is just a thing for convience. Normal commands like 'ls' will just skip over them so that you don't have to look at them.
These .filename and .directoryname files are generated automaticly when you run the programs for the first time. If you mess up a configuration all you have to do is either delete these files or move them to a backup directory and then re-run the prgoram and then it has all of it's defaults reset.
To see these directories you go:
# ls -lad ~/.*
So when you run gkrellm it automaticly makes a ~/.gkrellm2 directory. In there there will be a plugins directory.
So basicly you build this paticular peice of software with:
make
Then when it finishes you find the binary file that you built and then copy it to that ~/.gkrellm2/plugins directory.
Don't worry if it's a bit confusing. It's easy after a little bit. It's just one of those big differences between Linux and Windows.
If you want a different explaination of what is going on or better details then check out my sig link and look at 'hands on guide" to linux.