Software Engineering on Unix

tar

If you download software source code on unix it will likely be in a "tarball".
Likewise, if you want to give other people your code, you want to give them a tarball.

On Athena, GNU tar can be accessed with:
```
	add gnu;  gtar
	
```
Archives contain multiple files
Extract an archive foo.tar with:
```
	tar -xf foo.tar
	
```
Files will be extracted relative to current directory
List the contents of an archive foo.tar with:
```
	tar -tf foo.tar
	
```
Using -v option gives more verbose output, such as files being tarred
Archives can be compressed with gzip or compress
Gzipped tar archives end in .tar.gz or .tgz
Compressed tar archives end in .tar.Z or .taz
Using -z option to gtar will make it automatically compress or decompress archives (using gzip)
Using - as archive filename makes it read from stdin

With standard tar, uncompress and untar with:

	gunzip -c -d foo.tar.gz | tar -xvf -

or with gtar:

	gtar -xzvf -

To create an archive named foo.tar containing dir1 and file2, we'd do:
```
	gtar -czf foo.tar.gz dir1 file2
	
```
The arg following the -f is the archive file

gzipped tarballs
Named program-version.tar.gz (ie screen-3.9.11beta5.tar.gz)
Contains only sourcecode.
Extracts into a directory named program-version (screen-3.9.11beta5/). Doesn't put stuff in the current directory other than that file.
To build one:
- Put source into a directory named properly.
- Run
```
  tar -czf screen-3.9.11.tar.gz screen-3.9.1/
  
```
A file named README at the top of your source tree is good to tell people what to do with it.