GZip is a form of data compression — i.e. it takes a chunk of data and makes it smaller. The original data can be restored by un-zipping the compressed file.
It is relevant to web apps and websites because the HTTP protocol includes the ability to gzip data that is being sent.
This means that, when it is in use, your bandwidth costs for serving the site will be lower because people visiting the site will be downloading smaller files.
There are a few caveats to using GZip, but, overall, it’s usually better to use gzip than not to — for example, it does take time and processor power to zip and unzip the files, but, typically, this is not a problem because the time it takes to do that is often less than the time that is saved by downloading a larger file. Therefore, the overall effect is time-saving despite the browser having to unzip the file.
GZip can compress all files; it doesn’t make any difference what the file type is or the encoding. Obviously, some files can be compressed more effectively than others, so the bandwidth saving will vary – text files like HTML give the best results; images are not compressed so much by gzip because they already have some compression built-in. Some files (e.g. those that are already heavily compressed like .zip files) may actually get slightly bigger when gzipped because they can’t be compressed any further but gzip still needs to add its metadata to the file. But these are edge cases and don’t make much difference.
GZip across HTTP normally happens completely transparently. The end-user should be completely unaware that it is happening; the browser would do it behind the scenes for them. And from the webserver end, it is simply a matter of turning on a config setting in your web server software. From your perspective, that’s really all you need to know; just set the gzip setting on your server (or ask your ISP to do it). It’s quite possible it may already be active on your site without you even knowing.