What is FuseCompress?

FuseCompress provides a mountable Linux filesystem which transparently compresses its content.

Files stored in this filesystem are compressed on the fly and FUSE allows to create a transparent interface between compressed files and user applications.

FuseCompress supports different compression methods: LZO, gzip, bzip2, and LZMA.

Isn't there a new version?

This is the legacy C implementation of FuseCompress (0.9.x tree). It performs slightly less well than the current one implemented in C++ (1.99.x tree), but does not exhibit flaws such as data corruption or files growing larger and larger that make the current code (as of July 17 2008) unusable in a production environment. You can find the new version at Milan Svoboda's website. Because Milan is concentrating on the new version, I am now maintaining the old tree.

How well does it work?

It works well enough that I have trusted it to handle the vast majority of my private data. I am also using it for my $HOME directory, source trees, and I have (with some really evil hacks) succeeded in installing an openSUSE 11.0 system with FuseCompress as its root filesystem. (That, however, is not something you should currently try at home. At least not on a real system.)

LZMA support should be reasonably stable. The LZMA format has changed between versions 4.999.3 and 4.999.5 of the LZMA library, but it is stable now.

I am still occasionally fixing slight deviations from the expected behavior of a Unix filesystem. These problems are all minor, but may still cause data loss, depending on how robust your applications are.

FuseCompress now has experimental support for Mac OS X and MacFUSE.

Where to get it

• Packages for openSUSE
• Source code can be found in the SVN repository.

More information

• Build HOWTO
• How to use FuseCompress
• What compression method should I use?