I decide to rewrite flyback 0.4 code for a new project

by Forrest Sheng Bao http://fsbao.net

Flyback (http://code.google.com/p/flyback/) is a very good backing up tool for Linux like Time Machine for Mac.

But I really cannot stand the versions of flyback after v.0.4.

1. It uses Git instead of rsync (before v.0.4). Not to argue that rsync is small (file synchronization only) and Git is big (features for source code management), Git cannot handle files larger than the memory. I do medical image and signal processing. I have a lotta large files. 
2. flyback after v.0.4 is too complex. I can't even find where I can exclude a directory in backup. But Git actually provides such a function. I don't have the luxury to study the complex code to modify it. 
3. Icannot set up timed backup now. For example, I want the code to be backed up twice a day.

I looked around other open source backup solutions. They are all too complex for me. After reading their websites for 1 minute, I still don't know how they work. For desktop users' simple requirement, we need simple solution: I tell the problem which directories to or not to back up and the program will back up for me incrementally, and even periodically.

So I decide to work from flyback 0.4 source code which uses rsync and cp to do incremental backup. And create a new project which also uses GPL v2.

1. I wanna write a command line version of flyback 0.4. It will be designed very simple, just reading my configurations from a text file and calling basic Linux commands, e.g., rsync, chmod, cp and rm. This will be very easy and very useful on servers. 
2. I will simplify the flyback 0.4 code and add some new features, such as backup thru network using SSH. 
3. In such a way, the code can also work on Mac OS X. I can't stand the robustness of Time Machine. It sucks. 

I am quite busy now but I will slowly move this project as a hobby - which means I will have to eliminate my time on hanging out with girls :P .