Revision control system for library backup

[Man Eating Duck]

Hi! Does anyone else use this for backup? I do regular file backups, but I'd like to have a changelog and efficient storage/recovery of different file versions. Some kind of revision control system seems good for this, but I'm not too well versed in the various options, and am grateful for any tips

A good gui is a definite plus, but not a must. I'm interested in automating it (committing every day if changed) eventually, but I presume that most systems can do everything from the command line, making it script/cron-able.

I've looked at svn, but its approach to storing metadata in .svn-directories makes it a bit messy with Calibre. The system must also handle directory renaming with no file changes well. I've looked at git, which seems promising, but browsing for older versions is very cumbersome. Other candidates are Mercurial and Bazaar.

I'm on Linux by the way. If anyone already does this I'd be interested in hearing about your experience with abusing VCSes

[user_none]

calibre development uses bazaar for managing its source code.

[Man Eating Duck]

Originally Posted by user_none

calibre development uses bazaar for managing its source code.

Yes, I know, I've only ever used bazaar to check out calibre source

It seems that Bazaar doesn't handle folder renaming without special treatment; it also seems that git, for instance, works around this by using sha1 for identifying files. If I change "Lovecraft, H. P." to "Lovecraft, Howard Philip" I would end up with duplicates of all his books. As I'd probably do with git as well as the epubs are updated with new metadata, I guess.

Maybe it wasn't such a good idea after all

[Pistos]

If you are not a developer, and don't use version control on a regular basis, I would say it is too advanced (i.e. more trouble) for you. Try a regular backup system, like duplicity or plain rsync.