Subject: CVS commit: wip/baler
From: crocusino
Date: 2010-12-31 17:24:38
Message id: E1PYhmf-0006pp-W8@sfp-cvsdas-2.v30.ch3.sourceforge.com

Log Message:
Import baler-1.0 as wip/baler.

A simple tool which takes a set of files (e.g. encrypted compressed tar files
or just any files), stores them Byte by Byte into volumes, which I call bales
(just because the data are stored Byte by Byte similarly like hay is put into
bales), and generates a convenient table of contents (XML file) describing what
and where is stored in this balestack (i.e. the set of bales - the whole
archive). To get a particular file from the archive, we just look it up in the
XML table of contents, take the right bale (i.e. mount the right DVD) and
simply copy and paste the position and size to baler and that's all, folks!
Easy!

Typically I dedicate four DVDs as my archive media. Three of them contain bales
and the fourth one is used for the redundant data (so that a lost DVD can be
regenerated)

I prepare a few tar.xz.enc (i.e. encrypted compressed tarball) files and use
baler to create three bales and a table of contents. Then I use eor tool
(Exclusive OR generator for ordianary files) to generate the redundant data
from the three bales. Finally I just burn each bale plus the TOC file on the
corresponding DVD (multisession) and the redundant data plus TOC on the fourth
DVD. This process can be repeated with new balestacks until the DVDs are full.

This way the media are filled efficiently and the data are stored safely. When
I want a particular file I just open the TOC and see which of the DVDs I need.
And baler then gets the desired file from one or more bales very quickly.


Files:
RevisionActionfile
1.1importwip/baler/distinfo
1.1importwip/baler/Makefile
1.1importwip/baler/PLIST
1.1importwip/baler/DESCR
1.1importwip/baler/patches/patch-aa