subversion's storing / diffing techniques

Hi,

I am studying computer science at Bonn / germany and writing my diploma
thesis about

"version control in an xml-based content management system".

I know about RCS/CVS that it stores the current version as is and that
it uses backward deltas
for older versions to reduce storage amount (after diffing the text
files). In the subversion
book I read that even in case of binary data subversion is capable of
not having to store the complete file.

Now my questions:
Is there any documentation about the way subversion computes what has
changed in a file?
When requesting an old version of a file - how is it reconstructed?
How does subversion reduce the amount of data it has to store?
What is the actual storing technique used by subversion?
What information is stored in the underlying Berkley DB?

(Of course I am especially interested in the question what happens when
committing
xml data.)

Thanks in advance for answering my questions.

Yours sincerely

Michael Bartholomäus

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri May 21 23:52:19 2004