Why SVK instead of...
Performance (with some tips):
- "svk log" is much faster over big repo and history ('svn log' is equal to 'svk log -x')
- "svk status" is much faster than svn over large directory.
- 1.3G repository, 35k revisions.
- svn working copy of trunk will take 2.2G
- svk mirror will take 1.3G
- svk checkout is 750M
- So, 200M saved plus the whole history of thirty-five thousand revisions