Hi everybody! Usually developers read mailing lists to find out useful data to make a good choose. I hope this mail could be helpful for them.
People said that was impossible writing a decent cache for Subversion repositories using a rational database and get results with enough speed. Hence, tools which need this feature has develop its own mechanism to deal with millions of data. Fortunately, I gave Derby a chance and I used it to build a cache for the public hugest repositories around the World (KDE, Apache,...) and results were absolutely awesome. KDE: with more than 1 million of revisions, near to 2.5 millions of files and directory entries and more that 12 millions of actions on those files and directories... all them on a Derby embedded database. Awesome speed and absolutely stable. Fast, reliable and easy! Thanks a lot for all people who have made Derby. Pablo www.svnflash.com
