Hey all, I've constructed a dump file that we can load into repos/asf/steve/trunk/ which preserves all the history. Before I write up an Infra ticket to load it, please take a look: people.apache.org:~gstein/voter.dump.gz
This dumpfile was constructed with the following steps: 1. svnrdump for private/foundation/voter/ (about 34k revisions from first/last voter changes) 2. filter out all revisions that did not touch voter (leaving 221 revisions) 3. fix up all paths to refer to steve/trunk/ 4. filter a stray revision that tweaked svn:ignore on private/foundation/ This leaves us with 220 revisions for the voter/ history. My plan is to ask Infra to do a simple load of these revisions. They shouldn't even need to schedule (read-only) downtime, as the load should only take a few tens of seconds. Because the repository will be in read/write mode during loading, it *is* possible that other commits will occur during the load process. However, we have no need for "our" revisions to be in a contiguous range, so that shouldn't pose a problem. Further thoughts welcome! Cheers, -g
