RE: RE: Understanding Reflective/Cyclic Merges

>>svn merge ^/branches/sqlite-node-origins

>>Question: will merge tracking be smart enough to skip the revisions
>>that were merges from trunk to the sqlite-node-origins branch? Maybe
>>it just skips any merges to this branch that came from the copy-source
>>of the mergeinfoless-copies branch?

>No it is not smart here. Assuming it is the first merge, it can potentially repeat merges the trunk changes to >mergeinfoless-copies branch. issue 2897 does not take care of it. *Probably we need to do diff-summarize >equivalent on the merge source to see if at all the source is a result of some merge by looking for >'svn:mergeinfo' changes and drive our merge accordingly.*.

Sorry I was wrong, close look at the code suggests that it can handle such indirect reflective merges. Though there could be a confusion here.
Fix to 2897 avoids the commit-revision of merge from a target to source getting merged, this can skip some valid 'sqlite-node-origins only changes' which got merged to 'mergeinfoless-copies' along with some trunk merges(implicit via svn:mergeinfo prop merges) in one single commit.

With regards
Kamesh Jayachandran
Received on Mon Nov 19 21:12:45 2007