Konstantin Shvachko wrote:
HADOOP-3628 was just committed to trunk I believe by mistake. Looks like it was intended for the HADOOP-3628 branch. Please hold on with commits until this is reverted. Thanks --Konstantin
Konstantin -sorry for causing trouble. I'd seen that the text file had seemed to have gone in the right place so I pushed out the rest, but clearly SmartSVN had a different idea. I shall let any of your colleagues ask questions at apachecon about whether or not I understand how to use svn switch properly as part of the public humiliation for this.
Incidentally, because I dont trust svn switch that much, for our work releases we have forked vmware images for cutting releases of different branches -that not only guarantees that the wrong branch is never touched, it stops the .ivy repository accidentally containing artifacts from other builds, and ensures that the releases are from unmodified files. But I don't see that working for the main development cycle. I will rm and then check out the branch clean, to ensure my machines are fully bonded to the branch.
Again, many apologies to everyone. steve
