When I modify a property (say, svn:ignore) on the top most directory I checked out (most of the time it's trunk, anyway), using TortoiseSVN the client(s) recursively go through the complete project.
Big tree, much data, takes a long time.
How can I avoid this? There may be changes somewhere else but I know that I'm just not interested in them and uncheck them anyway.
I'm aware of the sparse checkout feature but I don't see how I can apply this here as I don't have a sparse checkout. Do I need to create a second, sparse checkout, apply the changes/commit and then update my other copy?