Re: Tree conflict resolution considered harmful

2018-09-03 Thread Daniel Shahaf
[moved from users@; https://svn.haxx.se/users/archive-2018-08/0052.shtml] In a nutshell, it concerns me that passing --non-interactive changes the default value of --accept. I think --non-interactive should use the same algorithm and have the same results as interactive merges. Sorry for being s

Re: BUG: Delete svn:special property on symlink; hilarity ensues.

2018-09-03 Thread Julian Foad
Branko Čibej wrote: > The only correct way to change a node's type is to replace the node itself. The crucial missing piece of the definition is the mapping between the client-side notions of node kinds and node lifetimes and their server-side counterparts. (Where client side treats specials as d