On Thu, Jun 28, 2012 at 12:57 PM, Greg Stein <[email protected]> wrote: >... >> + svn_wc__db_t *db, >> + const char *wri_abspath, >> + const char *marker_abspath, >> + apr_hash_t *original_props, >> + apr_hash_t *mine_props, >> + apr_hash_t *their_props, >> + apr_hash_t *conflicted_prop_names, >> + apr_pool_t *result_pool, >> + apr_pool_t *scratch_pool); > > As I mentioned in conflict-storage-2.0, there are four sets of > properties. Do we just never generating the incoming-base properties?
Answering myself: yes, we already have the four property values. This API and the storage should account for that. (and should CONFLICTED_PROP_NAMES be an array of char* rather than a hash?) Cheers, -g
