Different RA layers should use the same error code to signal the "path not
found" error condition.
Code like this appears in several places :
> err = editor->delete_entry(path, item->revision, parent_baton, pool);
> if (err)
> if (err->apr_err == SVN_ERR_FS_NOT_FOUND
> || err->apr_err == SVN_ERR_RA_DAV_PATH_NOT_FOUND)
These two errors both mean "the item was not present in the repository".
SVN_ERR_FS_NOT_FOUND is sent by the file or svn RA layers, and
SVN_ERR_RA_DAV_PATH_NOT_FOUND is sent by the http/https RA layers, with the
Similarly, there the problem is with
> svn_ra_get_locations(svn_ra_session_t *session,
> apr_hash_t **locations,
> const char *path,
> svn_revnum_t peg_revision,
> apr_array_header_t *location_revisions,
> apr_pool_t *pool)
If PATH doesn't exist at PEG_REVISION:
RA_LOCAL/RA_SVN: Returns SVN_ERR_FS_NOT_FOUND
RA_NEON: Returns two SVN_ERR_RA_DAV_PATH_NOT_FOUND errors
RA_SERF: Returns SVN_NO_ERROR, sets *LOCATIONS to an empty hash. 
See email thread . In that thread Karl suggests, "The fix(es) should happen
as close to the original (server-side) source of the error as possible, IMHO."
Barring that, the fix should be in the
particular RA implementation.
 Email thread: <http://svn.haxx.se/dev/archive-2008-03/0135.shtml>.
 Email thread: <http://svn.haxx.se/dev/archive-2007-12/0457.shtml>.
 This is reckoned to be a bug: see email thread