Stefan Sperling wrote on Mon, Jul 01, 2013 at 13:52:08 +0200: > On Mon, Jul 01, 2013 at 02:33:06PM +0300, Daniel Shahaf wrote: > > I don't remember whether I pointed this out when Stefan originally wrote > > that code: > > > > [[[ > > * subversion/tests/libsvn_fs/fs-test.c > > (filename_trailing_newline): Switch from a blacklist approach to > > to a whitelist approach, for defining backends that don't implement > > the API correctly. > > ]]] > > When a new backend comes along that doesn't handle \n due to a bug > in the design/implementation, this test is supposed to FAIL, not PASS! >
The current HEAD code will expect a new backend to return an error on \n. The code with my patch will expect a new backend to return SVN_NO_ERROR. It sounds like applying the patch will cause the code to do what you say you want it to do. What am I missing? > So I would strongly prefer keeping a whitelist.
