On Wed, Jun 15, 2022 at 09:30:52AM -0400, Derrick Stolee wrote: > On 6/14/22 8:37 PM, Taylor Blau wrote: > > `write_mtimes_file()` takes an optional parameter `mtimes_name`, which > > specifies where to write the mtimes file. If it is NULL, a location is > > generated with `odb_mkstemp()`. > > > > This imitates the pattern in `write_idx_file()`, and `write_rev_file()`, > > both of which have callers from the `index-pack` builtin which specify > > an exact location instead of generating one. > > I have a nearly-identical patch [1], but I'm happy to take Taylor's > instead. I'll plan on dropping that patch from my v2. ;-). Great minds think alike? Either that, or one of us (me) stopped reading the rest of this series before sending their patch. > [1] https://lore.kernel.org/git/b67e110bf60e820874de94c64ee8c32d69413877.1655242070.git.gitgitgadget@xxxxxxxxx/ Either version is fine with me. It's probably easier for the maintainer to just drop my patch and take this series in one go, so I'd err on the side of your version in [1]. Thanks, Taylor
