On Thu, May 28, 2009 at 01:44:36PM -0400, Jeff King wrote: > On Wed, May 27, 2009 at 11:11:17PM -0700, David Aguilar wrote: > > > +int git_mkstemps(char *path, size_t n, const char *template, int suffix_len); > > FWIW, I find this name not very descriptive. From the name, I would > expect it to do the exact same thing as mkstemps, but be our own > personal implementation. But it is actually a wrapper that behaves > somewhat differently. So I wonder if "mkstemps_tmpdir" or something > would be a better name. It does exactly what git_mkstemp() does, plus the extra suffix_len parameter. If we rename this function we have to rename both. > Is there a reason _not_ to always just use the pretty filename? It looks > like you turn it on for external diff, but off for textconv. I don't > think there is a reason not to use it for textconv. I was not aware of the other code paths and only wanted to affect the one that I knew about. I agree that making that the default behavior would be great, meaning we could drop the pretty_filename flag altogether. If you and others agree that the user-friendly names are a good thing to have by default then I can rework patch 2/2. > However, I suspect that all callers should use pretty filenames, and > then this bit would just go away. I fully agree. -- David -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html