Ruediger Pluem wrote: > > On 08/03/2009 11:07 AM, Bojan Smojver wrote: >> On Mon, 2009-08-03 at 10:58 +0200, Rüdiger Plüm wrote: >>> Bill, does this work on Windows as well? >> Doesn't this patch bring us back where we were (roughly)? >> > > Not really. Bills patch ensured that we consistently pickup > the headers from the mysql subdir or from one level above. > But you are correct my patch reverses the ordering Bill introduced. > Hence my question to him :-).
Yes, it does essentially revert, but throws away several legibility issues and bad practices (#define HAVE_THAT_H in a .c source? seriously ;-?) I can't follow the old code, but it is entirely possible to have picked up global/sys from one system install and mysql.h from another, which can't possibly be a good thing. Mine and Rued I've applied this, since apparently it's not possible to both compile against my_global/my_sys on unix and windows. On windows, this is broken since my_sys includes far too many #LL constants and such that seem solely destined for gcc (Mingw perhaps). What is interesting, it appears my_sys is not actually used, and the my_global header solves the my_init binding problem. It would make sense to enable only that one header, on Windows; issues solved.
