On Tue, Feb 11, 2014 at 12:18:04AM +0400, Oleg Kolosov wrote: > Hello All. > > I am happy to inform you that I have achieved some success converting > Chicken build system from Makefiles to CMake. This initial work is a > proof of concept and still very ugly and incomplete, but certainly will > be improved. I've created a fork of Chicken on > https://github.com/bazurbat/chicken-scheme/tree/cmake-build, so anyone > interested can track progress and make requests there. > > If you have proposals or suggestions for the new build system, let's > discuss.
Hi Oleg, First off, thanks for putting in this amount of effort already. Now, about the CMake build: Let me start by saying I'm extremely sceptical about doing this because of CHICKEN's history with CMake. Furthermore, *if* we do this I put down the same hard requirement that Felix put down when I refactored the GNU Make build: The build *must* work and be tested on *all* supported platforms before we switch. If people object to having CMake, we must consider their reasons as well. The build scripts which I can see look very clean, though, so it may be worth the switch. I am trying to grok the CMakeLists.txt; could you explain where "chicken_wrap_sources" and "chicken_add_source" are coming from, for example? I quickly scanned the sources but couldn't find its definition. If this is some implicit "function" (what do they call these?) from CMake, where can I find out about it in the documentation? I don't see any code to take care of any of the non-"standard" build flags (ie, anything aside the least common UNIX denominator). Does CMake take care of these things? And what about the prefixing and suffixing of the binaries that are built (PROGRAM_PREFIX and PROGRAM_SUFFIX), which is used for cross-compilation? Does cross-compilation work at all? Cheers, Peter -- http://www.more-magic.net _______________________________________________ Chicken-users mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/chicken-users
