2009/3/29, Piotr Dobrogost wrote: > I'd love to hear Yang Tse opinion on this. He's working on a new nmake > makefile for the project and should join this discussion. By the way; > how is this work going, Yang?
Some general comments... I'm another one of the many who on this list have openly admitted not having used CMake at all. I might be wrong but I think that the actual work-force who have admitted knowing or having used CMake on this list is just a couple of persons, and none of them seem to have been around here for much time, if these are going to hang around long enough or not is something that time will tell. And there has already been at least one report of someone which having used it has been badly bitten. Something not technical that truly bothers me, I'm getting old and grumpy, is how hard CMake is been pushed all over the net on all relevant open source projects by the CMake advocates. For example, at some point in this thread it has been suggested that cmake should replace all already existing build systems for curl and libcurl, when the only cmake contribution so far seems to not be able to complete the task even for a single platform. We the not-yet-cmake-believers, or at least me, might be short-seeing but wouldn't mind to be shown the powers of the new-religion, oops build-system, so good as to justify proposing elimination of other build systems on this project. The day there is enough cmake infrastructure for curl/libcurl capable of building what is already possible to get built for a specific platform will be the day i'll express my opinion on the matter of removing the other build system support for that specific platform. Meanwhile my personal view is to certainly not remove it. On the other hand my motivation to improve MSVC makefiles right now is close to zero. I'm awaiting how all this cmake-evangelism turns out, letting the dust settle down and postponing further decisions. Who wants to work for nothing on something that is going to be discarded? My 2 cents -- -=[Yang]=-
