|
Boost-Build : |
From: Douglas Gregor (doug.gregor_at_[hidden])
Date: 2007-10-05 14:47:05
On Oct 4, 2007, at 3:28 PM, Rene Rivera wrote:
> Which was the point of having someone experiment with Cmake. So
> here's a
> few questions about that experimentation...
>
> * How much of the Boost.Build v2 functionality does it implement?
Most of the functionality we use in Boost, certainly. We have
features, variants, and can build multiple variants in build step. We
have regression testing support. The CMake-based system builds all of
the libraries, tools, and tests in Boost 1.34.1. It does some things
better than BBv2 (e.g., detecting zlib and bzip2 for use in
iostreams), some things that BBv2 can't do (my favorite: graphical
installers), but not everything has been ported over (e.g., no
BoostBook support).
> * How much code did that take to implement?
It's about 8,000 lines of CMake, 4,700 of which are comments.
> * How descriptive are the project and target descriptions? And how
> does
> it compare to BBv2?
Here's the Boost.Thread library's description:
boost_add_library(
boost_thread
barrier.cpp condition.cpp exceptions.cpp mutex.cpp once.cpp
recursive_mutex.cpp thread.cpp tss_hooks.cpp tss_dll.cpp tss_pe.cpp
tss.cpp xtime.cpp
SHARED_COMPILE_FLAGS "-DBOOST_THREAD_BUILD_DLL=1"
STATIC_COMPILE_FLAGS "-DBOOST_THREAD_BUILD_LIB=1"
NO_SINGLE_THREADED
)
The boost_add_library macro is described here:
http://svn.boost.org/trac/boost/wiki/CMakeAddLibrary
> * Does Cmake support correct interleaving of build action output when
> doing parallel builds?
Yes.
- Doug
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk