Paul A Bristow a écrit :
Recently we had a flurry of discussion of why Boost is not more widely used.
I have used some bits of Boost happily for years, but recently I felt I had to use it 'properly' so concluded it was time I used the bjam build system - something I had felt looked a bit clever.
First of all : a disclaimer: I just skimmed through the whole discussion, so the points in my posts might be adressed somewhere else. I basically agree with this opinion, since it has also been a remark people make to me about boost (I'm the main boost proponent in my office). So here is my proposal: * First, easy to do, include more examples in the documentation. I've verified several times that when they don't want to read and understand a whole document, people try to skip to the examples. Just something like : To build all with VC++7.1 : (after calling the correct VCVARS.bat) bjam "-sTOOLS=vc-7_1" install To build only the serialization module with VC++6.0 : bjam "-sTOOLS=msvc" --with-serialization install To build only the multithread version vith BC++ bjam "-sTOOLS=borland" "-sBUILD=<threading>multi" install // I'm really not sure for this one * Second, provide a compiled bjam for major platforms directly in the distribution (ok, the distribution will be bigger. Is this a problem ?) * Third, provide some script to add boost to the compiler path (I do not know how to do this...) * Finally, installing boost takes longer than installing VC++... Is building from the source such an important point ? Would'nt it be possible to provide pre-built libraries for the 3 or 4 main compilers on windows with each release ? The whole package, zipped, is 80mb on my VC8. -- Loïc