<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;"><style type="text/css">#toc, .toc, .mw-warning { border: 1px solid rgb(170, 170, 170); background-color: rgb(249, 249, 249); padding: 5px; font-size: 95%; }#toc h2, .toc h2 { display: inline; border: medium none; padding: 0pt; font-size: 100%; font-weight: bold; }#toc #toctitle, .toc #toctitle, #toc .toctitle, .toc .toctitle { text-align: center; }#toc ul, .toc ul { list-style-type: none; list-style-image: none; margin-left: 0pt; padding-left: 0pt; text-align: left; }#toc ul ul, .toc ul ul { margin: 0pt 0pt 0pt 2em; }#toc .toctoggle, .toc .toctoggle { font-size: 94%; }body { font-family: 'Times New Roman'; color: rgb(0, 0, 0); widows: 2; font-style: normal; text-indent: 0in; font-variant: normal; font-weight: normal; font-size: 12pt; text-decoration: none; text-align: left; }table { }td { border-collapse: collapse; text-align: left; vertical-align: top; }p, h1, h2, h3, li { color: rgb(0, 0, 0); font-family: 'Times New Roman'; font-size: 12pt; text-align: left; }.plain_text { font-family: 'Courier New'; }</style> <div> <p></p> <p class="plain_text"></p> <p class="plain_text"></p> <p class="plain_text">I found out when goggling that boost is trying to create a python-based Boost.Build library. I wondering if any one thought about taking this one step further with swig and not limiting them self to just python? I think this might be a little easer to do but i don't no a lot about swig so it hard to say.</p> <p class="plain_text"></p> <p class="plain_text">The reason Why I think this would be good is for several reason ....</p> <p class="plain_text"></p> <p class="plain_text"><br></p><p class="plain_text">1.)It would allow for control by IDE not written in c++ or python.(java NetBeans for example could use it natively in there own language to control Boost-Build form there ide)</p> <p class="plain_text"></p> <p class="plain_text">2.)It would allow people to write praser(in other languages other that c++ or python) for more then just jam files(they could make a prase to read cmakeList.txt for example.)</p> <p class="plain_text"></p> <p class="plain_text">3.) It would I believe lead to my idea of a perfect build system.(but i could be wrong I sure it can be debated)</p> <p class="plain_text"></p> <p class="plain_text">IMHO I have not been very happy with the build system out there they all see half baked they have the following problems.</p> <p class="plain_text"></p> <p class="plain_text">1.)they demand to use there language when parsing files for example cmake can only read cmakeList.txt the can’t read jamfiles,MakeFiles or the format scons uses. I would like to see a build system actually smart enough to read (The developer choose build file format.) instead of demanding you to user there build format(you might say there language) </p> <p class="plain_text"></p> <p class="plain_text">what would be relay nice is have a build system smart enough to read a CmakeList.txt.makefile or what ever file and make a Jam file and vise versa.</p> <p class="plain_text"></p> <p class="plain_text"><br></p><p class="plain_text">2.) Most build-system I find hard to read Cmake,Scon,autoMake are all bad at this(please remember this is all my opinion) I think makefiles are better.<br></p> <p class="plain_text"></p> <p class="plain_text"><br></p><p class="plain_text">3.) They have a steep learning curve. I spent alot of time trying to leaner cMake and automake and got no ware <br></p> <p><br></p><p>I alway thought It would be good to have a build system that can handle build files ment for other build system. I think this could be implement that by make boost-build look for plugin(for a praser) or scripts (for a parser) in a directory(s) . For example we could have 1 plugin for a jamfile praser and another for cmake praser. </p> <p>I case you are wounding where i Got my Info <br>I found some doc at<br> <br>http://svn.boost.org/svn/boost/branches/BOOST_BUILD_PYTHON/boost/tools/build/v2/python/doc/Overview.txt <br><br>From the paragraph I read quote<br> "<br> Goal<br> </p> <p class="plain_text">----<br>Create a Python-based Boost.Build library.<br>All the functionality available to BBv2 users shall be provided in the form of a library with a well-documented API. Parsers for higher-level languages, such as bjam, may be developed and use that API.<br>An abstract interface between the library and the build engine shall be defined and used, thus avoiding mixing library code with direct calls to a specific build engine and simplifying integration with different build engines.<br></p> <p>" <br>***********************************************************</p> <p>Please forgive me I know it sound like (I’m complaining and I deserver nothing) </p> <p>this rely not my attempt. I just want to being my idea out in the open and tell you my hope for boost-build) and I don’t know how to word it any other way.</p> <p>***************************************************************</p> </div></td></tr></table>