We've been extensively using Ant/ivy in our C++ environment. The main reason was actually ivy to have a (transitive) dependency manager, but also to have the same "build/install/run/test" interface to all our projects.
The idea is pretty simple: We have a company-wide repository of common libraries (C++source and pre-compiled) in different versions that can be used in other projects. When building a project, the correct (in term of version) dependee modules are downloaded from the repository, extracted and locally build (with the normal build environment e.g. MSVC, make). We also implemented an auto-environment feature that creates suitable environment variables pointing to the include / import library paths of the locally built common modules. But be warned: We had to extensively use Python scripts to enhance Ant/Ivy to make it all work and it cost many man-months. In the end, we use ant as a standardized interface to our build management system, use ivy for dependency management and leave the rest to python scripts. But once set up, it works *really* good. -----Ursprüngliche Nachricht----- Von: Marcel Overdijk [mailto:marceloverd...@gmail.com] Gesendet: Donnerstag, 6. Dezember 2012 17:15 An: ivy-user@ant.apache.org Betreff: Ivy in C/C++ environment I wonder if somebody has some pointers for using Ivy in a C/C++ environment. a) how is dependency management done (e.g. using custom resolver?) b) how is building done (based on on de Ivy dependencies) I'm not looking for a complete solution, just wat to start a discussion about possibilities or perhaps best practices from people already having this set up. Unfortunately I can't find and information in the docs. Ivy is especially interesting as it is nog tight to Java dependency management. -- View this message in context: http://old.nabble.com/Ivy-in-C-C%2B%2B-environment-tp34767076p34767076.html Sent from the ivy-user mailing list archive at Nabble.com.
