>> I succeeded in: >> - correctly build all apps without any rpath (and without patch) >> - correctly manage dh_shlibdeps step > > I guess you are refering to the currently commented parts done in SVN > commit r14132, right? Without having tested this looks promising.
Yes. >> - create a small wrapper (that manages the LD_LIBRARY_PATH) >> - launch freemedforms from the wrapper >> >> Should I include the DEB_HOST_MULTIARCH in libs/plugins path? > > IMHO we should try to support multiarch if there are no strong reasons > not to do this. However, before we ask for new trouble we should > probably resolve the rpath issue which is claimed by lintian (missing > multiarch is not (yet) considered by lintian). Hummm, I think I should work on this already before releasing the 0.9.0. It will be easier for me as I'm currently documented. I already managed the build process but I have some trouble with the freemedforms internal code that I can easily fix. >> * there are no arch dependent code in the project > > Did you missed an 'in' in front od dependent? IMHO C++ code is always > arch dependent. Arf yes. I meant: I've no arch conditional code in the freemedforms project code ;) Eric
PGP.sig
Description: Ceci est une signature électronique PGP
