> Well, the current "python setup.py install" and the packaged RPMs that > are created using cx_Freeze put the binary and the library.zip and all > of the associated extension files in <prefix>/lib/<name>-<version> > where <name> and <version> are the values defined in the metadata in > setup.py and <prefix> is /usr or /usr/local or wherever you specify on > the command line. Then a symbolic link is created in <prefix>/bin > pointing to the executable(s) found in the above named directory. That > ensures that each package is independent but enables you to run it in > the normal fashion without changing the PATH environment variable or > anything like that. I believe this resolves all of your questions but > I'd appreciate your comments. In any case it seems to work well for > me. :-)
Thank you Anthony. I better understand now how cx_freeze expects to have its dependencies available. There is one more issue I want to bring to your attention. I suspect cx_freeze does not collect all .so dependencies. I followed the approach you recommended to produce a build directory with the binaries and the .so files, but running the binaries from that directory in a different machine does not work. For example, it reports that (for my particular application) libffi.so.5 is missing. This is a library that was needed to compile one of the python libraries needed by my application. What is the best way to handle this ? Should I manually copy in the distributed directory such libraries, or is this something cx_freeze should be handling ? ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ cx-freeze-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/cx-freeze-users
