On 07/29/2011 03:13 AM, Ryan Schmidt wrote:
"/opt/swig" would possibly be a good choice and consistent with FHS.
http://www.pathname.com/fhs/pub/fhs-2.3.html#OPTADDONAPPLICATIONSOFTWAREPACKAGES
Yes, /opt/swig sounds like a sensible recommendation.
The problem only occured recently as we discovered that Xcode 4.1
dropped SWIG entirely. This forced Clemens to find a better solution.
For now it's that we have to commit the machista_wrap.c to our
repository after making changes to the machista API.
This would be the first external library dependency for MacPorts which
we overcome by comitting the wrapper. But it is similar to the fact that
we are committing our autoconf generated scripts already.
Hopefully we can weed out the warnings from the code produced by SWIG
2.x so that there is no need to maintain a port for SWIG 1.x over the
long run.
Writing the bindings by hand is almost impossible as many structs and
functions are involved for which SWIG generates C wrappers. And we gain
a lot of speed from the implementation in C. Therefore we don't want to
abandon SWIG for this task.
Rainer
_______________________________________________
macports-dev mailing list
[email protected]
http://lists.macosforge.org/mailman/listinfo.cgi/macports-dev
