Dustin: First, thanks to you, the rest of the Zmanda team and the community contributors like Jean-Louis Martinea for all of your hard work on Amanda.
I guess I was just surprised that the run-time library path is different for the executables/libraries generated in the build directory as compared to the versions that end up in the install directory. Why not just copy the compiled versions over to the install directory so that you know you're running what you checked? Presumably because you want the *BUILD DIRECTORY* in the library search path of the build directory executables, but you want the *INSTALL DIRECTORY* in the library search path of the installed executables. I'm fine with this, but the "make install" process should ensure that this is the ONLY difference between the library search path otherwise the "make check" step is not only pointless, it is misleading. As it stands, you can do a "make check" and have everthing work, but then a "make install" can completely change the included libraries such that you're not running what you checked. I always figured that the "make install" just copied the checked executables over to the install dir, but this is not the case. Unfortunately I no experience with automake/libtool so I can't propose a solution. Hopefully documenting this issue here will assist anyone else who runs into this in the future. Sean > >On Sun, Jul 25, 2010 at 3:10 PM, Sean Walmsley ><s...@fpp.nuclearsafetysolutions.com> wrote: >> Thanks for any insight you can provide (well, besides the obvious >> suggestion to not leave older versions of Amanda about!). > >I have never understood Solaris's approach to linking very well, but I >can say that Amanda uses a very straightforward application of >automake/libtool to do its linking, so it shouldn't act appreciably >different from any other application. The general recommendation is >to make sure that other versions of Amanda are out of the linker paths >when building a new version. If you can find something we can do in >the source to make this better, I'm all ears. > >Dustin ================================================================= Sean Walmsley sean at fpp . nuclearsafetysolutions dot com Nuclear Safety Solutions Ltd. 416-592-4608 (V) 416-592-5528 (F) 700 University Ave M/S H04 J19, Toronto, Ontario, M5G 1X6, CANADA
