On Sun, Oct 9, 2011 at 3:38 PM, Bill Morris wrote: > I am not sure this is the ideal solution. > > First a patch should probably be pushed upstream if possible. I emailed > Philip Lamb at ARToolworks to see if they are accepting patches for the > GPL version. > This is an excellent point. Let me know if you hear back from Philip, and whether the patch can be pushed to the main ARToolKit repository. > Second, I would prefer not modifying the tarball directly as it is > supposed to be a snapshot of the ARToolkit SVN. I think the preferred > route is to use patch and replace the patch_v4l kludge. > > It is probably clearer what something like the patch system for the > canonical scan matcher is doing. > ccny-ros-pkg/scan_tools/csm/Makefile > ---- > patch: > ifneq (,$(wildcard $(BUILD_DIR)/.patched)) >        @echo Patching CSM >        cd $(SOURCE_DIR) && patch -p0 < ../../patch_a >        cd $(SOURCE_DIR) && patch -p0 < ../../patch_b >        cd $(SOURCE_DIR) && patch -p0 < ../../patch_c >        touch $(BUILD_DIR)/.patched > else >        @echo CSM previously patched, use 'make wipe' to completely > remove build directory. > endif > > wipe: clean >        rm -rf $(BUILD_DIR) > ---- > This looks much better. I can rewrite this using 2 patch files - one for the change the patch_v4l currently provides, and one for the the v4l stuff. Is this ok with the guys at CCNY? > Is there a REP for third party packages? > > On Sat, 2011-10-08 at 17:59 -0500, Piyush wrote: >> This is correct - the patch was only for Natty. For some weird reason >> I assumed that the debs could be built externally without pushing this >> change to the repository. It was a bad idea in the first place. >> >> I will see if I can work on this later tonight - the patch is still >> relatively easy based on what Bill suggests, and will require >> modifying ARToolKit's custom configuration script to provide the >> LIBV4L1 flag as necessary. >> >> Piyush > > > -- > Bill Morris > I Heart Engineering > http://www.iheartengineering.com > <3 > > _______________________________________________ > ros-users mailing list > ros-users@code.ros.org > https://code.ros.org/mailman/listinfo/ros-users > Piyush