This seems like the only way to get VW to find its own copy of gtest over one that is installed by the system. Normally this wouldn't be a problem but Macports and ROS want to install gtest 1.5 which causes cryptic errors against VW who wants to use gtest 1.6.
This allows the user to specify which protoc they'll be using as potentially they could be building against something older than what their system provides.
they're tied to library version, and we don't check the library version.
It didn't export the abs_srcdir and abs_builddir vars, and they weren't actually completely absolute anyway. Worked around it.
protoc handles paths as strings so it can't compare absolute vs. relative paths.
freebsd6 doesn't seem to set a default one... which is pretty lame.
use ./configure --disable-silent-rules or make V=1 to make them verbose again. This is only available on automake 1.11.
…stem that uses AMQP and Google Protocol Buffers for the RPC fabric. Began to work on a basic mosaic_worker program that can run in the background to process mosaic tiles as needed. Wrapped rabbitmq-c AMQP library in a C++ wrapper. This is all beginning to be a lot of code in this module, and at some point we may want to move it over into its own project.
…ost datetime detection
And clean up qt detection. Again.
…g. dunno. moving the function def to where it's used to compensate.
…move autotroll's hard-coded 2.60 dep to unbreak the build on some of the irg machines
…the first Vision Workbench graphical user interface tool. It allows you to read in any VW supported image and browse it in a QT/OpenGL window. Hovering your mouse over the image shows the real pixel values, and images with different channel types and pixel types are fairly well supported. High dynamic range images and DEMs can be displayed via image normalization. The UI for vwv mostly consists of special key bindings at the moment. I'll add documentation for these soon. Requires: Qt (detected in the build system by calling qmake), OpenGL 2.0 support, a relatively modern graphics card.\n
changes make it so the macros don't call ac_msg_error when they fail.
… new thirdparty/m4 directory
…e return (mostly for TerminalProgressCallback). Fixed a small bug in the cxxtest::TestSuite that appears to be a conflict with boost. Not sure why it hasn't manifested itself before...