Add a patch for doxygen management in cmake
Use dynamic PLIST for documentation

Bump PKGREVISION

Makefile
remove useless comments
remove doxygen stuff not needed for this package

PLIST
add newline at the end

first release in robotpkg

Release v2.0.0

 * Add the display method for normalization purposes.
 * Synchronize.
 * Remove deprecated method commanline.
 * Add the command line 'remove', alias of 'rm'.
 * For normalization purposes, use the DYNAMIC_GRAPH_ENTITY_DECL macro everywhere.
 * Correct the prototype of the method getClassName.
 * [cmake] Synchronize
 * Correct documentation generation.
 * Synchronize.
 * Update sot-concept for romeo-sot v1.1.1
 * Remove deprecated python scripts.
 * Correct the path to the plugin directory of dynamic-graph.
 * Resolve ambiguous constructor
 * [travis] Add missing dependencies
 * [travis] Add missing dependencies
 * Update README.md
 * [travis] Add Travis and coveralls.io support
 * Fixed bug in last commit
 * Fixed bug in setGain() and added reference input type in MetaTaskDynPosture.gotoq()
 * Correct CMakeLists.txt: remove depr. file installation
 * Remove deprecated scripts.
 * Add TaskDynPassingPoint as a dynamic task
 * Minor changes to the most simple kine and dyn tests
 * Add MetaTask for dynamic angle and velocity limits
 * Modified function name and command to set second order kinematics in order to uniform it to Device::setSecondOrderIntegration
 * Added underscore to attribute name secondOrderKinematics in solver-kine.cpp
 * Solved unreferenced pointer if empty stack
 * Added comment for method setSize in stack-template.t.cpp
 * Resolved timing problem on solver-kine. Now working on second order
 * Added control on task type when second order is set
 * Added control on task type when second order kinematics is set
 * Added secondOrderKinematics option with drift Jdot*dq compensation in solver-kine and userfriendly changes/commentaries in solver-op-space.cpp
 * Added gain utilisation to task-dyn-limits.cpp and  added some useful comments
 * Added Relative MetaTaskDyn and modified CMakeLists
 * Added secondOrderKinematics option with drift Jdot*dq compensation in solver-kine and userfriendly changes/commentaries in solver-op-space.cpp
 * Update reference to soth (point towards github not laas).
 * Added gain utilisation to task-dyn-limits.cpp and  added some useful comments
 * Added Relative MetaTaskDyn and modified CMakeLists
 * Resets the posture of the robot between configs in sot-concept.py
 * Install the tutorial python scripts.
 * Complete ros-walkromeo.py script
 * Add Simple test for dynamic sot
 * Add rotation part for gotoNd and goto6d
 * Simple typo.
 * Minor modification for ros-dynromeo.py
 * Add missing viewer commands for robot-viewer in sot-concept.
 * Change the missing tasks.
 * Correct the right hand task for unittest/dyn_romeo.py
 * Correct import pathes and robot name in the scripts.
 * Sur-constrain the walk of romeo.
 * Purposefully add some gaps in the walk script.
 * Clean and comment the script ros-walkromeo.
 * Clean and comment the script ros-planche with romeo.
 * rename ros/p105.py
 * Clean and comment the script p105 with romeo.
 * Clean and comment the kine script with romeo.
 * Clean, comment and correct the dynamic script with romeo.
 * Default romeo robot has toes.
 * Update 'classic' python script for romeo flying.
 * Use the Macro solver in ros scripts.
 * Add the simulation for romeo 'planing' with ros.
 * Correct path in meta_task_6d python script.
 * First draft for all python scripts.
 * Add p105_demo for the romeo robot.
 * Add a test for the dynamic SoT with romeo.
 * Remove useless file.
 * Synchronize.
 * Working version for Romeo, thanks Francois.
 * Modif robot specificities (path for romeo).
 * Script correction and modif for romeo.
 * Refact version with the minimal elements.
 * Corrected the unittest kinewalk.
 * Corrected version of the walk: work with HRP2-14.
 * Merge remote branch 'origin/master'
 * Moved robotSpecific to robot_specific, and install the generated python files in the binary dir.
 * Minor modifs of the 2013_coursens scripts.
 * Modify the robot specificy file.
 * Mv robot specific to robot_specific.
 * Avoid having hard-coded paths in robotSpecific.py
 * Minor modif.
 * IVIGIT: script of the ENS class 2013.
 * Added the meta tasks in the cmake.
 * Removed a trace in solver kine.
 * Moved file.
 * Deprecated filed removed.
 * Moved to python/tools.
 * Moved the script to python/unittest. None of them should be working yet, and need a little bit of correction and selection.
 * Moved the meta tasks.
 * Moved to sot.core.
 * Moved some files in dedicated repository of python/
 * Update cmake.
 * Minor robot spec modifs.
 * Added the task weight in the python init file.
 * Moved the matrix util python into the sot core package.
 * Minor modifs in the meta tasks.
 * Minor changes in thread loop.
 * Added the ALWAYS timer in attime.
 * Correction of the solver kine header.
 * IVIGIT task-weight.
 * Minor changes.
 * Finally, the active-set init-guess solver for inv kine.
 * Revert "Revert "Added an operator into solver-kine.""
 * Revert "Added an operator into solver-kine."
 * Added an operator into solver-kine.
 * Corrected a bug in task ineq.
 * Synchronize.
 * Add missing link.
 * Correct python scripts.
 * Avoid having hard-coded paths in robotSpecific.py
 * Correct the operation 'down' for the stack-template.
 * Using the meta_pg.
 * Synchronize
 * Correct link.
 * Correct link dependencies.
 * Correct the linkage.
 * Add missing dependencies
 * Win32: correct API
 * Add missing const in the declaration of getClassName
 * Correct the handling of python modules
 * Correct the link with boost
 * Add missing dependency between libs
 * Win32: Instead of a symbolic link, realize a dirty copy of the headers
 * Correct the exportation of symbols
 * Correct the CMakelists
 * ivigit.
 * Added the compilation instruction for the new entity contact-selecter.
 * Account for the error dot in the task computation.
 * Added an accessor to the support signal.
 * Add the new contact-selecter entity in python.
 * Corrected a bug in debug mode.
 * ivigit.
 * Accounting for the modif in the sdes branch of sot-core.
 * Accounting for the modif in cmake/python.cmake.
 * Accounting for the new feature.sdes.
 * Modify the output for OpenHRP seqplay.
 * Working version used for the novela.
 * Varying position of the COM.
 * Slight changes in the timing.
 * Accounting for the modifs in sot-core/feature.
 * Synchronize
 * Adding option to specify the robot model in the replay.py script (-R option). Now works with HRP2 and Romeo models.
 * Added some other specificities to handle romeo.
 * Corrected a hard-coded 36 dim.
 * ivigit.
 * Added n+10.
 * Added hide and c
 * Added some cases for matrix2vector convertion + corrected a probable bug in rpy.
 * Slight modif in restore function.
 * Added a empty constructor.
 * ivigit.
 * ivigit.
 * Minor change in the task update.
 * IVIGIT.
 * Remove a undesired trace.
 * Working version with option, but problem of initialization if the waist is not at 0.
 * First working version.
 * Added the rpy-to-matrix.
 * IVIGIT.
 * Robot viewer short cut.
 * IVIGIT.
 * Icra version.
 * IVIGIT.
 * Version I am very proud of :D
 * Retried on the robot: failure: to much on the front.
 * Maximale version, checked on the robot.
 * Slight refactoring.
 * Less demanding version.
 * IVIGIT, copy from imit.py.
 * Final version, to be tested on the robot.
 * Simple correction to handle a vector in goto6d.
 * Uses fast forward to correct the history
 * IVIGIt.
 * IVIGIT: working with taskCom, bug with support.
 * IVIGIT.
 * Cleaned version.
 * Working version for OpenHRP, problem with the stab.
 * Added a trajectory tracking to smooth the slide.
 * Rewrite the vectorToTupple to enable row and col vectors.
 * Added a fast forward to execute all events until a given position.
 * Added zmp in history.
 * Cleaned up version.
 * Final version, working on OHRP.
 * Complete version, but cannot replicate on openhrp.
 * ivigit.
 * Back from the robot Xp: worked for 2cm, better maybe with 1.5. Touch with 1. Try to slower the motion.
 * Working version, ok for ohrp.
 * IVIGIT.
 * Non sigular version, falling with OpenHRP.
 * IVIGIT.
 * IVIGIT.
 * Mutliple changes: added a spatial scaling, modify the return type of the position functions, added a anglePosition function.
 * IVIGIT.
 * Working motion, ready to be used on the robot ... wait, still a bug on the hip.
 * Added the global "ping" function.
 * Added vectorToTuple
 * The best motion from now.
 * Working with foot down.
 * Working version, with posture, no head no arm and foot.
 * Added the posture task.
 * ivigit.
 * Added a op point modifier. Should be ported to sot-core.
 * Added a time-scaled parser.
 * Minor bug correction.
 * IVIGIT.
 * IVIGIT.
 * IVIGIT.
 * Corrected a bug in signal refresh.
 * IVIGIT.
 * Changes to accept several arguments per iteration.
 * Minor changes.
 * Added the computation of the forces in each contact.
 * Corrected the bug in the altitude of the foot.
 * Added a command to apprehend eventual changes in the Jacobian while estimating its derivative.
 * Unified the QR and pseudo inversion.
 * Syntax error in debug mode.
 * Working version with the new colpiv solver.
 * WS removal.
 * Passed the solver to template.
 * Transfert all the class from the unittest to the header..
 * First complete working version.
 * IVIGIT.
 * Moved dynred2 has the nominal reduced solver.
 * Current working version, comparing both regular solver and reduced solver.
 * Added some traces to check the task validity.
 * Bug in the normal forces corrected. Version seems to work with reduced acceleration and forces.
 * With solver reduced 2.
 * With the second formulation of the reduced solver.
 * IVIGIT.
 * IVIGIT.
 * First working version: control ok for 2 contacts + COM task + brek down.
 * First working version.
 * Correction of bug at the compilation.
 * IVIGIT.
 * IVIGIT.
 * Using the new entity helper macro.
 * Accounting for the changes in hrp210
 * Accounting for new gpool().

first release in robotpkg

Release 2.0.0

* [cmake] Synchronize
 * Synchronize.
 * The unittests are compiled again
 * Add missing CMakeLists.txt in exe
 * Synchronize
 * Correct installation path of the library.
 * [travis] Compile tests
 * Update README.md
 * [travis] Add Travis and coveralls.io support
 * Remove useless file.
 * Modifying the license file: LGPL.
 * Another cleaning of exe/simple.:
 * Minor modif in the unittest.
 * Simplified the simple comparison test.
 * Corrected the initial QR decomposition to spare some time.
 * Refactoring of IJRR tests.
 * Minor debug.
 * Added an option in the getOptimalActiveSet function.
 * Clean the test dir.
 * Correct errors in CMakeLists.txt, Rename a file.
 * One last remaining warning.
 * Correction of the warnings. Non more warnings (!!!)
 * Some refactoring.
 * Adding some this-> to fix compilation errors with Apple clang version 4.1 compiler.
 * Add the flag -fpermissive to ensure C++ 4.7 compatibility.
 * Add a forward declaration (C++ 4.7 compatibility).
 * IVIGIT: correcting a missing file for compilation.
 * Removed useless include in tmanu.
 * Added a correction of the COD (used only in the test) in CODClean. This file should replace the src/COD.hpp, which is yet kept for laziness reasons.
 * Added documentation in the tmanu cone test.
 * IHQP test update.
 * Added a new random generator.
 * Minor compilation correctyion.
 * IVIGIT unittest.
 * Added an initial active set into the Stage class.
 * Added an initial active set.
 * ivigit.
 * Clean and reorganize the cmake files.
 * Reduce the compilation time of the unit tests.
 * Use the routine in cmake/eigen.cmake to find eigen.
 * Synchronize.

Fix linking with assimp when not installed in /usr

This synchronizes with the ros-indigo version provided by robotpkg.

Changes since 1.9.35:
    remove explicit dependency on urdfdom urdfdom is provided via urdf and
    catkin_* CMake variables. The current setup was unbalanced anyways because
    along with urdfdom, urdfdom_headers should have been being depended on and
    used. This precipitated from urdfdom's rosdep key changing as it became a
    system dependency in Indigo.
    Add ability to delete all markers in Marker plugin
    fix hidden cursor bug On some systems loading a pixmap from an svg file can
    fail. On these machines an empty cursor results, meaning the cursor is
    invisible inside Rviz. This works around the problem by using an arrow
    cursor when the desired cursor pixmap canot be loaded.
    Install rviz to the global bin
    Added display for sensor_msgs/RelativeHumidity
    Contributors: Acorn Pooley, Adam Leeper, Chad Rockey, Dave Coleman, William
    Woodall, hersh, trainman419

1.11.2 (2014-05-13)

    Fix an issue with rendering laser scans: issue 762
    Fix an issue with using boost::signal instead of boost::signal2 with tf tf
    recently moved to boost::signal2, so the effort display needed to be
    updated too I made it so that it would conditionally use boost::signal2 if
    the tf version is greater than or equal to 1.11.3 I also fixed some
    compiler warnings in this code closes issue 700
    Contributors: Vincent Rabaud, William Woodall

1.11.1 (2014-05-01)

    fix fragment reference in point_cloud_box.material Closes issue 759
    upgrade ogre model meshs with the OgreMeshUpgrader from ogre 1.9
    Changed TF listener to use a dedicated thread.
    Speed up point cloud rendering by caching some computations and using
    proper loop iterations
    Fixed rendering of mesh resource type markers with respect to texture
    rendering and color tinting
    Fix segfault on exit for OSX
    Fix memory leak in BillboardLine destructor (material not being destroyed
    correctly)
    Fix disabling of groups (issue 709) This was broken with commit 5897285, which
    reverted the changes in commit c6dacb1, but rather than only removing the
    change concerning the read-only attribute, commented out the entire check,
    including the parent_->getDisableChildren() call (which existed prior to
    commit 5897285).
    Add missing libraries to rviz link step, fixes OS X build.
    fix failing sip bindings when path contains spaces
    EffortDisplay: Added a check to avoid segfaults when receiving a joint
    state without efforts
    Contributors: Dirk Thomas, Hans Gaiser, Jordan Brindza, Mike Purvis, Mirko,
    Siegfried-A. Gevatter Pujals, Timm Linder, Vincent Rabaud, William Woodall

1.11.0 (2014-03-04)

    fixing problems with urdfdom_headers 0.3.0
    Contributors: William Woodall

1.10.14 (2014-03-04)

    Fixed a bug in tutorials caused by uninitialized ros::Time here.
    Contributors: Dave Hershberger, William Woodall

1.10.13 (2014-02-26)

    Use assimp-dev as a build_depend and leave assimp as the run_depend
    Contributors: Scott K Logan, William Woodall

1.10.12 (2014-02-25)

    Shiboken is now disabled when a version which would segfault is detected
    (fix issue 728)
    Eigen is now found using the FindEigen.cmake from the cmake_modules
    package.
    Added support for rendering rviz in stereo. For more information see this
    commit:
    https://github.com/ros-visualization/rviz/commit/9cfaf78e2ae8d34e4481de19568b353964846842
    Added a "Queue Size" option for the Range display type.
    Added Ogre-1.10 compatibility This allows rviz to compile (and work)
    against Ogre 1.10 (currently the latest version of ogre). It also still
    works with earlier versions of Ogre (tested with Ogre 1.7.4 as installed
    via debs on Ubuntu 12.04).
    Now includes ogre without OGRE prefix This is necessary to find Ogre files
    in the right place with compatibility between Ogre < 1.9 and Ogre
    >= 1.9. This is also necessary when 2 versions of Ogre are installed on the
    build machine.
    RVIZ doesn't use __connection_header from incoming messages, but only uses
    ros::MessageEvent's
    Better feature detection for assimp version The unified headers were
    introduced in Assimp 2.0.1150, so checking for Assimp 3.0.0 is not quite
    the best solution. See
    https://github.com/assimp/assimp/commit/6fa251c2f2e7a142bb861227dce0c26362927fbc
    Contributors: Acorn Pooley, Benjamin Chrétien, Dave Hershberger, Kevin
    Watts, Scott K Logan, Siegfried-A. Gevatter Pujals, Tully Foote, William
    Woodall, hersh

1.10.11 (2014-01-26)

    Fixed in selection_manager which allows interactive markers to work with
    orthographic cameras views
    Add support for yamlcpp 0.5 with backwards compatibility with yamlcpp 0.3
    Fixed message type for Polygon display. The polygon display type actually
    subscribes to PolygonStamped.
    Contributors: Austin, Ken Tossell, Max Schwarz, William Woodall

1.10.10 (2013-12-22)

    Fixed a severe memory leak with markers and marker arrays: issue 704 and issue 695
    Contributors: David Gossow, Vincent Rabaud

1.10.6 (2013-09-03)

    Added a new method for adding displays, by topic as opposed to by type.
    Added new exception handling for loading mesh files which have no content.

1.10.5 (2013-08-28 03:50)

    Removed run_dep on the media_export package
    All previous history is not curated, see the commit history
    <https://github.com/ros-visualization/rviz/commits/hydro-devel>.

This package defines messages commonly used in mapping packages.

This synchronizes with the ros-indigo version provided by robotpkg.

Changes since 1.5.14:
. allows to have range_cutoff > range_max.
. Fix mistake in end_time calculation for scan transformation.

While here, tidy DESCR, depend.mk, sort dependencies and fix PLIST.

This synchronizes with the ros-indigo version provided by robotpkg.
No functional changes since 1.9.9

While here, tidy the package a bit (fix dependencies, fix a few minor glitches
in the depend.mk).

Release 0.4.0

Update romeo.urdf and romeo_small.urdf with 0.0.9 Aldebaran version, however urdf version have been modified for sot needs
Add Readme to explain why we are not yet synchronize with Aldebaran repo
[cmake] Synchronize

correct path
remove useless variables (they were leading to search for dynamic-graph-bridge as a system package)

robotpkg:
- remove patch-aa, patch-ab, patch-ac : now integrated in cmake module
- remove patch-ad : integrated in the code
- add patch-ae, patch-af :  dont know how to use rospack in robotpkg

Release 1.1.0

Upgrade in accordance with 0.0.9 Aldebaran romeo urdf version …
[cmake] Synchronize
[Travis] Add submodule.
Add a controller and a device for Romeo.

Previously, it uses a d extension to mark debug libraries. It sounds
like a good idea, but the python caller are not ready for that. So,
well, just don't do smart things for debug libraries.

We now use a packaged version with a "version number" and not anymore a
CVS data. Fix required ABI to take this fact in consideration.

Makefile : add roslisp
PLIST : Correct

robotpkg :
patch-aa, patch-ac : update without modif (need to test if always needed)
patch-ab : deleted, has been integrated in the release
patch-ad : dont know how to use rospack inside robotpkg, need to investigate

Release 2.0.0

* Merge pull request #25 from stack-of-tasks/hydro
* Reflect new messages with dynamic_graph_bridge_msgs catkinization.
* Correct signal destruction in rosSubscription.
* Correct signal destruction in rosPublish.
* Add missing dependency in dl.
* Include runPythonFile service
* Update CMakeLists.txt and package.xml
* Adapt run_command to the extraction of the messages from dynamic_graph_bridge.
* Remove msg/srv generation, correct link with dg_bridge_msgs
* Use msg/srv defined in dynamic_graph_bridge_msgs
* Correct rosPublish.rm method.
* Synchronize.
* Merge pull request #20 from aclodic/master
* Update CMakeLists.txt to install libros_bridge

First release in robotpkg
Dynamic_graph_bridge_msgs provides messages and types for dynamic_graph_bridge
package

patch-aa : add a patch to handle doxygen.cmake issue
BUMP PKGREVISION