robotlocomotion / libbot Goto Github PK

For previous discussion, see: db3e32c#commitcomment-12640480

I just had a closer look at the offending code in pods.cmake, it says:

get_target_property(IS_TARGET ${__depend_target_name} LOCATION)
if (NOT IS_TARGET STREQUAL "IS_TARGET-NOTFOUND")
  ...
endif()

In CMake you can test whether a name is a target with:

if(TARGET ${__depend_target_name})

Code like this already exists in pods.cmake provided by RobotLocomotion/cmake.git

Maybe a low priority issue, but with this fix it could avoid setting the policy to OLD. I think it's best not to set policy to OLD because it can cover other bugs.

This repository is not forked off of what appears to be the (current) canonical upstream, https://github.com/libbot2/libbot, and the history is out of sync.

I propose that someone with sufficient permissions:

• Push a copy of the current master to some new branch name (to prevent garbage collection from pruning the old history until forks can be updates).
• Reset master to mwoehlke-kitware/libbot@e7ff4150c275, which has the same content as current master (@a889c5f10f31), but meshes with the upstream history.
• Ask github to attach this repo as a fork of https://github.com/libbot2/libbot so that PR's can be made between us and upstream.

The lcmgl_bot_renderer.h header isn't installed when building bot_lcmgl_renderer, which causes the bot-lcmgl build to fail later on. Has anyone else had this problem?

It's an easy fix, and I can open a pull request for it, but I'm curious if maybe it's just something wrong with my system (since I would expect this to cause the build to fail for everyone, which doesn't seem to be the case).

I think that most users of this repo are also using bot_core_lcmtypes. This can result in a bug if the lcmtypes from this project are installed over the those from bot_core_lcmtypes.

As a quick fix to avoid this bug, I propose that we add a CMake flag, SKIP_BOTCORE_LCMTYPES, which causes this repo to use the installed types from bot_core_lcmtypes and entirely skip over processing the core types in this project, libbot.

This bug has currently manifested itself in drake's install tree on Mac OSX: bot_core_lcmtypes are installed into the site-packages, while libbot's bot2-core/lcmtypes are installed to dist-packages. Ouch!

bot_core is missing a couple of the standard lcmtypes we used in the DRC: generic point cloud, IMU, joint state and presumably robot_state_t. Other to consider would be robot_urdf_t and force_torque_t

@patmarion and @RussTedrake would you find it useful to define standard types like this for future applications?

For pronto I keep having to redefine DRC types in pronto. Ideally there would be a submodule just with lcmtypes that I could use in my distro repo.

(Also, moving to use this repo in the mitdrc/drc would be good)

BTW, ROS has a set of clear and fully static standard sensor messages:
http://wiki.ros.org/sensor_msgs

...because header file gdk/gdkx.h no longer exists. I think this will be a problem for anyone doing a clean install with a new Mac.

https://support.apple.com/en-us/HT201341

LCM has been moved. You need to replace the existing link (in your Readme) with http://lcm-proj.github.io/

I manually move each commit from libbot-drc to mitdrc/libbot

• I figured out that libbot was forked at b8cfe1a
• this was r7667 of the DRC svn repository.
  Below are the commits that were made. I skipped a few that weren't useful.

This is on this branch:
https://github.com/mitdrc/libbot/tree/mfallon-merge-libbot-drc-changes-alpah
Which is based off of b8cfe1a.

r8334 | antone | 2015-03-10 15:35:41 +0000 (Tue, 10 Mar 2015) | 1 line

reconcile running state of sheriff with that of deputy at startup; this fixes behavior that erroneously starts stopped processes on sheriff restart

r8332 | russt | 2014-12-23 20:16:03 +0000 (Tue, 23 Dec 2014) | 1 line

obj support update from Russ

r8331 | antone | 2014-11-24 02:05:39 +0000 (Mon, 24 Nov 2014) | 1 line

added bot_core::images_t

r8329 | mfallon | 2014-08-28 01:45:21 +0100 (Thu, 28 Aug 2014) | 1 line

minor mod to param server republishing

r8328 | antone | 2014-08-28 00:28:37 +0100 (Thu, 28 Aug 2014) | 1 line

added lcm message based initialization for param server

r8325 | russt | 2014-03-25 16:08:52 +0000 (Tue, 25 Mar 2014) | 1 line

merge in change from github: enabling depth tests in lcmgl renderer

r8324 | russt | 2014-03-25 16:08:08 +0000 (Tue, 25 Mar 2014) | 1 line

merge change from github / libbot: make start/stop recording methods public

r8323 | antone | 2014-03-24 16:39:36 +0000 (Mon, 24 Mar 2014) | 1 line

fixed memory leak in which freeing elements from middle of config block would leave next elements unreferenced

r8322 | antone | 2014-03-21 21:11:32 +0000 (Fri, 21 Mar 2014) | 1 line

fixed bug in string array parsing

r8149 | antone | 2013-12-10 19:42:58 +0000 (Tue, 10 Dec 2013) | 1 line

fixed bug in param server value update when not an array

SKIPPED r7971 | mfallon | 2013-12-06 00:37:47 +0000 (Fri, 06 Dec 2013) | 1 line

adding printf request mesg

SKIPPED r7969 | mfallon | 2013-12-06 00:17:57 +0000 (Fri, 06 Dec 2013) | 1 line

adding fetch printf capability

SKIPPED r7946 | mfallon | 2013-12-05 20:38:51 +0000 (Thu, 05 Dec 2013) | 1 line

removing restart button

r7762 | russt | 2013-11-27 19:30:08 +0000 (Wed, 27 Nov 2013) | 1 line

merged non-Apple fix from main libbot

r7688 | antone | 2013-11-26 14:30:11 +0000 (Tue, 26 Nov 2013) | 1 line

fixed bugs in param client and server that prevented proper external 'set_x' methods from working properly; added new message field that distinguishes between different set types

r7687 | antone | 2013-11-26 14:28:01 +0000 (Tue, 26 Nov 2013) | 1 line

reverted russ's cmake changes from libbot head that were accidentally included in the clone

r7667 | antone | 2013-11-25 21:51:33 +0000 (Mon, 25 Nov 2013) | 1 line

made copy of libbot for local modifications