Add SRDFWriter to this repo

[velveteenrobot]

Fixes issue: moveit/moveit#66 from MoveIt.

[velveteenrobot]

Merge this before moveit/moveit#142.

[davetcoleman]

should be _SRDFDOM_SRDF_WRITER_

[davetcoleman]

This still needs to be fixed, should not have been merged yet

[rhaschke]

I fixed this myself before merging: 8bfdb1d

[davetcoleman]

@rhaschke this is committing directly to the main branch, which circumvents the review process necessary for stable code. if you don't want to wait for @velveteenrobot to make the change I think you should merge this then open a secondary PR rather than commit directly, like i did yesterday here

[rhaschke]

Merged. @130s Could you prepare a release of srdfdom to unblock moveit/moveit#142?
Thanks to @velveteenrobot.

[rhaschke]

Hm. Now that I merged, Travis fails in MoveIt. I will add a travis config to srdfdom too.

[rhaschke]

Fixed srdfdom. Updated .travis.yml to MoveIt's one. Travis is back to normal: https://travis-ci.org/rhaschke/srdfdom
@davetcoleman Do you have an idea, why https://travis-ci.org/ros-planning/srdfdom didn't build for ages?

[130s]

Could you prepare a release of srdfdom to unblock moveit/moveit#142?

Let me know when it's ready.

Do you have an idea, why https://travis-ci.org/ros-planning/srdfdom didn't build for ages?

Has travis been on for this repo (I see now it's on at http://travis-ci.org/profile/ros-planning though).

[rhaschke]

Do you have an idea, why https://travis-ci.org/ros-planning/srdfdom didn't build for ages?

Has travis been on for this repo (I see now it's on at http://travis-ci.org/profile/ros-planning though).

Yes, travis was configured to be on, but it nevertheless didn't triggered builds on pushes or PRs.
I just noticed that the detailed settings were configured to be off! Fixed.

[rhaschke]

OK, Travis triggers again: https://travis-ci.org/ros-planning/srdfdom/builds/154885328
@130s: IMHO you could release now. @davetcoleman?

[davetcoleman]

This is a violation of our review policy and standard software procedures - why are you committing some much unreviewed stuff to this repo?

[rhaschke]

I tried to hot-fixed the merge, which broke Travis of the main Moveit repo. That's why some urgent action was required. Before doing the original merge, I of course tested locally, and surprisingly I didn't observe any issues. Still not sure why.
Unfortunately, Travis was disabled for this repo, such that the issues didn't show up here automatically.

[130s]

Will make a release 0.3.2 now.

Yeah, I have to agree with @davetcoleman about direct commits. Maybe @rhaschke you could have hot-fixed and then if that worked you revert and open a PR?

[v4hn]

I also hot-fixed things in the past and I don't think it's fatal.
But hot-fixes should be limited to single commits.

a1dd532 is definitely no "hot-fix"
and enabling proper travis support isn't either.

[rhaschke]

a1dd532 is definitely no "hot-fix"

What else? The call to exit() and the use of ros/console.h, which wasn't declared a dependency, broke building of the package. For this reason unrelated PRs of the main moveit repo failed on Travis:
https://travis-ci.org/ros-planning/moveit/pull_requests (237-238). Because of my hot-fix, the down-time was kept as short as possible.

[v4hn]

On Thu, Aug 25, 2016 at 05:41:40PM -0700, Robert Haschke wrote:

The call to exit() and the use of ros/console.h, which wasn't declared a dependency, broke building of the package.

Then why didn't you write the explanation in the commit message? :-)
Please do that next time.

[130s]

0.3.2 along with this change now made to shadow repo for Indigo and JK.