-
Notifications
You must be signed in to change notification settings - Fork 935
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Noetic changelogs and version bump (1.1.2) #2423
Conversation
Codecov Report
@@ Coverage Diff @@
## master #2423 +/- ##
==========================================
+ Coverage 60.47% 60.55% +0.09%
==========================================
Files 351 351
Lines 26434 26462 +28
==========================================
+ Hits 15982 16021 +39
+ Misses 10452 10441 -11
Continue to review full report at Codecov.
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, but why does Travis fail on clang?
We didn't had a single successful master branch build on Travis since 18 days, namely since the Pilz planner (#1893) was merged! Additionally, there is a flaky unit test:
|
I started debugging the test that is flaky yesterday but don't have a fix for it. Some testing yesterday showed that that notice on Travis is not telling the truth and it is running ci (for now atleast). I have started working on circleci and am using GitHub actions on one customer project. I have yet to get better performance out of circleci and wouldn't recommend it yet. GitHub actions seems to be faster than Travis (Microsoft might be throwing money at more powerful machines to try to get people to switch). To use GitHub actions the easiest way will be to use industrial_ci. There are a handful of things we have in moveit_ci that we would lose until we get those things upstreamed. One of them is code coverage which is almost in industrial_ci. Another decision we should consider is (grasp) breaking up the repo. One sort of obvious thing we could factor out is anything that depends on qt. That would make for faster builds and have the advantage that people who depend on moveit but don't want the qt stuff but don't have them on their target can stop having to deal with our qt dependency. On top of all this the noetic fcl thing isn't building and causing problems for the official buildfarm. |
What's the culprit?
Can you provide any evidence, e.g. a link to a buildfarm build? |
I don't know yet. |
@rhaschke The PR builds before the merge did succeed. But the check takes veeryy long:
|
Picknik now paying for the 1 runner plan on Travis to have a temporary solution to the timeouts. This will probably make CI take longer but not timeout. In the long-ish term I will be investigating the other providers and doing work to support them to try to find a better solution than Travis. |
@tylerjw: The ROS build farm is complaining about missing pilz packages, e.g.: moveit_resources_prbt_moveit_config Didn't you already release those? |
@tylerjw It seems, there hasn't been a release of ros-planning/moveit_resources since october?! |
I remember releasing moveit_resources when I did the last noetic release. It appears I didn't though. We should release it so the build farm is happy. |
@jschleicher I am planning on doing a release once some of the tests become more robust. With the current state of our runners on travis the move_group interface test and subframes test has become flaky. I'm working with @bostoncleek to get that resolved so it stops blocking the OMPL constrained space PR and many others. I'm hoping that we'll have a solution for that within the next couple weeks. |
@tylerjw We're still waiting for a binary release including the pilz_industrial_motion_planner. Is there any news on your side?
|
closing this as 1.1.2 just went out |
Description
Changelogs and version bump for Noetic release 1.1.2