4 ➡️ 5

[chapulina]

Port ign-rendering4 to main.

Branch comparison: main...ign-rendering4

Note to maintainers: Remember to Merge with commit (not squash-merge or rebase)

[codecov]

Codecov Report

Merging #246 (13db3f6) into main (c6fc79f) will increase coverage by 0.24%.
The diff coverage is 78.76%.

@@            Coverage Diff             @@
##             main     #246      +/-   ##
==========================================
+ Coverage   57.01%   57.25%   +0.24%     
==========================================
  Files         151      151              
  Lines       15010    15181     +171     
==========================================
+ Hits         8558     8692     +134     
- Misses       6452     6489      +37     

Impacted Files	Coverage Δ
include/ignition/rendering/ArrowVisual.hh	100.00% <ø> (ø)
include/ignition/rendering/AxisVisual.hh	100.00% <ø> (ø)
include/ignition/rendering/Camera.hh	100.00% <ø> (ø)
include/ignition/rendering/CompositeVisual.hh	100.00% <ø> (ø)
include/ignition/rendering/Geometry.hh	100.00% <ø> (ø)
include/ignition/rendering/Image.hh	100.00% <ø> (ø)
include/ignition/rendering/Light.hh	100.00% <ø> (ø)
include/ignition/rendering/Material.hh	100.00% <ø> (ø)
include/ignition/rendering/Mesh.hh	100.00% <ø> (ø)
include/ignition/rendering/Node.hh	100.00% <ø> (ø)
... and 12 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c6fc79f...13db3f6. Read the comment docs.

[chapulina]

This new test failure could be related to #189 or #235:

/Users/jenkins/workspace/ignition_rendering-ci-pr_any-homebrew-amd64/ign-rendering/test/integration/thermal_camera.cc:233
The difference between boxTemp and temp is 14, which exceeds boxTempRange, where
boxTemp evaluates to 310,
temp evaluates to 296, and
boxTempRange evaluates to 3.

[adlarkin]

This new test failure could be related to #189 or #235

I took a look at homebrew CI. The failure is for the ThermalCameraBoxesUniformTemp test, which was actually already implemented before #189 and #235. That being said though, #189 and #235 did change some of the code slightly for processing uniform temperatures, so maybe these PRs are indeed related to this new failure. It's interesting that this failure only happened on MacOS and not Ubuntu. All checks passed on #235 as well, so I am not sure what's going on here.

@iche033 any ideas?

[chapulina]

this failure only happened on MacOS and not Ubuntu.

I haven't looked closely into the code, but when this happens, I usually look for anything in the test that relies on timing, because the Homebrew CI machines are usually less powerful than the Ubuntu ones and may require more sleeping.

In any case, I retriggered the build to see if the test passes this time. I don't think we need to block this PR on a flaky test, but I think it would be good to address the flakiness back on ign-rendering4.

[adlarkin]

I retriggered the build to see if the test passes this time

It looks like the test still failed on MacOS CI. I'm taking a closer look at this and will post an update here if I find anything that needs to be addressed. Let's hold off on merging this for now in case something needs to be fixed.

[iche033]

hmm not sure why. The tests seem to be passing in ign-rendering4, e.g. in #241. Maybe we've changed something in edifice.

[adlarkin]

I tried to do some debugging about the new test failure. I opened up two draft PRs in order to trigger/compare CI (feel free to take a look at the comments I left in these PRs for more details):

• Fix flaky thermal camera test on MacOS #247: Branches off of this PR in order to do more digging into the specifics of the test failure
• Get current status of MacOS tests on edifice #248: Branches off of main (edifice) to see what the current CI results are without this forward port

I found that all tests pass on #248, but I am still seeing the MacOS failure on #247 even after trying a few workarounds. This makes me think that the new test failure is due to a combination of change(s) in this forward port and changes in edifice, since tests were passing on the dome and edifice branches without this forward port.

The interesting thing is that based on my testing in #247, it looks like only one pixel in the image fails the check (all of the other pixels pass), and I have no idea why (I'd expect that all pixels would either pass or fail, but not a mixture of each since the box should take up the whole image at this point in the test): #247 (comment)

I'm out of ideas at this point. Does anyone have any thoughts on what may be going on here?

[chapulina]

I'll merge this to get some PRs moving forward while you peeps are looking into the thermal camera failure.

[adlarkin]

I'll merge this to get some PRs moving forward while you peeps are looking into the thermal camera failure.

I create an issue for this so that it can be tracked: #253