show_frame: Use workaround updateFrame pending upstream merge

[EricCousineau-TRI]

Upstream PR:
RobotLocomotion/director#621

For issue, see Slack discussion:
https://drakedevelopers.slack.com/archives/C43KX47A9/p1554239704004700

EDIT (2019/10/07):

Copy of Slack transcript:

jwnimmer-tri Apr 2nd at 5:15 PM
One of our d-v plugs seems to have bitrotted?  Does this ring a bell for anyone?
drake/tools/workspace/drake_visualizer/plugin/show_frame.py", line 36, in handle_message
    vis.updateFrame(transform, name, parent=folder, scale=0.1)
  File ".../external/drake_visualizer/lib/python2.7/site-packages/director/visualization.py", line 879, in updateFrame
    obj.copyFrame(frame)
AttributeError: 'ContainerItem' object has no attribute 'copyFrame'
(edited)

eric.cousineau  6 months ago
Er... Nuh uh. Can I ask if there's a repro recipe?
Will try on Anzu with latest Drake master.

jwnimmer-tri  6 months ago
bazel-bin/tools/drake_visualizer &
bazel-bin/examples/kuka_iiwa_arm/kuka_simulation &
(edited)

eric.cousineau  6 months ago
Huh... oddly enough, if I run those two without a sleep, it renders the frame, but misses the load message:
(
    set -eux
    trap 'kill -9 $(jobs -p)' SIGINT
    bazel-bin/tools/drake_visualizer &
    bazel-bin/examples/kuka_iiwa_arm/kuka_simulation &
    wait
)
(edited)

eric.cousineau  6 months ago
But if I inject sleep 2 between them, I can reproduce your error...

jwnimmer-tri  6 months ago
When it happened, I turned off that plug-in to stop the spam.  Turning it back on again, there was no more spam even after restarting the simulation.  Something somewhere is stateful.

jwnimmer-tri  6 months ago
I am OK to ignore this for now, it was easy enough to work around.

eric.cousineau  6 months ago
Aye.

eric.cousineau  6 months ago
It's b/c the simulation publishes iiwa_link_ee, which is also a link elsewhere in the object model.
If I change the names published, it's fine.

eric.cousineau  6 months ago
Yeah, updateFrame completely discards the parent=folder arg there. Meh. Will submit a 2min fix.

---

This change is 

[EricCousineau-TRI]

+@jwnimmer-tri for both reviews, please.

Reviewable status: all discussions resolved, LGTM missing from assignee jwnimmer-tri, platform LGTM missing (waiting on @jwnimmer-tri)

[jwnimmer-tri]

TBH None of this code is automatically tested so I'm hesitant to patch this without upstream merging the fix first, in case the fix isn't really good enough, or has other problems. And I'm not planning on rebuilding and running the demo where I saw this anytime soon.

Reviewed 1 of 1 files at r1.
Reviewable status: all discussions resolved, LGTM missing from assignee jwnimmer-tri, platform LGTM missing

[EricCousineau-TRI]

SGTM. I think we should up the priority if another user runs into it, but I'm fine with this otherwise chilling pending upstream updates.

[EricCousineau-TRI]

(Or closing this for the time being, as long as it's easy to refer back to.)

[jwnimmer-tri]

Sure. They would be in a good position to test it, in that case.

[EricCousineau-TRI]

Aye. Closing, and will post a tracer-comment in #visualization Slack channel.