Added failing test for #5111 #5157
Conversation
|
Is this bug still present in ED@2.15.0 ? |
workmanw
force-pushed
the
failing-test-for-5111
branch
from
4bbbec7
to
2eb64f1
Compare
|
@urbany Unfortunately so. I just rebased against master and pushed so you'll see a travis failure here shortly. I also ran the tests locally and this test is still failing. |
workmanw
force-pushed
the
failing-test-for-5111
branch
2 times, most recently
from
bfb9e58
to
7d5d942
Compare
| run(() => { | ||
| allPeople.objectAt(0).unloadRecord(); | ||
|
|
||
| assert.equal(get(allPeople, 'length'), 2, 'Unload does not complete until the end of the loop'); |
There was a problem hiding this comment.
i think we want this to happen before the loop ends to reduce # of wtfs
There was a problem hiding this comment.
I definitely agree that that would be more ideal. I had an extended conversation with @stefanpenner about that on slack. IIRC his near term objective was to get things back to how they were at 2.12, then later down the line rethink some of this stuff. That was several months back so it may have changed. But that was the thought process behind this test.
workmanw
force-pushed
the
failing-test-for-5111
branch
from
7d5d942
to
aa0d0bf
Compare
workmanw
force-pushed
the
failing-test-for-5111
branch
from
aa0d0bf
to
1db6b65
Compare
|
I've confirmed that #5378 will resolve this issue, and that PR has a test covering this case. Thanks so much for the failing test, made things easy to investigate! |
No description provided.