fix(common/core/web): error from early fat-finger termination due to OS interruptions 🍒#5491
Conversation
|
Huh... |
mcdurdin
left a comment
There was a problem hiding this comment.
LGTM. Any sanity user testing we can do?
"Can do?" We can always run the same tests from the PR against |
I'm looking more for regression tests -- make sure nothing else gets broken. I'm fairly confident that this is true, but can you add a user test for Makara similar to the one we did for master? |
Gave it the ol' |
|
Eh, that's fine. (The |
User Testing [retest on Android for good measure]@keymanapp/testers Platform: Android 6.0.1 emulated on Nexus 5 Test build #14.0.280-test at 2 Aug 01:52
|
|
Retested for good measure using the stable build #14.0.280-test at 2 Aug 01:52. Confirm that all tests have PASSED. Things are working as expected. |


A 🍒-pick of #5479.
User Testing
@keymanapp/testers
It's literally the same set of tests that was used for #5479.
Much of this is taken from standard KMW acceptance testing and adapted toward the aspects of KMW modified by this PR, though we'll only worry about testing against a single platform here.
Platform: Chrome emulation / Android
Ensure that you see the "Console" tab while performing these tests.
Console tab:
It should resemble Windows'
CMDprompt and the macOS Terminal.Utilize the "Test unminified Keymanweb" testing page to ensure the following:
sil_ipa. (For our currently-deprecated IPA keyboard.)ŋshould be a subkey ofn- ensure it works.khmer_angkorspanish.Utilize the "Prediction - robust testing" testing page for the following:
.to output'and then presse. The two characters should not combine.pand SHIFT + long-pressgdisplays properly and that the subkeys produce the expected output.L(shift layer)k(default layer)v(default layer)LoveandLive, along with one other random suggestion.