-
Notifications
You must be signed in to change notification settings - Fork 44
#1588 Fixing up Android buffer problem and failing tests #1589
Conversation
SocketThreadBase - Turns out that the logic that got checked in could end up setting buffers to zero because the Bluetooth socket get size calls could return 0. For now I decided to just remove the dynamic size setting to simplify things. This should hopefully fix Android. testThaliMobileNative - We have a condition where we could test something after we have called t.end but this is necessary because our peer might be done with the test but the other might not. So now the test will only cause a failure if it actually fails in which case we should fail. testThaliMobileNativeWrapper - In addition to an identical issue to the one described above we also had tests that were written before the skip functionality so I fixed that. Also some of the test names were so similar it made doing searches hard so I added some text to make them different. CoordinatedClient - Added the original result object to the error makeIntoCloseAllServer - Linting
PR is added to the queue for testing as 1. task. (6721f4c) |
Test 96008345 (6721f4c) build started. |
Test (Fail) 96008345 build is completed (6721f4c) See https://github.com/ThaliTester/TestResults/tree/960083456721f4c__1588_Fixing_up_Android_buffer_problem_and_failing_tests_yaronyg/ for the logs |
Reviewed 5 of 5 files at r1. test/www/jxcore/bv_tests/testThaliMobileNativeWrapper.js, line 448 at r1 (raw file):
It should be Comments from Reviewable |
PR is added to the queue for testing as 1. task. (ac97228) |
Test 96008345 (ac97228) build started. |
Test (Success) 96008345 build is completed (ac97228) See https://github.com/ThaliTester/TestResults/tree/96008345ac97228__1588_Fixing_up_Android_buffer_problem_and_failing_tests_yaronyg/ for the logs |
Test 96008345ac97228(ac97228) has failed See https://github.com/ThaliTester/TestResults/tree/96008345ac97228__1588_Fixing_up_Android_buffer_problem_and_failing_tests_yaronyg/ for the fail logs |
Chapko pointed out that using isMobile is silly since that is literally always true, what I was looking for is _isRealMobile
PR is added to the queue for testing as 1. task. (c2441a7) |
Test 96008345 (c2441a7) build started. |
Review status: 3 of 6 files reviewed at latest revision, 1 unresolved discussion. test/www/jxcore/bv_tests/testThaliMobileNativeWrapper.js, line 448 at r1 (raw file): Previously, chapko (Eugene Chapko) wrote…
An excellent point! Fixed Comments from Reviewable |
Test (Success) 96008345 build is completed (c2441a7) See https://github.com/ThaliTester/TestResults/tree/96008345c2441a7__1588_Fixing_up_Android_buffer_problem_and_failing_tests_yaronyg/ for the logs |
Test 96008345c2441a7(c2441a7) has failed See https://github.com/ThaliTester/TestResults/tree/96008345c2441a7__1588_Fixing_up_Android_buffer_problem_and_failing_tests_yaronyg/ for the fail logs |
SocketThreadBase - Turns out that the logic that got checked in could end up setting buffers to zero
because the Bluetooth socket get size calls could return 0. For now I decided to just remove the dynamic
size setting to simplify things. This should hopefully fix Android.
testThaliMobileNative - We have a condition where we could test something after we have called t.end but
this is necessary because our peer might be done with the test but the other might not. So now the test
will only cause a failure if it actually fails in which case we should fail.
testThaliMobileNativeWrapper - In addition to an identical issue to the one described above we also had tests
that were written before the skip functionality so I fixed that. Also some of the test names were so similar it
made doing searches hard so I added some text to make them different.
CoordinatedClient - Added the original result object to the error
makeIntoCloseAllServer - Linting
This change is