Closed Bug 1000870 Opened 11 years ago Closed 9 years ago

Tests for check support Pointer Events

Categories

(Core :: DOM: Events, defect)

Product:
Core
Component:
DOM: Events
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED FIXED
Milestone:
mozilla48
Tracking Flags:
Tracking Status
firefox48 --- fixed

People

(Reporter: alessarik, Assigned: alessarik)

References

(Depends on 1 open bug)

Details

Attachments

(3 files, 20 obsolete files)

original_official_tests_ver2.diff
295.44 KB, patch
smaug
: review+
Details | Diff | Splinter Review
testing_feauters_ver4.diff
1.48 KB, patch
smaug
: review+
Details | Diff | Splinter Review
off_tests_ver20.diff
225.85 KB, patch
smaug
: review+
Details | Diff | Splinter Review 
FireFox should have internal auto tests for check support Pointer Events during build
Attached patch off_tests_ver5.diff (obsolete) — — Splinter Review 
Some of test to check support Pointer Events
Attachment #8411798 - Flags: feedback?(romaxa)
Attachment #8411798 - Flags: feedback?(nicklebedev37)
Attachment #8411798 - Flags: feedback?(bugs)
So this is waiting for w3c tests to be reviewed.
Component: Untriaged → DOM: Events
Product: Firefox → Core
Comment on attachment 8411798 [details] [diff] [review]
off_tests_ver5.diff

Do you have try build with all these tests passing?
Overall it looks great to me.
Attachment #8411798 - Flags: feedback?(romaxa) → feedback+
(In reply to Oleg Romashin (:romaxa) from comment #3)
> Do you have try build with all these tests passing?
https://tbpl.mozilla.org/?tree=Try&rev=3b3cfeb83a2b
But most of tests were finished by timeout without any results.
So what parts of the tests are coming from w3c test suite (which is still being reviewed, I believe)?
Comment on attachment 8411798 [details] [diff] [review]
off_tests_ver5.diff

Review of attachment 8411798 [details] [diff] [review]:
-----------------------------------------------------------------

We have similar bug about pointer events tests: 966961.
I believe we need to either tie them or close one of them as duplicate.
Attachment #8411798 - Flags: feedback?(nicklebedev37) → feedback+
Are all the w3c tests now reviewed?
Attached patch off_tests_ver6.diff (obsolete) — — Splinter Review 
All test were got from
https://github.com/w3c/web-platform-tests/tree/master/pointerevents

I made minimal changes (to support mochitest system)
Some test were passed succesfully. Some were failed.
I try to check fails.

https://tbpl.mozilla.org/?tree=Try&rev=43d29e9c5602
Attachment #8411798 - Attachment is obsolete: true
Attachment #8411798 - Flags: feedback?(bugs)
Attachment #8446538 - Flags: review?(peterv)
Attachment #8446538 - Flags: review?(jst)
Attachment #8446538 - Flags: review?(bugs)
Attachment #8446538 - Flags: review?(peterv)
Attachment #8446538 - Flags: review?(jst)
Comment on attachment 8446538 [details] [diff] [review]
off_tests_ver6.diff

So the tests don't pass on try.
Attachment #8446538 - Flags: review?(bugs)
Also, would it be possible to get the diff for the changes you've made to the original files?
That would make reviewing _a_lot_ easier.
Attached patch off_tests_ver7.diff (obsolete) — — Splinter Review 
+ Some changes in test were added
Attachment #8446538 - Attachment is obsolete: true
Attachment #8451027 - Flags: review?(jonas)
Attachment #8451027 - Flags: review?(bzbarsky)
Attachment #8451027 - Flags: review?(bugs)
Attached patch testing_features_ver1.diff (obsolete) — — Splinter Review 
+ Some changes which needs for correct testing
Attachment #8451033 - Flags: review?(jonas)
Attachment #8451033 - Flags: review?(bzbarsky)
Attachment #8451033 - Flags: review?(bugs)
Attachment #8451033 - Flags: feedback?(oleg.romashin)
Attachment #8451033 - Flags: feedback?(nicklebedev37)
Attached file original_official_tests.diff (obsolete) — 
Original official tests from W3C at 2014.07.01
Attachment #8451027 - Flags: review?(bzbarsky)
Attachment #8451033 - Flags: review?(bzbarsky)
Attachment #8451033 - Flags: review?(bugs) → review+
Comment on attachment 8451027 [details] [diff] [review]
off_tests_ver7.diff


>-MOCHITEST_MANIFESTS += ['test/mochitest.ini']
>+MOCHITEST_MANIFESTS += [
>+#    'test/mochitest.ini',
Why this? 


But I can't really review this all.
Please provide a diff which shows which changes you have made to the original w3c tests.
Attachment #8451027 - Flags: review?(bugs) → review-
Comment on attachment 8452251 [details]
original_official_tests.diff

So if you have these, then do diff comparing all the files in this patch
to the patch you asked a review for.
Or interdiff of this patch and the other patch might work too.
(In reply to Olli Pettay [:smaug] from comment #15)
> >-MOCHITEST_MANIFESTS += ['test/mochitest.ini']
> >+MOCHITEST_MANIFESTS += [
> >+#    'test/mochitest.ini',
> Why this? 
Only for testing. In finish version will be at normal state.
Attached patch original_official_tests.diff (obsolete) — — Splinter Review 
Initial state of tests.

TRY: https://tbpl.mozilla.org/?tree=Try&rev=524acfdfdd22
Attachment #8452251 - Attachment is obsolete: true
Attachment #8455950 - Flags: review?(bugs)
Attachment #8455950 - Flags: feedback?(oleg.romashin)
Attachment #8455950 - Flags: feedback?(nicklebedev37)
Comment on attachment 8455950 [details] [diff] [review]
original_official_tests.diff

I really need the interdiff, in other words I should review only the 
changes you've made and I could rely on the
tests from w3c being right (mbrubeck et al have reviewed those).
Attachment #8455950 - Flags: review?(bugs)
Attached patch off_tests_ver8.diff (obsolete) — — Splinter Review 
+ Changes: Patch is difference over original_official_tests.diff
Attachment #8451027 - Attachment is obsolete: true
Attachment #8463978 - Flags: review?(bugs)
Comment on attachment 8463978 [details] [diff] [review]
off_tests_ver8.diff


>                         test(function() {
>-                            assert_true(event.button == -1, "If mouse buttons are released button attribute is -1")
>+                            assert_equals(event.button, -1, "If mouse buttons are released button attribute is -1")
>                         }, "If mouse buttons are released button attribute is -1");
>                         test(function() {
>-                            assert_true(event.buttons == 0, "If mouse buttons are released buttons attribute is 0")
>+                            assert_equals(event.buttons, 0, "If mouse buttons are released buttons attribute is 0")
>                         }, "If mouse buttons are released buttons attribute is 0");

Why these changes?



>@@ -79,15 +85,15 @@
>             function run() {
>                 on_event(target0, "pointerover", function (event) {
>                     log("pointerover", target0);
>                     if(isPointerCapture) {
>                         test_pointerover_capture.done();
>                         if (!isRelatedTargetValueTested) {
>                             test(function() {
>-                                assert_true(event.relatedTarget==null, "relatedTarget is null when the capture is set")
>+                                assert_equals(event.relatedTarget, null, "relatedTarget is null when the capture is set")
I don't understand this change.



Those fixed, I think we could get this in.
(We may want to change the setup at some point, but w3c harness lacks some features.)
Attachment #8463978 - Flags: review?(bugs) → review+
(In reply to Olli Pettay [:smaug] from comment #21)
> Why these changes?
> I don't understand this change.
In this case we can read more information in logs.
For example, we can see actual result and expected value.
Without this changes, we can see only "FAIL" message without details of error.
But we want as few changes as possible, so that we can update the test easily if there are
changes to w3c tests.
Attachment #8455950 - Flags: feedback?(nicklebedev37)
Attachment #8451033 - Flags: feedback?(nicklebedev37)
Maksim, any update here.
Flags: needinfo?(alessarik)
Unfortunately, I have no update for this at this time.
But we should get new version of tests from official repository.
There are many changes in repo. But original_official_tests.diff is very older.
Maksim, would be good to get the tests landed.
(In reply to Olli Pettay [:smaug] from comment #26)
> Maksim, would be good to get the tests landed.
I will try to synchronize tests with latest changes in official repo.
Blocks: 960316
Test suite from official repo at 2015.03.01
Attachment #8455950 - Attachment is obsolete: true
Attachment #8455950 - Flags: feedback?(oleg.romashin)
Attachment #8586103 - Flags: review?(bugs)
Comment on attachment 8451033 [details] [diff] [review]
testing_features_ver1.diff

This changes already were pushed in another bug.
Attachment #8451033 - Attachment is obsolete: true
Attachment #8451033 - Flags: feedback?(oleg.romashin)
Comment on attachment 8586103 [details] [diff] [review]
original_official_tests_ver2.diff

I don't see how this patch makes us to run the tests automatically.
Are these the manual-only tests? If so, I don't think we need those in the tree.

Or do I misunderstand something here now.
Attachment #8586103 - Flags: review?(bugs) → review-
So we need tests, but we need tests to be run automatically.
Attached patch testing_feauters_ver2.diff (obsolete) — — Splinter Review 
+ Add posibility to send no-pressed button in event
Attachment #8587496 - Flags: review?(bugs)
(In reply to Olli Pettay [:smaug] from comment #30)
> I don't see how this patch makes us to run the tests automatically.
> Are these the manual-only tests? If so, I don't think we need those in the tree.
> 
> Or do I misunderstand something here now.
You always asked (in comment 10, comment 15, comment 16, comment 19)
to provide two patches: official patch and patch with small differencies.
Attached patch off_tests_ver9.diff (obsolete) — — Splinter Review 
+ Each test now has own support file (To decrease changes in official tests and
    turn on pointer events in some tests before loading page)
- Script testharnessreport.js was removed from each official test.
    (Each official test is loading in iframe.
    In this case testharnessreport.js generates exception).
Attachment #8463978 - Attachment is obsolete: true
Attachment #8587510 - Flags: review?(bugs)
Attachment #8587496 - Flags: review?(bugs) → review+
Comment on attachment 8587510 [details] [diff] [review]
off_tests_ver9.diff

So I need this to be split to two patches. First one patch containing all the tests without any changes. That I would r+ pretty much automatically (assuming there is the other patch to review too) (perhaps attachment 8586103 [details] [diff] [review] is that first patch?)
then another patch on top of that adding all the changes need to run this stuff automatically. I would expect the second patch to be a lot smaller than the first one.
Attachment #8587510 - Flags: review?(bugs) → review-
(In reply to Maksim Lebedev from comment #33)
> You always asked (in comment 10, comment 15, comment 16, comment 19)
> to provide two patches: official patch and patch with small differencies.
Exactly. And I saw only the official tests -patch in the queue, not the patch which does the changes.
I'd like to see them both.
(In reply to Olli Pettay [:smaug] from comment #35)
> So I need this to be split to two patches.
I provided three patches for more convinience.
> First one patch containing all the tests without any changes.
It is "original_official_tests.diff" with some version.
> That I would r+ pretty much automatically
That patch got r-, unfortunately.
> then another patch on top of that adding all the changes need to run this stuff automatically.
It is "off_tests.diff" with some version.
> I would expect the second patch to be a lot smaller than the first one.
Unfortunately, it has a lot of additionally info.
(For example, the largest file in folder is "mochitest.ini". It is situated in "off_tests" patch).
Also, another changes were discribed in comment 34.

(In reply to Olli Pettay [:smaug] from comment #36)
> I'd like to see them both.
Both patches were provided.
And also third patch was provided with changes in CPP source code for more comfortable review.
(In reply to Maksim Lebedev from comment #37)
> It is "original_official_tests.diff" with some version.
> > That I would r+ pretty much automatically
> That patch got r-, unfortunately.
Because at that time there wasn't another patch to make those tests automatically testable.


> > then another patch on top of that adding all the changes need to run this stuff automatically.
> It is "off_tests.diff" with some version.
Ahaa, now I'm starting to see your setup

ok, reviewing again
Comment on attachment 8586103 [details] [diff] [review]
original_official_tests_ver2.diff

rs+ (rubberstamp)
Attachment #8586103 - Flags: review- → review+
Attachment #8587510 - Flags: review- → review?(bugs)
(In reply to Olli Pettay [:smaug] from comment #38)
> Because at that time there wasn't another patch to make those tests automatically testable.
There were some issues with work of auto test system. That issues now are discribed in bug 1150091.
Do we need bug 1150091 fixed before the patches in this bug can land? Or if bug 1150091 doesn't get fixed soon, should the patch in this bug be changed?
Comment on attachment 8587510 [details] [diff] [review]
off_tests_ver9.diff

Apparently a new patch coming.
Attachment #8587510 - Flags: review?(bugs)
Blocks: 1122211
Attached patch testing_feauters_ver3.diff (obsolete) — — Splinter Review 
+ Update code according with GetButtonsFlagForButton() -> nsContentUtils::GetButtonsFlagForButton()
Assignee: nobody → alessarik
Attachment #8587496 - Attachment is obsolete: true
Status: UNCONFIRMED → NEW
Ever confirmed: true
Attachment #8606376 - Flags: review?(bugs)
Attached patch off_tests_ver10.diff (obsolete) — — Splinter Review 
+ Four tests were disabled for future investigation
+ Some issues in tests were resolved

Suggestions and comments and objections are very welcome.
Attachment #8587510 - Attachment is obsolete: true
Flags: needinfo?(alessarik)
Attachment #8606386 - Flags: review?(bugs)
Attachment #8606386 - Flags: feedback?(jmathies)
Comment on attachment 8606376 [details] [diff] [review]
testing_feauters_ver3.diff

actually, could you add a new flag to buttonsFlag, 
eNoButtonFlag = 0x00, and return that in GetButtonsFlagForButton.
It would make the code a bit easier to understand.

With that change, r+
Attachment #8606376 - Flags: review?(bugs) → review+
Comment on attachment 8606386 [details] [diff] [review]
off_tests_ver10.diff

>+// Helper function to send PointerEvent with different parameters
>+function sendPointerEvent(int_win, elemId, pointerType, inputSource, params) {
>+  var elem = int_win.document.getElementById(elemId);
>+  if(!!elem) {
>+    var rect = elem.getBoundingClientRect();
>+    var eventObj = {type: pointerType, inputSource: inputSource};
>+    if(params && "button" in params)
>+      eventObj.button = params.button;
>+    if(params && "isPrimary" in params)
>+      eventObj.isPrimary = params.isPrimary;
>+    else if(SpecialPowers.Ci.nsIDOMMouseEvent.MOZ_SOURCE_MOUSE == inputSource)
You don't need SpecialPowers here.
Just MouseEvent.MOZ_SOURCE_MOUSE.
Similar also elsewhere.


>+      eventObj.isPrimary = true;
>+    console.log(elemId, eventObj);
>+    var salt = ("pointermove" == pointerType) ? 1 : 2;
I don't understand this. how could pointerType be "pointermove"?
So, I guess pointerType should be renamed to pointerEventType


Mostly rs+ (rubberstamp+)
manual tests won't be run automatically, but better to have them in tree.
Attachment #8606386 - Flags: review?(bugs)
Attachment #8606386 - Flags: review+
Attachment #8606386 - Flags: feedback?(jmathies)
Make sure to run the tests couple of times on tryserver before the patch is pushed to mozilla-inbound.
+ Added WidgetMouseEvent::buttonsFlag::eNoButtonFlag
Attachment #8606376 - Attachment is obsolete: true
Attachment #8607535 - Flags: review?(bugs)
Attached patch off_tests_ver11.diff (obsolete) — — Splinter Review 
+ Changed pointerType -> pointerEventType in sendPointerEvent()
+ Changed SpecialPowers.Ci.nsIDOMMouseEvent.MOZ_SOURCE_* -> MouseEvent.MOZ_SOURCE_*

Suggestions and comments and objections are very welcome.
Attachment #8606386 - Attachment is obsolete: true
Attachment #8607539 - Flags: review?(bugs)
Comment on attachment 8607539 [details] [diff] [review]
off_tests_ver11.diff

interdiff to the previous version of the patch would have been nice ;)

Mostly rs+ based on the comment about the changes.
Attachment #8607539 - Flags: review?(bugs) → review+
Attachment #8607535 - Flags: review?(bugs) → review+
Attached patch off_tests_ver12.diff (obsolete) — — Splinter Review 
+ Added small changes at "pointerevent_capture_mouse-manual.html" test

Suggestions and comments and objections are very welcome.
Attachment #8607539 - Attachment is obsolete: true
Attachment #8623612 - Flags: review?(bugs)
Comment on attachment 8623612 [details] [diff] [review]
off_tests_ver12.diff

What is the status of this bug? What is blocking landing patches?
Attachment #8623612 - Flags: review?(bugs) → review+
(In reply to Olli Pettay [:smaug] from comment #56)
> What is the status of this bug? What is blocking landing patches?
Looks like a most part of tests have stable results, so we can land this changes.
If there are no objections, I put checkin-needed flag.
Keywords: checkin-needed
sorry had to back this out for test failures like https://treeherder.mozilla.org/logviewer.html#?job_id=10960451&repo=mozilla-inbound
Flags: needinfo?(alessarik)
Attached patch off_tests_ver13.diff (obsolete) — — Splinter Review 
+ Added skipping two tests on linux platform

Suggestions and comments and objections are very welcome.
Attachment #8623612 - Attachment is obsolete: true
Attachment #8625739 - Flags: review?(bugs)
Looks like test results from Comment 65 have no issues related with current patches.
Flags: needinfo?(cbook)
Any idea why that test might fail on linux?
Comment on attachment 8625739 [details] [diff] [review]
off_tests_ver13.diff

But we can fix that in a different bug to get these test landed sooner than later.
Please file a followup to figure out those linux-only issues.
Attachment #8625739 - Flags: review?(bugs) → review+
(In reply to Olli Pettay [:smaug] from comment #68)
> Any idea why that test might fail on linux?
I have no prepared environment on linux. At least at current time.
Fails looks very strange because it happens unstable and probably only in debug mode.
So maybe it need deeper investigation.
If there are no objections, I put checkin-needed flag.
Status: NEW → ASSIGNED
Flags: needinfo?(alessarik)
Keywords: checkin-needed
Backed out for the test failures noted in bug 1151152.
https://hg.mozilla.org/mozilla-central/rev/716df1829f4f
Flags: needinfo?(cbook)
Attached patch off_tests_ver14.diff (obsolete) — — Splinter Review 
+ Skiped [test_pointerevent_capture_suppressing_mouse-manual.html] on B2G Emulator
+ Changes in [test_pointerevent_setpointercapture_inactive_button_mouse-manual.html]
+ Added some comment about bugs and issues

Suggestions and comments and objections are very welcome.
Attachment #8625739 - Attachment is obsolete: true
Attachment #8629357 - Flags: review?(bugs)
Attachment #8629357 - Flags: review?(bugs) → review+
Attached patch off_tests_ver15.diff (obsolete) — — Splinter Review 
+ Skiped some tests on B2G ICS Emulator
+ Added small correcting changes into tests

Suggestions and comments and objections are very welcome.
Attachment #8629357 - Attachment is obsolete: true
Attachment #8631465 - Flags: review?(bugs)
Comment on attachment 8631465 [details] [diff] [review]
off_tests_ver15.diff

Please explain those changes to the tests. Why we need them?

But I can rs+ this anyhow.
Attachment #8631465 - Flags: review?(bugs) → review+
(In reply to Olli Pettay [:smaug] from comment #83)
> Please explain those changes to the tests. Why we need them?
Issues were in test related with lostpointercapture events, because such events work as async I have added some syntetized pointermove events for better working.
If there are no objections, I put checkin-needed flag. Let's try to push it again.
Keywords: checkin-needed
(In reply to Maksim Lebedev from comment #85)
> If there are no objections, I put checkin-needed flag. Let's try to push it
> again.

I object. Your last Try run shows failures on both Android and B2G debug.

https://treeherder.mozilla.org/logviewer.html#?job_id=9173724&repo=try
https://treeherder.mozilla.org/logviewer.html#?job_id=9175530&repo=try
Keywords: checkin-needed
(In reply to Ryan VanderMeulen [:RyanVM UTC-4] from comment #86)
> I object. Your last Try run shows failures on both Android and B2G debug.
Is there any construction for skip all tests on B2G ICS emulator or I should write "skip" word into each small test?
Flags: needinfo?(ryanvm)
You can add |skip-if = toolkit == 'gonk'| to the [DEFAULT] section of the manifest to skip all tests on B2G emulator builds. If you want to skip on only debug, it'd be |skip-if = (toolkit == 'gonk' && debug)|.

Note that there are intermittent Android failures as well. Frankly, I'd like to see multiple (10+) green retriggers before landing this to avoid new intermittents.

Also, can you please try to use more targeted Try pushes for this instead of running all builds and all tests across all platforms? Doing so uses over 300hr of machine time.
Flags: needinfo?(ryanvm)
Attached patch off_tests_ver16.diff (obsolete) — — Splinter Review 
+ Moved "skip-if" statement from several separated tests into default section

Suggestions and comments and objections are very welcome.
Attachment #8631465 - Attachment is obsolete: true
Attachment #8641713 - Flags: review?(bugs)
"Debug B2G ICS Emulator" is very strange configuration with unstable behavior.
So all pointer events tests were disabled on that configuration until happy future.
Attachment #8641713 - Flags: review?(bugs) → review+
If there are no objections, I put checkin-needed flag. Let's try to push it again.
Keywords: checkin-needed
sorry had to back this out for test failures like https://treeherder.mozilla.org/logviewer.html#?job_id=12416096&repo=mozilla-inbound
Flags: needinfo?(alessarik)
Attached patch off_tests_ver17.diff (obsolete) — — Splinter Review 
+ Skiped tests on Android in Debug mode

Suggestions and comments and objections are very welcome.
Attachment #8641713 - Attachment is obsolete: true
Attachment #8699870 - Flags: review?(bugs)
Testing: https://treeherder.mozilla.org/#/jobs?repo=try&revision=159492c4f1c0
Looks like, there are no failing tests which are related with pointer events.
Comment on attachment 8699870 [details] [diff] [review]
off_tests_ver17.diff

Could you please provide interdiff.
(In reply to Olli Pettay [:smaug] from comment #99)
> Could you please provide interdiff.
Interdiff between off_tests_ver16 and off_tests_ver17 is very simple:
> [DEFAULT]
> -skip-if = debug && (toolkit == 'gonk') # Bug 1178701 - Issue on 'B2G ICS Emulator'
> +skip-if = debug && ((toolkit == 'gonk') || (os == 'android')) # Bug 1178701 - Issue on 'B2G ICS Emulator' and 'Android'
> support-files =
Attachment #8699870 - Flags: review?(bugs) → review+
Today it is good time to finish this long-running task :-)

If there are no objections, I put checkin-needed flag.
Let's try to push it again.
Keywords: checkin-needed
Attached patch off_tests_ver18.diff (obsolete) — — Splinter Review 
Add skip for one test.
>  [test_pointerevent_setpointercapture_inactive_button_mouse-manual.html]
> + skip-if = debug && (os == 'linux') && e10s
Attachment #8699870 - Attachment is obsolete: true
Attachment #8713517 - Flags: review?(bugs)
Comment on attachment 8713517 [details] [diff] [review]
off_tests_ver18.diff

r+, but please do still another try run using all the platforms. This has been backed out so many times
Attachment #8713517 - Flags: review?(bugs) → review+
Attached patch off_tests_ver19.diff (obsolete) — — Splinter Review 
Unfortunately, it is needed to skip several more tests.
Difference:
> -skip-if = debug && ((toolkit == 'gonk') || (os == 'android'))
> +skip-if = (toolkit == 'gonk') || (os == 'android')
Attachment #8713517 - Attachment is obsolete: true
Attachment #8715278 - Flags: review?(bugs)
Attachment #8715278 - Flags: review?(bugs) → review+
Do you have a link to the failures on Android?
(In reply to Olli Pettay [:smaug] (high review load) from comment #109)
> Do you have a link to the failures on Android?
Yes: https://treeherder.mozilla.org/#/jobs?repo=try&revision=1d4a20419521
Looks like Android-opt didn't run all of that test for a long time before latest time.
So, test-run-results looks better after latest changes.
https://treeherder.mozilla.org/#/jobs?repo=try&revision=d3c9a3312e60
So, if there are no objections, I put checkin-needed flag.
Let's try to push it again.
I hope it will be the latest iteration of this bug :-)
Keywords: checkin-needed
It seems rather unfortunate that we seem to have added a large number of testharness.js tests to a Mozilla-only location rather than testing/web-platform/tests where they would be upstreamed to wpt and shared with other vendors to improve interop. Is there any good reason these tests can't be moved to web-platform-tests?
If some subset of the tests are using mozilla-only APIs testing/web-platform/mozilla is still a better location because then we know that the tests are there when we are looking for things to convert later when the standard harness has more features.
Difference for test setpointercapture_inactive_button_mouse
> - skip-if = debug && (os == 'linux') && e10s # Bug 1180188 - Issue on Linux
> + skip-if = (os == 'linux') && e10s # Bug 1180188 - Issue on Linux

Testing: https://treeherder.mozilla.org/#/jobs?repo=try&revision=254573d6464e
Attachment #8715278 - Attachment is obsolete: true
Attachment #8723522 - Flags: review?(bugs)
Comment on attachment 8723522 [details] [diff] [review]
off_tests_ver20.diff

Please make sure this passes the test on all the platforms. Run the tests couple of times in tryserver before asking for check-in.
Attachment #8723522 - Flags: review?(bugs) → review+
(In reply to Olli Pettay [:smaug] (HIGH REVIEW LOAD) from comment #118)
> Please make sure this passes the test on all the platforms. Run the tests
> couple of times in tryserver before asking for check-in.
I checked all failures every time running on try-server. And all times I cannot see any issues.
Issues happens only after checkin procedure. Maybe try-server has different environment?
So, if there are no objections, I put checkin-needed flag. Let's try it again.
Keywords: checkin-needed
Did you trigger the tests several times on tryserver?
(In reply to Olli Pettay [:smaug] from comment #121)
> Did you trigger the tests several times on tryserver?
Yes. As I wrote earlier all latest try-server tests (with pointer events) showed good results.
One more time: https://treeherder.mozilla.org/#/jobs?repo=try&revision=fb03b0cd37c0
This doesn't apply to mozilla-central. Please rebase.
Flags: needinfo?(alessarik)
Keywords: checkin-needed
(In reply to Ryan VanderMeulen [:RyanVM] from comment #123)
> This doesn't apply to mozilla-central. Please rebase.
Checked on 2016-04-19 night. All patches were applied ok.
Right sequence: testing_feauters_ver4.diff - original_official_tests_ver2.diff - off_tests_ver20.diff
Flags: needinfo?(ryanvm)
Keywords: checkin-needed
thanks maksim for the checkin instructions, this helped landing this.
Flags: needinfo?(ryanvm)
Perfectly! Congratulations! 2-year-long bug.
I hope it will help to protect pointer-events from regressions!
Blocks: 1180188
No longer depends on: 1180188
Depends on: 1267310
Depends on: 1267311
Depends on: 1270903
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: