1866784 - moveTo when using geckodriver doesn't work if called on 0, 0

Status: RESOLVED DUPLICATE of bug 1850148

(Testing :: geckodriver, defect)

Description

[nikhakerlab]

Attachment: hover-test.html

Steps to reproduce:

1. Create session in geckodriver:

curl --location 'http://localhost:4444/session' \
--data '{"capabilities":{"alwaysMatch":{"browserName":"firefox"},"firstMatch":[{}]},"desiredCapabilities":{"browserName":"firefox"}}'

2. Open the html page attached to this issue from step 1
3. Now, try manually hovering over different cells. Then try to set pointer to 0,0:

curl --location 'http://localhost:4444/session/c2d9e552-8888-4589-a56a-7e9c6c54bc9a/actions' \
--data '{
    "actions": [
        {
            "id": "action2",
            "type": "pointer",
            "parameters": {
                "pointerType": "mouse"
            },
            "actions": [
                {
                    "type": "pointerMove",
                    "button": 0,
                    "width": 0,
                    "height": 0,
                    "pressure": 0,
                    "tangentialPressure": 0,
                    "tiltX": 0,
                    "tiltY": 0,
                    "twist": 0,
                    "altitudeAngle": 0,
                    "azimuthAngle": 0,
                    "x": 0,
                    "y": 0,
                    "duration": 100,
                    "origin": "viewport"
                }
            ]
        }
    ]
}'

Try different x/y coords and mix requests with manual hovering

Actual results:

Sometimes the request to move pointer to 0,0 is successful, sometimes nothing happens and pointer stays where it was. The behaviour is inconsistent.

Expected results:

The pointer consistently moves to 0,0 after each curl request. It does so in chrome.

Comment 1

[BugBot [:suhaib / :marco/ :calixte]]

The Bugbug bot thinks this bug should belong to the 'Core::Layout' component, and is moving the bug to that component. Please correct in case you think the bot is wrong.

Comment 2

[Henrik Skupin [:whimboo][⌚️UTC+2]]

This is most likely a dupe of bug 1850148.

nikhakerlab would you mind attaching a trace log from geckodriver that show-cases that? I assume that in the case when it does not work you freshly created the action chain so that the mouse pointer internally is already at the position (0, 0). As such no pointer move will be executed. Chrome is currently doing that wrong and will also change its behavior.

Comment 3

[Henrik Skupin [:whimboo][⌚️UTC+2]]

No further reply from reporter. As such I'm going to assume it's indeed a dupe of bug 1850148.