Open AndreiGorea opened 1 week ago
If the screenshot is very large you may need to use a smaller ratio.
Hi, thanks for a quick reply @garris but the problem is not with the ratio, because I didn't make any changes to the tests that a ran before.
I will show you an example:
This is my scenario array from backstop.json
"scenarios": [
{"misMatchThreshold": 0.1,"requireSameDimensions": true,"postInteractionWait": 1000,"label": "Dashboard Page","url": "https://dxnet-test-de.crossinx.com/crossnet/en_DE_xchannel/layouts/finx/common/welcome","waitForSelector": "fa fa-power-off, .fa-power-off","removeSelector": "#selectvalue ","hideSelectors": [".d-flex > small:nth-child(1) > strong:nth-child(1), [class='horizontalbarchart-container']"],"delay": 1000}
],
I have this reference photo:
And I made a test and I have this test photo:
Obviously there is a difference, in test photo we don't have data in the fields
But backstopjs in html report tells that test is passed and there are no changes!
Maybe there were made some changes to backstop? It's a strange and new behaviour.
There were somewhat recent changes that might have interacted. https://github.com/garris/BackstopJS/pull/1558
Maybe that?
I've noticed for some time that backstop doesn't compare pictures well, for example if we have a field with some data in it like the date and we do a test and remove this date with removeSelector Backstop won't see this change. Before about 2 weeks ago Backstop was noticing this change.
What would be the problem?
I use misMatchThreshold": 0.1