Regression Issue

[WilliamParks]

Hello!
I've been playing around with your tests via uniquemachine.org, and noticed that recently something seems to have broken.

Previously, I had very good stability among browsers for the graphics specific tests. Now, a number of tests are different between browsers. The attached image has Safari Version 10.1 (12603.1.30.0.34) on the left, Firefox 53.0.2 in the middle, and Chrome 58.0.3029.110 on the right. The check marked tests on Safari and Firefox are the ones that differ from Chrome. All of these browsers are running on Mac OS 10.12.4, on a macbook pro with an AMD Radeon R9 M370X graphics card, if that helps reproduce the issue.

Since this worked previously, I figured you'd want to know!

Also, I'm curious as to how you're doing the hashing, if you're doing some sort of feature extraction on the images, or simply hashing the image as a whole. I did a quick look through the code in the repo, but missed exactly how you implemented it. Thanks!

[shockwavemoto]

Previously, I had very good stability among browsers for the graphics specific tests. Now, a number of tests are different between browsers. The attached image has Safari Version 10.1 (12603.1.30.0.34) on the left, Firefox 53.0.2 in the middle, and Chrome 58.0.3029.110 on the right. The check marked tests on Safari and Firefox are the ones that differ from Chrome. All of these browsers are running on Mac OS 10.12.4, on a macbook pro with an AMD Radeon R9 M370X graphics card, if that helps reproduce the issue.

Tried to reproduce it but sorry couldn't
Any more information would be helpful to sort out this issue

[WilliamParks]

@shockwavemoto I'm not exactly sure what else I can add, other than my system setup and browser version numbers which were in the original comment. Anything in particular?

[nebi-frame]

Is this issue unsolved yet?