applyMask() fix in Gd/Image

[ninze]

Created applyMask() fix for #460

[ausi]

I’m not sure if this fix is correct. I think applyMask() is meant to be used with a grayscale image.

Did you try to use ->applyMask($mask->mask()); as described in #460 (comment) ?

[ninze]

@ausi yes I did without success. The fix makes imagine work consistently with both GD and Imagick drivers. We are using this fix in production since April without any issues.

[mlocati]

Is it possible to have a test case that covers what this PR is fixing?

[ninze]

It probably is, will look into it.

[ninze]

Done. I don't have gmagick installed, but GD and Imagick produce identical results after the fix.

[jakubjo]

@ausi yes I did without success. The fix makes imagine work consistently with both GD and Imagick drivers. We are using this fix in production since April without any issues.

Maybe it's about time someone merges this?

[ninze]

After more than 2 years since the pull request was created, I also think someone should either merge or close this.

[ausi]

Tested this pull request locally and all worked correctly.

In the old code ->dissolve(…) could have never ended up with a negative value and therefore could never have masked anything.

[mlocati]

Thanks for reviewing this PR, @ausi !

[chekalsky]

@mlocati thank you! is it possible to release a new version of Imagine wit this fix any time soon?

[mlocati]

@chekalsky I'd publish a new version after tests have been fixed (I'd move them to GitHub Actions)