Open ld-kerley opened 3 weeks ago
I think no issues with the suggested change.
Thanks Masuo, for the Hextile PR I think we can roll with the refactor - I just wanted to log the issue here so we can remember to perhaps investigate a more robust solution moving forwards, or at least have the issue here for future people to discover if they need the same work around.
While looking at getting Masuo's Hextiling PR working #2094 for MSL, I discovered that GLSL and MSL differ in how structs are initialized.
I think its an easy refactor for the Hextiling PR, but noting this difference here, so we can investigate a more robust solution in the future.
Below is the problematic code from Masuo's PR.
The workaround is to create the object and then initialize each member separately.
But I can imagine cases where this might not be desirable.