Chris Blume

I think I agree that private color types should still be a thing. The fix would be just adding it to that list. I don't think this allows private bit...

The original suggestion mentioned a bool. This would be better served as a list of enumumerated scaling algorithms, in order of preference. Or as a single preferred behavior as mentioned...

Agreed. The idea makes sense. But I think we should have some pressure from industries before we standardize it. In fact, I think that is a piece of W3C's requirements.

Thank you for the work and informing us. > sensitive to any relative movement of the iDOT/IDAT chunks This one seems like a problem for existing tooling. I think an...

I would revise the "or have any space between them" part. I could imagine the first thread decodes until it gets to its end, which is a reset marker (or...

Indeed, the `pHYs` chunk and `oFFs` are not suitable for geospacial information. This is related to #15. In fact, they're two separate pieces of the original post. The original post...

Just because some decoders decide to reject the PNG doesn't mean the spec must require them to reject the PNG. It's the other way around. That said, I do strongly...

Oh, right. I thought you were saying decoders couldn't ignore them. You are right that encoders must write them. I don't see a way we could remove that requirement without...

GitHub does allow renaming repos, I think. I'm okay either way. Someone who knows what W3C is will immediately understand "/png" is the spec. So "/PNG-spec" is redundant to them....

Sounds good. I'll look into renaming it w3c/png.