Document DImage transformation

[robinp]

In https://archives.haskell.org/projects.haskell.org/diagrams/haddock/Diagrams-TwoD-Image.html, I wanted to prescale the DImage before converting to Diagram with image, so I sticked in the transform in the constructor (it was a simple scalingX <> scalingY).

But that apparently did something strange. I got the desired effect by applying the scaling to the Diagram itself.

Would be nice to document what the Transformation piece of the constructor is for.

[bergey]

There is a comment in the code from 2015 "should we use the transform here?". On a quick reading, I think yes, we should. It seems now the `Transformable` instance does the obvious thing, but then we throw out the `Transformation` when creating the `Diagram`. This doesn't seem right.

…

On 2017-12-24 at 12:28, Robin Palotai ***@***.***> wrote: In https://archives.haskell.org/projects.haskell.org/diagrams/haddock/Diagrams-TwoD-Image.html, I wanted to prescale the DImage before converting to Diagram with image, so I sticked in the transform in the constructor (it was a simple scalingX <> scalingY). But that apparently did something strange. I got the desired effect by applying the scaling to the Diagram itself. Would be nice to document what the Transformation piece of the constructor is for. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.*

[byorgey]

@bergey I think I agree with you. Does #307 look reasonable to you?