Re: [css-images] more midpoint issue

Please ignore this. I was confused how midpoints were treated.

On Sun, Sep 28, 2014 at 10:08 PM, Rik Cabanier <cabanier@gmail.com> wrote:

> While writing the validation checker for gradient midpoints, I ran into an
> issue: since positions can be defined as a mixture of length units and
> percentages, it's impossible to say that gradient midpoints are invalid.
>
> Maybe we can say that if a midpoint is in an invalid state, it should be
> ignored for rendering instead?
> Invalid would be if the midpoint is positioned:
> - at the beginning or the end of the grdient line
> - next to another midpoint
> - on top of a regular colorstop
>