RE: [css3-images] Resolving on gradient issues

Alan Gresley [mailto:alan@css-class.com]:
> On 5/08/2011 3:45 AM, Brian Manthos wrote:
> > Brad Kemper:
> >> That is a slanted way of characterizing it. The way I see it,
> >> 'repeating-linear-gradient' is a way of simulating
> >> 'background-repeat' on a possibly rotated canvas. background-repeat
> >> can already create a repeating gradient (it just doesn't look good
> >> when the gradient path is angled).
> >
> > Background-repeat has a specific meaning and capabilities.
> 
> 
> Brad is talking about 'repeating-linear-gradient' looking odd when the
> background tile is not rotated. To achieved that, you need to rotate
> the
> background tile but then that bring up the new issue of how keywords
> work for rotated background tiles along with how they are used (will be
> used) with keywords with gradients.

You're reading more into it than what he said.  That may be what he meant, but it's not what he said.

I don't see repeating-linear-gradient as a way of simulating background-repeat.  It's a variation on linear-gradient with its own behavior.  The fact that some flavors of linear-gradient + background-repeat combinations look similar to repeating-linear-gradient is perhaps interesting, but doesn't suggest or imply that they are designed to suit the same needs.

Further, Brad wrote:
> And there already is very little need/demand for 'repeating-radial-gradient'.

I believe quite the opposite.  The current repeating-radial-gradient provides capabilities that can't be simulated with background-repeat or any other properties without significant shenanigans.

Lastly, let's imagine you did "something interesting" with background-repeat, and tossed out repeating-linear-gradient and repeating-radial-gradient.  Well, you've just removed the capability of repeating gradients from list-style-image and generated content.  For no good reason.


I think the current repeating-linear-gradient and repeating-radial-gradient provide capabilities not found in other places, and aren't easily replaced with other syntax.

I recommend we stop fiddling here -- leave these well-defined and well-understood repeating gradients as is.