# Re: 24. Gradient & stretched backgrounds

From: Rasmus Kaj <rasmus@kaj.a.se>
Date: Fri, 15 Jan 1999 09:39:42 -0500 (EST)
Message-Id: <19990115152558F.kaj@interbizz.se>

```>>>>> "IH" == Ian Hickson <py8ieh@bath.ac.uk> writes:

IH> On Thu, 14 Jan 1999, David Perrell wrote:
>> How about two more: 'stretch-x' and 'stretch-y'?
>> Stretch-x would size-to-fit in the x direction and repeat in the y,
>> stretch-y would size-to-fit in the y and repeat in the x. This would
IH> Good idea.

Agreed.

IH> Anyway. The full description becomes:

IH>  'background-repeat'
IH>     Value:       repeat | repeat-x | repeat-y | no-repeat |
IH>                [INS:] fit | fit-x | fit-y | [:INS] inherit

(snip)

IH>   fit-x
IH>          Analogous to 'fit', but only scaled to fit in the x direction.
IH>          In the y-direction, the image is repeated, the height of each
IH>          step being the intrinsic height of the image (as in 'repeat-y').

I'd like to suggest a further change in this, so that in the
y-direction the image is neither fitted or repeated, unless repeat-y
is given. The same change applies to fit-y (and repeat-[xy])

This means that the value definition would have to be altered to
accept two values. Then repeat, fit and no-repeat would be shorthand
values for repeat-x repeat-y, fit-x fit-y and no-repeat-x no-repeat-y
respectively.

The benefit of this is
a) that the author gets greater freedom
b) that the implementator have independent values for the x and y
axes, which I think simplifies the implementation (yes, I am
writing a CSS implementation, so I think I can have opinions of
what implementators (or at least one implementator) thinks).

// Rasmus Kaj

--
rasmus@kaj.a.se ----------------- Rasmus Kaj - http://www.e.kth.se/~kaj/
\                          Microsoft: How long do U want to wait today?