Re: [css3-images] Reintroduce object-fit: none

On Thu, Jan 27, 2011 at 2:35 PM, fantasai <fantasai.lists@inkedblade.net> wrote:
> On 01/27/2011 09:18 AM, Tab Atkins Jr. wrote:
>> Ok, so I'm hearing some good justifications for both 'cover' and
>> 'none' (and I think 'scale-down' has similar nice justifications).
>
> I'm still unsure about 'none'. The use case Leif has given is
> panning the image, and in that case you'd probably want more than
> just the ability to pan at 100% zoom level. If that's the use case
> we're addressing, we should add percentages or something.
>
>> I'm also hearing support from Safari and Opera about making replaced
>> elements *always* hide their overflow.
>>
>> Is that acceptable to you, dbaron?
>
> I agree that the contents of replaced elements should always be
> clipped to the content edge. (I thought I'd made those changes
> already. Must've been to css3-background.)

I'm trying to wrap up this issue.  There are a few aspects of the
property that need adjustment:

1) replaced elements should *always* clip their content.  This is a
non-controversial change; I just need to adjust some wording and
change the example image.

2) Add the 'scale-down' value.  The use-case for this (image viewer
that wants to scale the image to the viewing area if it's too large,
but display it at the natural size if it's small enough to fit in the
viewing area) seems appropriate, and the implementation for this value
seems easy.

3) Add the 'none' value.  The stronger use-case here seems to be
panning a "window" over a portion of a larger image, a la Google Maps.
 A somewhat weaker imo use-case is vertically centering an image; I
think this should be addressed by a more general centering mechanism
like Flexbox.

4) (optional) Add percentages as a possible value, similar to
background-size, so you can arbitrarily size the content.  'fill'
would be equivalent to '100% 100%'.

Any opinions yay/nay on these?  The different proposals are presented
in descending order of how much I want to add them, so I'll do #1
unless there's a very strong reason against, I'll do #2 unless there's
a reasonably strong reason against, I'll do #3 if there's a reasonable
reason *for*, and I'll do #4 if there's a strong use-case presented
for it.

~TJ

Received on Tuesday, 24 May 2011 00:07:27 UTC