Re: [css-round-disiplay][mediaqueries] Orientation media feature consideration

Hi Hyojin Song,

Thank you for the feedback on media queries.

> On 04 Jun 2015, at 10:53, Hyojin Song <hyojin22.song@lge.com> wrote:
> 
> In the CSS Round Display spec, 'device-radius' property has been described as the Media Queries extension. I think this property is proper level between expressiveness and simplicity to check what the display shape is. We can extend the property to express another shapes like star, triangle, diamond.

I agree, trying to support detection of arbitrary shapes at this point would be going beyond the situations we understand well, and we would risk over-engineering the solution.

Maybe even if we do not define support for other shapes, we should consider the syntax carefully to make sure we can easily extend that support later. I do not have any particular proposal in that regard, though.

However, I think we do need to be careful in how we define the device-radius media feature so that it is not ambiguous what happens with non-square and non-circular shapes. See points 7, 8, 9, 10 and 11 in this mail:
http://www.w3.org/mid/24C96E06-013A-40C0-87A4-6C76F23AC384@rivoal.net

> BTW, I noticed the orientation property as one of media features in CSS Media Queries. Its values can be 'portrait' or 'landscape'. 'orientation: portrait' has a true value, when the value of the 'height' media feature is greater than or equal to the value of the 'width' media feature.
>  
> In the case of a regular round display, the value of the 'width' is equal to the value of the 'height', but the situation that should detect the device's orientation can happen if I want to show my watch's screen to a friend. The orientation of the watch might be changed when I show the watch to someone from my view or when I stretch out my hand straight from bended state.

Could you clarify a little? Is the behavior of smart-watch similar to smart-phones, rotating the viewport when a physical orientation change is detected? If yes, then I don't think authors really need a media query to react to the change.

If the viewport itself isn't rotated, but you would like a media query to detect this and opt into some changes despite the viewport staying the same, then I agree that the 'orientation' media feature will not work, as it is really linked to the aspect ratio of the viewport, not to the physical position of the device.
 
> In this situation, I'm wondering the definition of orientation media feature's values in CSS Media Queries should be changed, or new property can be defined, or nothing to do.

I guess we could think of a new media query that changes whether "up" in the document matches "up" in the physical world, or whether the device is tilted. At the same time, I am not exactly sure how it would be used in practice, so I have a bit of a hard time considering whether it should be keyword based (straight | tilted-left | tilted-right | upside-down) or be a numeric media features indicating an angle, and used with range syntax. Also, what about the fact that physical orientation is in 3D?

I think having a few example showing how this would be used, to get a better sense of what an author would do with it, would help.

Best regards,
 - Florian

Received on Thursday, 4 June 2015 12:23:09 UTC