Re: Home for viewport-related issues

> On Jan 22, 2019, at 7:44 AM, David Bokan <bokan@chromium.org> wrote:
> 
> Hi www-style,
> 
> I've collected some viewport-related interop research in an explainer I wrote a while back[1]. This doc tracked issues like how we decide what page scale to load a page at, how position: fixed elements and overflow: hidden work in the presence of pinch-zoom, etc.
> 
> bballo@mozilla.com <mailto:bballo@mozilla.com> has been working on squashing some of these interop issues and so far we've been using my GitHub as a place to collect issues and discussion. I think it'd be beneficial to move the discussion into a more official+public place, perhaps a WICG GitHub repo. I was wondering if anyone had thoughts on where this might belong or if there's an existing spec/repo viewport work might fit well into.
> 
> I think these kind of issues are closely aligned with the device-adaptation spec [2]. There's also the question of the visualViewport API spec [3]. This is currently a WICG draft but I believe others are now implementing it so we could probably start moving it out of "incubation". That's also closely related to these discussions.
> 
> WDYT?
> 
> Thanks,
> David
> 
> [1] https://github.com/bokand/bokand.github.io/blob/master/web_viewports_explainer.md <https://github.com/bokand/bokand.github.io/blob/master/web_viewports_explainer.md>
> [2] https://www.w3.org/TR/css-device-adapt-1/ <https://www.w3.org/TR/css-device-adapt-1/>
> [3] https://wicg.github.io/visual-viewport/index.html <https://wicg.github.io/visual-viewport/index.html>
It would be great to have a spec for viewports (basically covering the stuff the explainer[1] talks about). That spec also needs to define what coordinate systems are used for ALL THE THINGS (events, getBoundingClientRect etc etc), as well as talking about what triggers the firing of scroll events. This spec feels related to CSS OM View, so I'm not sure if it would go into a level 2 of that spec, be its own spec.

WebKIt has recently enabled the Visual Viewport API by default, so we support that proposal.

I don't think we're as keen on the CSS @viewport rule.

Simon