[M58] exo: Confine windows to primary display

[M58] exo: Confine windows to primary display

Multi-display support in ARC was punted to the next milestone. This CL
temporarily prevents windows from being dragged to secondary displays
by using display rather than screen coordinates for window bounds.

BUG=714832
TEST=ARC apps cannot be dragged to secondary displays.

[Dominik Laskowski, 2017-04-27 14:40:23]

domlaskowski@chromium.org changed reviewers:
+ oshima@chromium.org, reveman@chromium.org

[Dominik Laskowski, 2017-04-27 14:40:24]

PTAL.

[reveman, 2017-04-27 16:08:08]

Any reason we shouldn't land this in ToT as well? This only affects arc clients,
correct?

[Dominik Laskowski, 2017-04-27 17:26:44]

On 2017/04/27 16:08:08, reveman wrote:
> Any reason we shouldn't land this in ToT as well? This only affects arc
clients,
> correct?

We'd need a special case for ARC clients without multi-display support. Given
that client-side support is landing in M59, I don't think that's worthwhile.

[Dominik Laskowski, 2017-04-27 17:38:08]

Note that clients prior to M58 would coincidentally revert the drag, but this CL
drops the window instead.

[reveman, 2017-04-27 18:13:54]

On 2017/04/27 at 17:38:08, domlaskowski wrote:
> Note that clients prior to M58 would coincidentally revert the drag, but this
CL drops the window instead.

OK. I would still prefer if we landed this in ToT first and then merged it back
to both M59 and M58. We can revert the change in ToT after merging it back.
Having everything go through ToT is nice as it allows a bisect to point to this
change if it's causing an unexpected issue.

[Dominik Laskowski, 2017-04-27 18:32:02]

On 2017/04/27 18:13:54, reveman wrote:
> OK. I would still prefer if we landed this in ToT first and then merged it
back
> to both M59 and M58. We can revert the change in ToT after merging it back.
> Having everything go through ToT is nice as it allows a bisect to point to
this
> change if it's causing an unexpected issue.

It doesn't merge cleanly to ToT and M59, and seems innocuous enough to avoid
that churn.

[reveman, 2017-04-28 19:50:37]

On 2017/04/27 at 18:32:02, domlaskowski wrote:
> On 2017/04/27 18:13:54, reveman wrote:
> > OK. I would still prefer if we landed this in ToT first and then merged it
back
> > to both M59 and M58. We can revert the change in ToT after merging it back.
> > Having everything go through ToT is nice as it allows a bisect to point to
this
> > change if it's causing an unexpected issue.
> 
> It doesn't merge cleanly to ToT and M59, and seems innocuous enough to avoid
that churn.

Fair enough. Is it possible to land a separate change in ToT to make it
consistent? I'm a bit scared of using a behavior on a branch that we never
tested in ToT.

[Dominik Laskowski, 2017-04-28 21:35:24]

On 2017/04/28 19:50:37, reveman wrote:
> Fair enough. Is it possible to land a separate change in ToT to make it
> consistent? I'm a bit scared of using a behavior on a branch that we never
> tested in ToT.

We don't want this in ToT/M59, because it would break multi-display support.

There is little to no risk in landing this, i.e. the code is equivalent unless
there are multiple displays.

[reveman, 2017-04-28 21:38:14]

lgtm

[oshima, 2017-05-01 21:58:12]

lgtm

[Dominik Laskowski, 2017-05-04 21:31:12]

Description was changed from

==========
[M58] exo: Confine windows to primary display

Multi-display support was punted to M59. This CL prevents windows from
being dragged to secondary displays by using display rather than screen
coordinates for window bounds.

BUG=714832
TEST=Manual on Caroline M58.
==========

to

==========
[M58] exo: Confine windows to primary display

Multi-display support in ARC was punted to the next milestone. This CL
temporarily prevents windows from being dragged to secondary displays
by using display rather than screen coordinates for window bounds.

BUG=714832
TEST=ARC apps cannot be dragged to secondary displays.
==========

[Dominik Laskowski, 2017-05-04 22:14:49]

Closing. Will cherry-pick from ToT instead.