Third part of porting Chrome Ash to Windows 7

Third part of porting Chrome Ash to Windows 7

This is the Chrome side that
1- removes windows 8 specific checks
2- disconnects the fake window_tree_host_win
3- connects the remote_window_tree_host_win

With this change you can now see the ash desktop but not interact with
it since input events have not been enabled.

Here is the second part
https://codereview.chromium.org/211863003/

First part with rationale:
https://codereview.chromium.org/199843004/

BUG=356475
TEST=none

[cpu_(ooo_6.6-7.5), 2014-04-11 02:14:06]

.

[cpu_(ooo_6.6-7.5), 2014-04-11 02:15:58]

please review:

sky : chrome stuff
ananta :  all
oshima : ash

[ananta, 2014-04-11 02:25:23]

https://codereview.chromium.org/227573007/diff/40001/chrome/browser/metro_uti...
File chrome/browser/metro_utils/metro_chrome_win.cc (right):

https://codereview.chromium.org/227573007/diff/40001/chrome/browser/metro_uti...
chrome/browser/metro_utils/metro_chrome_win.cc:23: if (base::win::GetVersion() <
base::win::VERSION_WIN8)
Would it make sense to launch the ASH version on Win 7 here?

https://codereview.chromium.org/227573007/diff/40001/win8/metro_driver/displa...
File win8/metro_driver/display_properties.cc (right):

https://codereview.chromium.org/227573007/diff/40001/win8/metro_driver/displa...
win8/metro_driver/display_properties.cc:21: if (base::win::GetVersion() <
base::win::VERSION_WIN8)
Please put a todo here to return the correct dpi scale. A bug would be great.

https://codereview.chromium.org/227573007/diff/40001/win8/metro_driver/metro_...
File win8/metro_driver/metro_driver_win7.cc (right):

https://codereview.chromium.org/227573007/diff/40001/win8/metro_driver/metro_...
win8/metro_driver/metro_driver_win7.cc:18: switch (message) {
Please add some comments here about why we need the g_window_count variable.

[oshima, 2014-04-11 03:38:31]

Hi, I haven't looked into the code yet but this will conflict with my change
(which I'm trying to land)

https://codereview.chromium.org/201573015/

Notable change is that the window tree host factory is gone.

[sky, 2014-04-14 13:55:17]

https://codereview.chromium.org/227573007/diff/40001/chrome/browser/browser_p...
File chrome/browser/browser_process_platform_part_aurawin.cc (right):

https://codereview.chromium.org/227573007/diff/40001/chrome/browser/browser_p...
chrome/browser/browser_process_platform_part_aurawin.cc:53:
CHECK(metro_viewer_process_host_->LaunchViewerAndWaitForConnection(
We're not going to use the metro viewer code for this, are we?

[cpu_(ooo_6.6-7.5), 2014-04-14 20:02:06]

https://codereview.chromium.org/227573007/diff/40001/chrome/browser/browser_p...
File chrome/browser/browser_process_platform_part_aurawin.cc (right):

https://codereview.chromium.org/227573007/diff/40001/chrome/browser/browser_p...
chrome/browser/browser_process_platform_part_aurawin.cc:53:
CHECK(metro_viewer_process_host_->LaunchViewerAndWaitForConnection(
On 2014/04/14 13:55:17, sky wrote:
> We're not going to use the metro viewer code for this, are we?

The current plan includes the viewer process in metro_driver.dll. Just like in
win8.

We could use a different thread on the browser but I felt that it could beget
the issues we had with the previous code in which activation & focus bugs are
masked because of the differences on message pumping.

I am going to be OOO for a few days and this CL has been abandoned and ananta is
going to continue it on a different CL.

[cpu_(ooo_6.6-7.5), 2014-04-14 20:04:16]

===============================
 this CL has been abandoned
===============================

Please the successor here:

https://codereview.chromium.org/235933004/