| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Mouse locking is used for e.g. by WebGL and is implemented in the browser by entering a state in the web view
where in we ensure that the mouse stays within the boundaries of the view and we send only mouse increment/decrement
information to the renderer.
We had support for this in Chrome AURA which did not work correctly due to a number of bugs.
1. We need to implement the MoveCursorTo override in the DesktopRootWindowHostWin, DesktopRootWindowHostX11 and
RemoteRootWindowHostWin classes which implement the aura RootWindowHost interface. In desktop Chrome windows we
invoke the SetCursorPos API to ensure that the cursor is set to the position specified. In desktop Chrome linux we
invoke the XWarpPointer API to achieve this. In ASH chrome on Windows we cannot call SetCursorPos in the
RemoteRootWindowHostWin class's implementation of the MoveCursorTo function.
As mouse input is received by the viewer process and sent to the browser, we have scenarios wherein we receive stale
mouse input from the viewer after invoking the SetCursorPos API in the browser. To workaround this we call SetCursorPos
API in the viewer process from an IPC MetroViewerHostMsg_SetCursorPos which is sent from the RemoteRootWindowHostWin::MoveCursorTo
function. We send out an ACK from the viewer process when the call finishes and in the meantime we ignore any mouse
messages coming from the viewer while we wait for the ACK.
If we don't do this then we have issues with the image locking up on screen in ASH. In the viewer process we also
send out a fake mousemove IPC after the SetCursorPos API as we don't receive the corresponding real mouse move for
some reason. This is needed as RVHAura class sets a flag which is reset when it receives the mousemove corresponding to
the cursor pos.
2. The RenderWidgetHostViewAura::ShouldMoveToCenter function which checks if the mouse cursor has moved beyond the window
bounds needs to convert the window rect to screen.
3. The MakeWebMouseEvent function in the web_input_Event_aura.cc file should use not set the globalX and globalY members in
the WebMouseEvent structure from the root coordinates if we have a native mouse event for Windows. This breaks the
lock mouse coordinate calculations.
I also fixed the implementation of the RootWindowHost::QueryMouseLocation function in the RemoteRootWindowHostWin class and
the DesktopRootWindowHostWin class. We can now safely enable the AshNativeCursorManagerTest.DisabledQueryMouseLocation test
with some windows specific fixes.
BUG=222497
R=cpu@chromium.org, palmer@chromium.org, sky@chromium.org
Review URL: https://codereview.chromium.org/14574010
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@198582 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=236029
R=ananta@chromium.org
Review URL: https://chromiumcodereview.appspot.com/14520027
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@198323 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Chrome AURA when a DesktopRootWindowHostWin/
ASH root window is activated.
To achieve this we register a handler for the Activate event on the ASH core Window in Windows 8 and send an IPC
MetroViewerHostMsg_WindowActivated from the metro viewer process along with a bool which indcates whether we
gained or lost activation. In the RemoteRootWindowHostWin class we inform the RootWindow whether it gained or lost
focus based on the active flag passed in the IPC.
BUG=235128
R=cpu,sky
TBR=palmer
Review URL: https://codereview.chromium.org/14070017
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@196835 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Windows 8.
The current behavior is that the select folder picker displays in desktop mode even when
when performed in ASH.
Fixes are as below:-
1. In the Chrome browser added plumbing in the SelectFileDialogImpl class to forward this
operation of to the metro viewer process via the newly added function HandedSelectFolder
which lives in the aura namespace in the remote_window_host_win.cc file.
2. In the metro viewer code added a new class FolderPickerSession which provides the functionality
for displaying the metro select folder picker and returning the results from the same.
3. The following IPC messages provide the necessary functionality to funnel the request
and the result between the browser and the metro viewer process.
MetroViewerHostMsg_DisplaySelectFolder
MetroViewerHostMsg_SelectFolderDone
I also added code to save away the metro root window in the RemoteRootWindowHostWin class
when we receive the MetroViewerHostMsg_SetTargetSurface IPC from the viewer process. We
return this window in the RemoteRootWindowHostWin::GetAcceleratedWidget function. This
change although not needed for this CL seems correct.
BUG=230087
R=cpu,sky
Review URL: https://codereview.chromium.org/14282002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@195224 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The XP bots are still red, and it's still whining about URL.DLL
> Lets try this again.
>
> Revert 193983 "Revert 193968 "Roll the DEPS for google-url.""
>
> > Revert 193968 "Roll the DEPS for google-url."
> >
> > > Roll the DEPS for google-url.
> > >
> > > And update all the references from build/temp_gyp/googleurl.gyp to url/url.gyp.
> > >
> > > This also changes googleurl.gyp to reference the files under url/, so that we
> > > don't break the Blink tree.
> > >
> > > BUG=229660
> > > R=brettw@chromium.org
> > > TBR=brettw@chromium.org
> > >
> > > Review URL: https://chromiumcodereview.appspot.com/14089011
> >
> > TBR=tfarina@chromium.org
> > Review URL: https://codereview.chromium.org/14028012
>
> TBR=michaeln@google.com
> Review URL: https://codereview.chromium.org/14109014
TBR=michaeln@google.com
Review URL: https://codereview.chromium.org/14263002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@194151 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Revert 193983 "Revert 193968 "Roll the DEPS for google-url.""
> Revert 193968 "Roll the DEPS for google-url."
>
> > Roll the DEPS for google-url.
> >
> > And update all the references from build/temp_gyp/googleurl.gyp to url/url.gyp.
> >
> > This also changes googleurl.gyp to reference the files under url/, so that we
> > don't break the Blink tree.
> >
> > BUG=229660
> > R=brettw@chromium.org
> > TBR=brettw@chromium.org
> >
> > Review URL: https://chromiumcodereview.appspot.com/14089011
>
> TBR=tfarina@chromium.org
> Review URL: https://codereview.chromium.org/14028012
TBR=michaeln@google.com
Review URL: https://codereview.chromium.org/14109014
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@193992 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
> Roll the DEPS for google-url.
>
> And update all the references from build/temp_gyp/googleurl.gyp to url/url.gyp.
>
> This also changes googleurl.gyp to reference the files under url/, so that we
> don't break the Blink tree.
>
> BUG=229660
> R=brettw@chromium.org
> TBR=brettw@chromium.org
>
> Review URL: https://chromiumcodereview.appspot.com/14089011
TBR=tfarina@chromium.org
Review URL: https://codereview.chromium.org/14028012
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@193983 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
And update all the references from build/temp_gyp/googleurl.gyp to url/url.gyp.
This also changes googleurl.gyp to reference the files under url/, so that we
don't break the Blink tree.
BUG=229660
R=brettw@chromium.org
TBR=brettw@chromium.org
Review URL: https://chromiumcodereview.appspot.com/14089011
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@193968 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
FindCoreWindow.
BUG=15171
TEST=NONE
Review URL: https://chromiumcodereview.appspot.com/13933009
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@193719 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The main thing is the discovery of a clean way to retrieve the
corewindow hwnd based on ICoreWindowInterop, that removes some iffy
heuristic based on window name before.
From that we realize that we don't need some globals, namely the window
handle the thread id that owns the window.
BUG=151718
TEST=ash still works.
Review URL: https://codereview.chromium.org/13849005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@193120 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The touch events coming in from desktop chrome AURA and ASH have hardcoded touch ids
which breaks the gesture recognizer as it is unable to figure out that multiple touch points
are involved.
Fix on desktop Chrome AURA is to use the OS generated touch id mapped to the gesture recognizer range.
For ASH Windows 8 we pass in the mapped touch ids from the metro driver in the touch IPC messages.
Fixes bug https://code.google.com/p/chromium/issues/detail?id=179977
BUG=179977
R=cpu,sky
Review URL: https://codereview.chromium.org/13003004
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@191015 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The 2012 compiler can handle the "enum class" so don't force patching if compiling with 2012. Cast a limited selection
logging of enum class types to avoid needing to add overloads to base/logging for those types (they don't decay to int).
R=robertshield@chromium.org
TEST=ninja -C out\Debug metro_driver with 2010/2012 and see if check_sdk_patch is generated/not.
Review URL: https://chromiumcodereview.appspot.com/12676005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@190073 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
keyboard.
When a menu is displayed we enter a nested message loop in the Chrome Browser. This loop
expects to receive native events for WM_KEYDOWN/WM_KEYUP/WM_CHAR, etc. In desktop AURA this
works well because the native events come in through the standard OS mechanism.
In ASH we send over fabricated keyboard events from the viewer process to the browser where
these are dispatched to the root window. This causes the secondary loop to never receive these
events.
Fix for now is to check if we are in a nested loop in the RemoteRootWindowHostWin IPC handlers
and post the corresponding native event back to the queue.
The other change in this CL is to add an event handler to listen for accelerator events in the viewer
process. This is necessary to receive keystrokes like Alt, etc. We send over the same keydown/keyup/character
IPC's when we receive this event.
BUG=180738
R=cpu
TBR=ben
TEST=Will look into an ash based test for this in a subsequent CL.
Review URL: https://codereview.chromium.org/12558008
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@186797 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
| |
TBR=sky
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@184344 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
| |
Review URL: https://codereview.chromium.org/12217101
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@181640 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
This updates headers that forward-declare it and a few random places to use the namespace explicitly. There us a using declaration in file_path.h that makes the rest compile, which we can do in future passes.
Review URL: https://codereview.chromium.org/12163003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@180245 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Chrome ASH.
Added the following functions to the Aura namespace.
1. HandleOpenFile
2. HandleOpenMultipleFiles
3. HandleSaveFile
These functions are invoked from the windows implementation of SelectFileDialogImpl::SelectFileImpl
for AURA. Internally these functions forward off to the corresponding implementations in the singleton
instance of the RemoteRootWindowHostWin class which is instantiated for Chrome ASH on Windows.
These functions take in a callback object which is invoked when the file picker operations complete.
The SelectFileDialogImpl::SelectFileImpl function uses the newly added ShellDialogsDelegate interface to
determine if the owning native window is in Windows 8 metro. This interface is implemented by a global
object ShellDialogsDelegateWin which lives in the chrome_browser_main_extra_parts_ash.cc file. It uses the
IsNativeViewInAsh helper function to determine this.
Added the following IPC's which are sent from the browser to the metro viewer process.
1. MetroViewerHostMsg_DisplayFileOpen
2. MetroViewerHostMsg_DisplayFileSaveAs
The following IPC's are sent from the metro viewer to the browser process and inform the browser about the
status of the file open/file save operations along with the file names.
1. MetroViewerHostMsg_FileOpenDone
2. MetroViewerHostMsg_MultiFileOpenDone
3. MetroViewerHostMsg_FileSaveAsDone
The handlers for these IPCs live in the RemoteRootWindowHostWin class and invoke the callback objects which are
passed in.
In the metro driver the file_picker_ash.cc/.h files contain the code for displaying the metro style file pickers.
These sources are copied from the file_picker.cc/.h files which perform this functionality for regular Chrome metro.
BUG=170483
Review URL: https://codereview.chromium.org/12041089
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@179921 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
Remove a dependency that was eliminated when
crbug.com/140545 was fixed.
BUG=166496
BUG=140545
Review URL: https://chromiumcodereview.appspot.com/11929041
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@179327 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
This prevents the Metro Chrome viewer process from crashing at exit when using the component build which is annoying for tests that start and stop the viewer process frequently.
BUG=154081
TEST=Metro Chrome does not crash on exit when using the component build.
Review URL: https://chromiumcodereview.appspot.com/12040040
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@178388 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
(I looked at changing size_t to unsigned int elsewhere, but then the interactions with std::vector get messier then).
R=cpu@chromium.org
BUG=166496
Review URL: https://chromiumcodereview.appspot.com/11876033
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@177790 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes C4267 warnings, and errors regarding passing size_t* when they chould be unsigned int *.
winrt_utils:
WindowsCreateString takes a UINT32 as size. Truncation would mean
shorter than expected string, but presumably not a severe problem
in this case.
file_picker:
The interface uses unsigned int, not size_t, so change implementation and callers to unsigned int.
R=cpu@chromium.org
BUG=166496
Review URL: https://chromiumcodereview.appspot.com/11884021
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@176854 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
ref: http://msdn.microsoft.com/en-us/library/windows/desktop/ms644898(v=vs.85).aspx
R=cpu@chromium.org
BUG=166496
Review URL: https://chromiumcodereview.appspot.com/11882029
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@176847 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
This is to stop people accidentally using the ERROR macro for other uses,
e.g. VLOG(ERROR) is not meaningful.
BUG=
Review URL: https://chromiumcodereview.appspot.com/11669010
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@175998 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
To change the cursor we need to send a message from
the browser to the viewer with the right HCURSOR. Since the
cursors are a GDI shared resource, we can call ::LoadCursor
in the browser and ::SetCursor in the viewer.
BUG=151718
TEST= moving pointer to a chrome window edge changes cursor to resize cursor
Review URL: https://codereview.chromium.org/11446078
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@172265 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
to the browser process.
cpu:- Please review everything.
sky:- Please review changes to content\browser, chrome\browser and ui\aura.
This effectively also means that we won't be passing the fake mouse messages generated by Windows for touch events
from the Ash viewer process to chrome.
This only fixes touch handling in Metro Ash. Touch handling for desktop chrome Aura will be fixed in a subsequent CL.
BUG=151718
R=cpu,sky
Review URL: https://codereview.chromium.org/11472013
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@171888 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
is snapped, the new window should be snapped.
Currently the new chrome window which becomes visible ends up with the wrong size when metro chrome is snapped. We should ensure
that the new chrome window is also snapped and the IDC_METRO_SNAP_ENABLE message is sent to it. This ensures that the behavior remains
consistent, i.e the omnibox remains invisible, etc.
BUG=163458
R=robertshield
Review URL: https://codereview.chromium.org/11442004
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@171265 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
triggered from hotdog -> Bookmarks when in Metro mode.
To perform the metrics recording, this CL includes changes
to provide a callback mechanism for the asynchronous tile
creation in metro_driver.dll to report back the result of
the use gesture to the browser process. Note that both
metro_driver.dll and chrome.dll link statically to base, so
they have distinct data segments holding UMA histograms,
hence the callback.
BUG=160840
TEST=In Windows 8 Metro Mode, Hotdog -> Bookmarks -> Pin to start page:
- pin and cancel (escape/click/touch off popup)
- pin and confirm
- (when pinned) unpin and cancel
- unpin and confirm
Above should all work (and report UMA histogram data).
Review URL: https://chromiumcodereview.appspot.com/11280112
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@170916 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
The breakpad export has been unused for the last five months, so I suppose it's safe to remove it entirely.
BUG=162405
TEST=Chromium Win (aura) bot should be able to compile metro_driver.dll.
Review URL: https://chromiumcodereview.appspot.com/11412265
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@170555 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
state between mouse pressed/released events and pass it down to MouseEvent through flags.
MouseEvent now automatically converts a ET_MOUSE_MOVED event into an ET_MOUSE_DRAGGED event if the flags show that any mouse button is down (mimicking drag behavior of EventTypeFromNative() used for native events).
NOTE: From experimentation, it seems like Metro only sends a PointerPressed event for the first button pressed and the last button released in a sequence of mouse events.
For example, a sequence of LEFT_DOWN, RIGHT_DOWN, LEFT_UP, RIGHT_UP results only in PointerPressed(LEFT)/PointerReleased(RIGHT) events.
Thus, we cannot accumulate mouse down events and know each button that is down, nor can we wait for a matching mouse up event to cancel that state.
TBR=cdn
BUG=151718
TEST=Can now drag windows in Metro Ash.
Review URL: https://chromiumcodereview.appspot.com/11411155
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@170292 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
BUG=151718
TEST=The browser process launched for the metro/ash viewer process is not subject to Metro-style message-loop timeout killing.
Review URL: https://chromiumcodereview.appspot.com/11419069
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@169376 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
the core window by invoking the ICoreApplicationExit::Exit function in a task.
At times metro chrome does not exit after sending the Alt + F4 key combination to the core window. This occurs at times when tabs are closed via the keyboard.
To ensure that chrome exits in this case we invoke the ICoreApplicationExit::Exit function in a background delayed task.
Fixes bug https://code.google.com/p/chromium/issues/detail?id=159147
BUG=159147
R=cpu
Review URL: https://codereview.chromium.org/11308128
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@168895 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
host desktop type.
This is for use in those places where there's no context from which to figure out if something should be popped on the native or ash desktop.
Original approach reviewed here: https://chromiumcodereview.appspot.com/11188064/
BUG=129187
TEST=none
Review URL: https://chromiumcodereview.appspot.com/11358090
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@168074 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
background window.
BUG=151718
TEST=Launch metro ash, see it start.
Review URL: https://chromiumcodereview.appspot.com/11367060
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@167409 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=160332
Review URL: https://chromiumcodereview.appspot.com/11275261
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@167383 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
is for developer Chrome only.
TBR=grt@chromium.org
(already LGTMed in https://codereview.chromium.org/11362092/ (patch set 5), but needs to be dcommitted separately)
Review URL: https://codereview.chromium.org/11369101
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@166248 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
-Wheel, r-click and l-click now discriminated
-More efficient, clean mouse event handling
Note; wheel event not properly handled. That is a bigger CL
that will come later.
BUG=151718
TEST= right click and left click
Review URL: https://codereview.chromium.org/11312025
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@166034 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
BUG=151718
TEST=Close last Chrome window in metro/ash.
TBR=cpu
Review URL: https://chromiumcodereview.appspot.com/11368081
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@165991 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
please review everything,
sky: Please review the changes to ui and chrome\browser.
We need to send character events to the RenderViewHost class in addition to the KeyDown/Keyup events
for this to work.
To achieve this we register the character received event on the Windows 8 CoreWindow class and send the
character info via a new IPC message MetroViewerHostMsg_Character.
The chrome side handling for this message is performed in the MockInputMethod::DispatchFabricatedKeyEvent
function where we dispatch the message to the focused TextInputClient.
While working on this I also noticed that we were not sending the event flags like shift/alt/Ctrl, etc to
Chrome from the Metro code. Added support for that. This ensures that we are able to select text in the edit
fields on the page and the omnibox, etc.
The above changes will most likely not work with IME's. Will look into that in a subsequent change.
BUG=158637
R=cpu,sky
Review URL: https://codereview.chromium.org/11348053
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@165496 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
When pages are pinned to the start screen, if there is a favicon available for the site it is used to generate the tile.
BUG=140942
Review URL: https://chromiumcodereview.appspot.com/11198025
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@163943 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
ChromeAppViewFactory now produces ChromeAppView or ChromeAppViewAsh depending on the use_aura gyp flag.
The two versions have very little in common. Overt time they might share more code.
BUG=151718
TEST=see bug
Review URL: https://codereview.chromium.org/11233070
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@163860 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
This centralizes this definition and removes it from the various places it is currently replicated. This also allows external projects that rely on Chromium base to compile successfully, as otherwise this (required) definition is missing.
BUG=
Review URL: https://chromiumcodereview.appspot.com/11190060
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@163299 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is the first part since there IME are issues
that need to be sorted out before this can fully
work. In particular, InputMethodEventFilter::PreHandleKeyEvent
needs changes not to assume native events.
BUG=151718
TEST=none
Review URL: https://codereview.chromium.org/11194044
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@162781 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
Just removing condition because 2008 is toast.
BUG=143646
R=cpu@chromium.org
TBR=ben@chromium.org
Review URL: https://codereview.chromium.org/11184012
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@162280 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Chrome in Windows 8 mode. The crash
occurs because we attempt to initiate a navigation when there are no frame windows registered with the metro driver.
Fix is to ignore the navigation request in this case.
I also updated the ChromeUrlLaunchHandler::GetUrlFromLaunchArgs function to return the chrome://newtab url if the verb
coming in is opennewwindow. This should address bug http://code.google.com/p/chromium/issues/detail?id=135122
BUG=155281,135122
R=cpu
Review URL: https://codereview.chromium.org/11193007
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@162269 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Missed the actual mouse sending in the previous CL.
https://codereview.chromium.org/11047012/
Note that to view the aura-in-metro you need
0) chrome as default browser
1) start chrome with --open-ash
2) hold shift-f11 while clicking on the chrome metro tile
BUG=151718
TEST=see bug
Review URL: https://codereview.chromium.org/11088083
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@161387 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
invoke a caller specified handler function
with a string parameter.
Currently the only consumer of this functionality is the download notification display code which when invoked displays
the downloaded file in the shell.
BUG=151141
R=cpu
Review URL: https://codereview.chromium.org/11096013
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@160988 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
- Wires metro mouse move, mouse click to aura viewer
- Introduces a new RootWindowHost, one that is remote
BUG=151718
TEST=see bug
Review URL: https://codereview.chromium.org/11047012
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@160489 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
Render to the surface it provides to us.
BUG=151718
Review URL: https://codereview.chromium.org/10984007
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@159559 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
metro mode are at the top of the
window list maintained by the metro driver. This regressed with my change to fix the OSK scrolling bug:
http://code.google.com/p/chromium/issues/detail?id=150848
BUG=150848
R=cpu
Review URL: https://codereview.chromium.org/11014005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@159374 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
last week.
BUG=none
TBR=cpu
Review URL: https://codereview.chromium.org/10982081
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@159147 0039d316-1c4b-4281-b951-d872f2087c98
|