Ties up OOBE controllers and WebUI handlers.

Ties up OOBE controllers and WebUI handlers.

OobeDisplay ownership was moved out of WizardController. In case of WebUI, WebUI implements OobeDisplay. In case of views, OobeDisplay is owned by the LoginDisplayHost. Also new flag was added to activate WebUI for OOBE.

BUG=chromium-os:15923
TEST=manual

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=88517

[Nikita (slow), 2011-06-07 09:29:52]

As discussed, please make sure that only one WebUI host window is created for
OOBE / login and it is navigated to proper URLs.

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches.cc
File chrome/common/chrome_switches.cc (right):

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches....
chrome/common/chrome_switches.cc:44: // Allows WebUI based OOBE. Note, WebUI
will be used iff kWebUILogin is also
nit: iff > if

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches....
chrome/common/chrome_switches.cc:46: const char kAllowWebUIOobe[]               
= "allow-webui-oobe";
Just webui-oobe.

[altimofeev, 2011-06-07 14:54:52]

Login window was moved to the WebUILoginDisplyaHost. Please take another look.

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches.cc
File chrome/common/chrome_switches.cc (right):

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches....
chrome/common/chrome_switches.cc:44: // Allows WebUI based OOBE. Note, WebUI
will be used iff kWebUILogin is also
On 2011/06/07 09:29:52, Nikita Kostylev wrote:
> nit: iff > if

I meant "if and only if" here. Is it ok?

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches....
chrome/common/chrome_switches.cc:46: const char kAllowWebUIOobe[]               
= "allow-webui-oobe";
On 2011/06/07 09:29:52, Nikita Kostylev wrote:
> Just webui-oobe.

Actually, it will not work without kWebUILogin. So I think that "allow" part
makes it more correct.

[Nikita (slow), 2011-06-07 16:18:06]

LGTM if it doesn't break current WebUI login page.

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches.cc
File chrome/common/chrome_switches.cc (right):

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches....
chrome/common/chrome_switches.cc:44: // Allows WebUI based OOBE. Note, WebUI
will be used iff kWebUILogin is also
On 2011/06/07 14:54:53, altimofeev wrote:
> On 2011/06/07 09:29:52, Nikita Kostylev wrote:
> > nit: iff > if
> 
> I meant "if and only if" here. Is it ok?

Sorry, wasn't obvious to me.

http://codereview.chromium.org/7111033/diff/15/chrome/common/chrome_switches....
chrome/common/chrome_switches.cc:46: const char kAllowWebUIOobe[]               
= "allow-webui-oobe";
On 2011/06/07 14:54:53, altimofeev wrote:
> On 2011/06/07 09:29:52, Nikita Kostylev wrote:
> > Just webui-oobe.
> 
> Actually, it will not work without kWebUILogin. So I think that "allow" part
> makes it more correct.

Ok, makes sense, to make it more clear.

http://codereview.chromium.org/7111033/diff/5001/chrome/browser/chromeos/logi...
File chrome/browser/chromeos/login/webui_login_display_host.cc (right):

http://codereview.chromium.org/7111033/diff/5001/chrome/browser/chromeos/logi...
chrome/browser/chromeos/login/webui_login_display_host.cc:9: #include
"chrome/browser/chromeos/login/webui_login_view.h"
nit: alpha sort.

http://codereview.chromium.org/7111033/diff/5001/chrome/browser/chromeos/logi...
chrome/browser/chromeos/login/webui_login_display_host.cc:20: const char
kLoginURL[] = "chrome://login";
Please move to chrome/common/url_constants.h,
kChromeUILoginURL

http://codereview.chromium.org/7111033/diff/5001/chrome/browser/chromeos/logi...
chrome/browser/chromeos/login/webui_login_display_host.cc:22: const char
kOobeURL[] = "chrome://oobe";
Use kChromeUIOobeURL from chrome/common/url_constants.h

http://codereview.chromium.org/7111033/diff/5001/chrome/browser/chromeos/logi...
File chrome/browser/chromeos/login/webui_login_view.cc (right):

http://codereview.chromium.org/7111033/diff/5001/chrome/browser/chromeos/logi...
chrome/browser/chromeos/login/webui_login_view.cc:42:
webui_login_->SetVisible(true);
Ryan, does this call is needed? I think that DOMView control will be created
with visibility = true.

[Nikita (slow), 2011-06-07 16:34:33]

http://codereview.chromium.org/7111033/diff/9001/chrome/browser/chromeos/logi...
File chrome/browser/chromeos/login/webui_login_display_host.h (right):

http://codereview.chromium.org/7111033/diff/9001/chrome/browser/chromeos/logi...
chrome/browser/chromeos/login/webui_login_display_host.h:38: virtual void
StartWizard(const std::string& first_screen_name,
OVERRIDE for both?

http://codereview.chromium.org/7111033/diff/9001/chrome/browser/chromeos/logi...
chrome/browser/chromeos/login/webui_login_display_host.h:43: virtual
WizardController* CreateWizardController();
OVERRIDE?

[rharrison, 2011-06-07 18:12:16]

LGTM, with following requirements:
- Please run through linux_touch trybot, since this isn't run on commit at the
moment
- Please test and confirm that there are no regressions for the login screen on
touch and non-touch. If you don't regularly perform touch builds I can test on
touch for you.

[altimofeev, 2011-06-09 09:59:43]

Also run it on device and run touch version on linux. Committing...

http://codereview.chromium.org/7111033/diff/9001/chrome/browser/chromeos/logi...
File chrome/browser/chromeos/login/webui_login_display_host.h (right):

http://codereview.chromium.org/7111033/diff/9001/chrome/browser/chromeos/logi...
chrome/browser/chromeos/login/webui_login_display_host.h:38: virtual void
StartWizard(const std::string& first_screen_name,
On 2011/06/07 16:34:33, Nikita Kostylev wrote:
> OVERRIDE for both?

Done.

http://codereview.chromium.org/7111033/diff/9001/chrome/browser/chromeos/logi...
chrome/browser/chromeos/login/webui_login_display_host.h:43: virtual
WizardController* CreateWizardController();
On 2011/06/07 16:34:33, Nikita Kostylev wrote:
> OVERRIDE?

Done.