Adding a format device button to UI.

Adding a format device button to UI.
Additionally local path translation is now supported by filebrowser extension in formatting routine. 


BUG=chromium-os:4541, chromium-os:17071
TEST=Try to format removable device


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=96682

[sidor.dev, 2011-08-10 18:51:05]

Hi, 

Could you please review my code

Thanks,
Szymon

[tbarzic, 2011-08-10 21:04:56]

I haven't gone through js code, yet.

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
File chrome/browser/extensions/extension_file_browser_private_api.cc (right):

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_file_browser_private_api.cc:1310: if
(!files.size()) {
files.size() != 1 (you can format only one device)

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_file_browser_private_api.cc:1315:
std::string mount_path = files[0].value().data();
files[0].value() should be sufficient..

You could also do this (it's up to you):
std::string mount_path(files[0].value());

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_file_browser_private_api.cc:1316: if
(mount_path[mount_path.length()-1] != '/')
spaces around -

Also, I don't think you shouldn't add '/' at the end of the string..

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
File chrome/browser/extensions/extension_file_browser_private_api.h (right):

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_file_browser_private_api.h:293: const
FilePathList& files, void* context) OVERRIDE;
does const FilePathList& files fit into previous line?
If it does move it there ;)

[sidor.dev, 2011-08-10 22:38:16]

Done.

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
File chrome/browser/extensions/extension_file_browser_private_api.cc (right):

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_file_browser_private_api.cc:1310: if
(!files.size()) {
On 2011/08/10 21:04:56, toni wrote:
> files.size() != 1 (you can format only one device)

Done.

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_file_browser_private_api.cc:1315:
std::string mount_path = files[0].value().data();
On 2011/08/10 21:04:56, toni wrote:
> files[0].value() should be sufficient..
> 
> You could also do this (it's up to you):
> std::string mount_path(files[0].value());

Done.

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_file_browser_private_api.cc:1316: if
(mount_path[mount_path.length()-1] != '/')
spaces done.

second issue - discussed.

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
File chrome/browser/extensions/extension_file_browser_private_api.h (right):

http://codereview.chromium.org/7583041/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_file_browser_private_api.h:293: const
FilePathList& files, void* context) OVERRIDE;
Done + fixed others in that file.

[rginda, 2011-08-10 23:08:09]

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
File chrome/browser/resources/file_manager/js/file_manager.js (right):

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1615:
FileManager.prototype.DisplayFormattingTask1_ = function(entries) {
JavaScript naming convention has leadingLowercase for anything that's not a
Constructor or a CONSTANT.

Since this may not actually create a button, and 'render' is used in a simliar
context near by, I'd suggest 'maybeRenderFormattingTask'.

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1623:
FileManager.prototype.DisplayFormattingTask2_ = function(mountPath,

The convention elsewhere in this file is to declare the async callbacks as
nested functions with descriptive names.  In this case, 'onMountPointsFound'
would be appropriate.

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1632: for (var i = 0; i
< mountPoints.length; ++i) {
i++

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1642:
FileManager.prototype.DisplayFormattingTask3_ = function(volumeMetadata) {
and onVolumeMetadataFound

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1649:
this.renderTaskButton_(task);
You should probably omit the entries parameter to the outer function and instead
rely on this.selection.  Then at the point where the volume metadata is
returned, you can verify that the selection has not changed during the async
calls.  If it has, just bail out on the button rendering, since another
invocation will be taking care of it.

[sidor.dev, 2011-08-10 23:51:21]

Thanks for review. I made the requested changes

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
File chrome/browser/resources/file_manager/js/file_manager.js (right):

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1615:
FileManager.prototype.DisplayFormattingTask1_ = function(entries) {
On 2011/08/10 23:08:09, rginda wrote:
> JavaScript naming convention has leadingLowercase for anything that's not a
> Constructor or a CONSTANT.
> 
> Since this may not actually create a button, and 'render' is used in a simliar
> context near by, I'd suggest 'maybeRenderFormattingTask'.

Done.

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1623:
FileManager.prototype.DisplayFormattingTask2_ = function(mountPath,
On 2011/08/10 23:08:09, rginda wrote:
> 
> The convention elsewhere in this file is to declare the async callbacks as
> nested functions with descriptive names.  In this case, 'onMountPointsFound'
> would be appropriate.

Done.

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1632: for (var i = 0; i
< mountPoints.length; ++i) {
On 2011/08/10 23:08:09, rginda wrote:
> i++

Done. Out of curiosity - why is that a convention in JS? In C++ it's usually
++i...

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1642:
FileManager.prototype.DisplayFormattingTask3_ = function(volumeMetadata) {
On 2011/08/10 23:08:09, rginda wrote:
> and onVolumeMetadataFound

Done.

http://codereview.chromium.org/7583041/diff/5001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1649:
this.renderTaskButton_(task);
On 2011/08/10 23:08:09, rginda wrote:
> You should probably omit the entries parameter to the outer function and
instead
> rely on this.selection.  Then at the point where the volume metadata is
> returned, you can verify that the selection has not changed during the async
> calls.  If it has, just bail out on the button rendering, since another
> invocation will be taking care of it.

Done.

[tbarzic, 2011-08-11 00:58:18]

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
File chrome/browser/resources/file_manager/js/file_manager.js (right):

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1631: function
onVolumeMetadataFound(volumeMetadata) {
what if selection has changed by now?

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1644: var
maybeMountPath = this.selection.entries[0].fullPath;
why  maybe here?

[sidor, 2011-08-11 04:28:50]

Let me know what do you think.

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
File chrome/browser/resources/file_manager/js/file_manager.js (right):

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1631: function
onVolumeMetadataFound(volumeMetadata) {
On 2011/08/11 00:58:18, toni wrote:
> what if selection has changed by now?

Hmm... Then we are screwed. But hey, to change the selection we need user
clicking right? So it shouldn't be a problem. 

Especially that this call don't go the any of libcros packets right? Metadata is
cached?

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1644: var
maybeMountPath = this.selection.entries[0].fullPath;
Because it might be a regular directory path.

[tbarzic, 2011-08-11 04:49:48]

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
File chrome/browser/resources/file_manager/js/file_manager.js (right):

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1631: function
onVolumeMetadataFound(volumeMetadata) {
On 2011/08/11 04:28:51, sidor wrote:
> On 2011/08/11 00:58:18, toni wrote:
> > what if selection has changed by now?
> 
> Hmm... Then we are screwed. But hey, to change the selection we need user
> clicking right? So it shouldn't be a problem. 
> 
> Especially that this call don't go the any of libcros packets right? Metadata
is
> cached?

maybe you could do something like
if (this.selection.entries.length != 1 ||
    '/media/'+this.selection.entries[0].fullPath != volumeMetadata.mountPath)
  return;

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1644: var
maybeMountPath = this.selection.entries[0].fullPath;
On 2011/08/11 04:28:51, sidor wrote:
> Because it might be a regular directory path. 

maybe entryPath or selectedPath would be better name...

[sidor.dev, 2011-08-11 18:39:43]

done.

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
File chrome/browser/resources/file_manager/js/file_manager.js (right):

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1631: function
onVolumeMetadataFound(volumeMetadata) {
On 2011/08/11 04:49:48, toni wrote:
> On 2011/08/11 04:28:51, sidor wrote:
> > On 2011/08/11 00:58:18, toni wrote:
> > > what if selection has changed by now?
> > 
> > Hmm... Then we are screwed. But hey, to change the selection we need user
> > clicking right? So it shouldn't be a problem. 
> > 
> > Especially that this call don't go the any of libcros packets right?
Metadata
> is
> > cached?
> 
> maybe you could do something like
> if (this.selection.entries.length != 1 ||
>     '/media/'+this.selection.entries[0].fullPath != volumeMetadata.mountPath)
>   return;

Done.

http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1644: var
maybeMountPath = this.selection.entries[0].fullPath;
On 2011/08/11 04:49:48, toni wrote:
> On 2011/08/11 04:28:51, sidor wrote:
> > Because it might be a regular directory path. 
> 
> maybe entryPath or selectedPath would be better name...

Done.

[tbarzic, 2011-08-11 19:51:04]

On 2011/08/11 18:39:43, sidor.dev wrote:
> done.
> 
>
http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
> File chrome/browser/resources/file_manager/js/file_manager.js (right):
> 
>
http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
> chrome/browser/resources/file_manager/js/file_manager.js:1631: function
> onVolumeMetadataFound(volumeMetadata) {
> On 2011/08/11 04:49:48, toni wrote:
> > On 2011/08/11 04:28:51, sidor wrote:
> > > On 2011/08/11 00:58:18, toni wrote:
> > > > what if selection has changed by now?
> > > 
> > > Hmm... Then we are screwed. But hey, to change the selection we need user
> > > clicking right? So it shouldn't be a problem. 
> > > 
> > > Especially that this call don't go the any of libcros packets right?
> Metadata
> > is
> > > cached?
> > 
> > maybe you could do something like
> > if (this.selection.entries.length != 1 ||
> >     '/media/'+this.selection.entries[0].fullPath !=
volumeMetadata.mountPath)
> >   return;
> 
> Done.
> 
>
http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
> chrome/browser/resources/file_manager/js/file_manager.js:1644: var
> maybeMountPath = this.selection.entries[0].fullPath;
> On 2011/08/11 04:49:48, toni wrote:
> > On 2011/08/11 04:28:51, sidor wrote:
> > > Because it might be a regular directory path. 
> > 
> > maybe entryPath or selectedPath would be better name...
> 
> Done.

As far as I'm concerned, LGTM,
but Robert might have some more suggestions..

[tbarzic, 2011-08-11 19:55:28]

On 2011/08/11 19:51:04, toni wrote:
> On 2011/08/11 18:39:43, sidor.dev wrote:
> > done.
> > 
> >
>
http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
> > File chrome/browser/resources/file_manager/js/file_manager.js (right):
> > 
> >
>
http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
> > chrome/browser/resources/file_manager/js/file_manager.js:1631: function
> > onVolumeMetadataFound(volumeMetadata) {
> > On 2011/08/11 04:49:48, toni wrote:
> > > On 2011/08/11 04:28:51, sidor wrote:
> > > > On 2011/08/11 00:58:18, toni wrote:
> > > > > what if selection has changed by now?
> > > > 
> > > > Hmm... Then we are screwed. But hey, to change the selection we need
user
> > > > clicking right? So it shouldn't be a problem. 
> > > > 
> > > > Especially that this call don't go the any of libcros packets right?
> > Metadata
> > > is
> > > > cached?
> > > 
> > > maybe you could do something like
> > > if (this.selection.entries.length != 1 ||
> > >     '/media/'+this.selection.entries[0].fullPath !=
> volumeMetadata.mountPath)
> > >   return;
> > 
> > Done.
> > 
> >
>
http://codereview.chromium.org/7583041/diff/6001/chrome/browser/resources/fil...
> > chrome/browser/resources/file_manager/js/file_manager.js:1644: var
> > maybeMountPath = this.selection.entries[0].fullPath;
> > On 2011/08/11 04:49:48, toni wrote:
> > > On 2011/08/11 04:28:51, sidor wrote:
> > > > Because it might be a regular directory path. 
> > > 
> > > maybe entryPath or selectedPath would be better name...
> > 
> > Done.
> 
> As far as I'm concerned, LGTM,
> but Robert might have some more suggestions..

Actually, you have to edit tests for GetVolumeMetadata (change expected disk
values in chrome/test/data/exetnsions/api_test/filebrowser_mount/test.html

[rginda, 2011-08-11 20:20:13]

http://codereview.chromium.org/7583041/diff/9002/chrome/browser/resources/fil...
File chrome/browser/resources/file_manager/js/file_manager.js (right):

http://codereview.chromium.org/7583041/diff/9002/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1660:
chrome.fileBrowserPrivate.getMountPoints(onMountPointsFound.bind(this));
The convention for nested functions is to create a local variable called 'self'
to hold the this object, rather than calling bind().  Then reference self.foo
inside the nested functions.

http://codereview.chromium.org/7583041/diff/9002/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1728: if
(!confirm(str('FORMATTING_WARNING')))
App modal dialogs are no longer allowed here.  You'll need to use this.confirm
here instead.  Watch out, it's another async method, so you'll have to refactor
a little.

[sidor.dev, 2011-08-12 02:05:19]

Style updates + 
Tony - why u no let my look at your fixes of my code before comitting? :)

http://codereview.chromium.org/7583041/diff/9002/chrome/browser/resources/fil...
File chrome/browser/resources/file_manager/js/file_manager.js (right):

http://codereview.chromium.org/7583041/diff/9002/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1660:
chrome.fileBrowserPrivate.getMountPoints(onMountPointsFound.bind(this));
On 2011/08/11 20:20:14, rginda wrote:
> The convention for nested functions is to create a local variable called
'self'
> to hold the this object, rather than calling bind().  Then reference self.foo
> inside the nested functions.

Done.

http://codereview.chromium.org/7583041/diff/9002/chrome/browser/resources/fil...
chrome/browser/resources/file_manager/js/file_manager.js:1728: if
(!confirm(str('FORMATTING_WARNING')))
On 2011/08/11 20:20:14, rginda wrote:
> App modal dialogs are no longer allowed here.  You'll need to use this.confirm
> here instead.  Watch out, it's another async method, so you'll have to
refactor
> a little.

Done.

[tbarzic, 2011-08-12 18:29:35]

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
File chrome/browser/chromeos/cros/mount_library.cc (right):

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:307: it != self->disks_.end();
++it) {
You should encapsulate path conversion into separate method

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:462: // In mount library we use
device path to identify device.
I don't think this comment is needed

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:635: it != disks_.end(); ++it) {
use 5 space indent here

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:637: path =
it->second->device_path().c_str();
Why don't you loose path and use device_path directly here?

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:643: << "device that is being
formatted.";
align << with << in the previous row.

[sidor.dev, 2011-08-12 19:14:20]

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
File chrome/browser/chromeos/cros/mount_library.cc (right):

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:307: it != self->disks_.end();
++it) {
On 2011/08/12 18:29:35, toni wrote:
> You should encapsulate path conversion into separate method

Done.

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:462: // In mount library we use
device path to identify device.
On 2011/08/12 18:29:35, toni wrote:
> I don't think this comment is needed

Sorry, forgot to delete it.

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:635: it != disks_.end(); ++it) {
On 2011/08/12 18:29:35, toni wrote:
> use 5 space indent here

Done.

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:637: path =
it->second->device_path().c_str();
On 2011/08/12 18:29:35, toni wrote:
> Why don't you loose path and use device_path directly here?

Done.

http://codereview.chromium.org/7583041/diff/18001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:643: << "device that is being
formatted.";
On 2011/08/12 18:29:35, toni wrote:
> align << with << in the previous row.

Done.

[tbarzic, 2011-08-12 19:31:12]

http://codereview.chromium.org/7583041/diff/20001/chrome/browser/chromeos/cro...
File chrome/browser/chromeos/cros/mount_library.cc (right):

http://codereview.chromium.org/7583041/diff/20001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:274: private:
add comment // static.

http://codereview.chromium.org/7583041/diff/20001/chrome/browser/chromeos/cro...
File chrome/browser/chromeos/cros/mount_library.h (right):

http://codereview.chromium.org/7583041/diff/20001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.h:170: virtual std::string
FilePathToDevicePath(std::string file_path) = 0;
Please, add this one to MountLibraryImpl, not MountLibrary.
Also, it should return const char*.

[sidor.dev, 2011-08-12 20:27:09]

http://codereview.chromium.org/7583041/diff/20001/chrome/browser/chromeos/cro...
File chrome/browser/chromeos/cros/mount_library.cc (right):

http://codereview.chromium.org/7583041/diff/20001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:274: private:
On 2011/08/12 19:31:13, toni wrote:
> add comment // static.

no.

http://codereview.chromium.org/7583041/diff/20001/chrome/browser/chromeos/cro...
File chrome/browser/chromeos/cros/mount_library.h (right):

http://codereview.chromium.org/7583041/diff/20001/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.h:170: virtual std::string
FilePathToDevicePath(std::string file_path) = 0;
On 2011/08/12 19:31:13, toni wrote:
> Please, add this one to MountLibraryImpl, not MountLibrary.
> Also, it should return const char*.

Done.

[rginda, 2011-08-12 20:33:59]

> On 2011/08/10 23:08:09, rginda wrote:
>> i++
>
> Done. Out of curiosity - why is that a convention in JS? In C++ it's usually
> ++i...

I don't know.  I prefer pre-increment too, but for whatever reason the js
convention is post.

LGTM

[tbarzic, 2011-08-12 20:49:17]

LGTM with a nit

http://codereview.chromium.org/7583041/diff/19003/chrome/browser/chromeos/cro...
File chrome/browser/chromeos/cros/mount_library.cc (right):

http://codereview.chromium.org/7583041/diff/19003/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:308: "device that is being
formatted.";
see below

http://codereview.chromium.org/7583041/diff/19003/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:627: LOG(ERROR) << "Error while
handling disks metadata. Cannot find " <<
LOG(ERROR) << "Error..."
           << "device.."

[sidor.dev, 2011-08-12 20:52:05]

http://codereview.chromium.org/7583041/diff/19003/chrome/browser/chromeos/cro...
File chrome/browser/chromeos/cros/mount_library.cc (right):

http://codereview.chromium.org/7583041/diff/19003/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:308: "device that is being
formatted.";
On 2011/08/12 20:49:17, toni wrote:
> see below

Done.

http://codereview.chromium.org/7583041/diff/19003/chrome/browser/chromeos/cro...
chrome/browser/chromeos/cros/mount_library.cc:627: LOG(ERROR) << "Error while
handling disks metadata. Cannot find " <<
On 2011/08/12 20:49:17, toni wrote:
> LOG(ERROR) << "Error..."
>            << "device.."

Done.

[commit-bot: I haz the power, 2011-08-13 04:38:28]

Change committed as 96682