Add a new version of the power save blocker class, to be implemented for all platforms.

Add a new version of the power save blocker class, to be implemented for all platforms in subsequent CLs.
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=140460

[Mike Mammarella, 2012-06-04 21:53:42]

https://chromiumcodereview.appspot.com/10511021/diff/1003/content/browser/pow...
File content/browser/power_save_blocker2.h (right):

https://chromiumcodereview.appspot.com/10511021/diff/1003/content/browser/pow...
content/browser/power_save_blocker2.h:43: PowerSaveBlockerImpl* impl;
I could also make this a scoped_refptr if it turns out that's what all the
implementations want. I am pretty sure the Linux and Mac versions will want
that, for instance.

[Avi (use Gerrit), 2012-06-04 21:58:52]

https://chromiumcodereview.appspot.com/10511021/diff/6001/content/browser/pow...
File content/browser/power_save_blocker2.h (right):

https://chromiumcodereview.appspot.com/10511021/diff/6001/content/browser/pow...
content/browser/power_save_blocker2.h:32: explicit
PowerSaveBlocker2(PowerSaveBlockerType type);
Can we get a string to describe why?

https://chromiumcodereview.appspot.com/10511021/diff/6001/content/browser/pow...
content/browser/power_save_blocker2.h:41: PowerSaveBlockerImpl* impl;
Huh. I was going to go for #ifdef-ed member variables here, but I'm OK with
this.

[Avi (use Gerrit), 2012-06-04 22:00:30]

https://chromiumcodereview.appspot.com/10511021/diff/1003/content/browser/pow...
File content/browser/power_save_blocker2.h (right):

https://chromiumcodereview.appspot.com/10511021/diff/1003/content/browser/pow...
content/browser/power_save_blocker2.h:43: PowerSaveBlockerImpl* impl;
I have baggage per-blocker, so I'd prefer a scoped_ptr. Worst case I'll take the
raw pointer and delete it in the destructor.

[Mike Mammarella, 2012-06-04 22:21:28]

https://chromiumcodereview.appspot.com/10511021/diff/6001/content/browser/pow...
File content/browser/power_save_blocker2.h (right):

https://chromiumcodereview.appspot.com/10511021/diff/6001/content/browser/pow...
content/browser/power_save_blocker2.h:32: explicit
PowerSaveBlocker2(PowerSaveBlockerType type);
On 2012/06/04 21:58:52, Avi wrote:
> Can we get a string to describe why?

Done.

https://chromiumcodereview.appspot.com/10511021/diff/6001/content/browser/pow...
content/browser/power_save_blocker2.h:41: PowerSaveBlockerImpl* impl;
On 2012/06/04 21:58:52, Avi wrote:
> Huh. I was going to go for #ifdef-ed member variables here, but I'm OK with
> this.

We could do that too I suppose. In the Linux version (which I haven't actually
started on yet) I am pretty sure I'll want a refcounted delegate with a
different lifetime to pass in messages to another thread, and that's where all
the real state will be anyway. So I figured it's likely that other
implementations could do that too...

[rvargas (doing something else), 2012-06-04 23:28:30]

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/content_bro...
File content/content_browser.gypi (right):

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/content_bro...
content/content_browser.gypi:441: 'browser/power_save_blocker2.h',
Can we just use the same file as before so that there's no need to rename the
file at the end?

[rvargas (doing something else), 2012-06-04 23:34:48]

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/browser/pow...
File content/browser/power_save_blocker2.h (right):

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/browser/pow...
content/browser/power_save_blocker2.h:16: class CONTENT_EXPORT PowerSaveBlocker2
{
... and I believe this now has to be part of the content namespace.

[Avi (use Gerrit), 2012-06-04 23:50:50]

I'd be good implementing this. LGTM

[Avi (use Gerrit), 2012-06-05 00:21:52]

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/content_bro...
File content/content_browser.gypi (right):

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/content_bro...
content/content_browser.gypi:441: 'browser/power_save_blocker2.h',
Clever! I like it, as long as we comment PowerSaveBlocker as "the deprecated
way" and PowerSaveBlocker2 as "the way that doesn't work yet".

[Mike Mammarella, 2012-06-05 00:28:42]

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/browser/pow...
File content/browser/power_save_blocker2.h (right):

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/browser/pow...
content/browser/power_save_blocker2.h:16: class CONTENT_EXPORT PowerSaveBlocker2
{
On 2012/06/04 23:34:48, rvargas wrote:
> ... and I believe this now has to be part of the content namespace.

Done.

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/content_bro...
File content/content_browser.gypi (right):

https://chromiumcodereview.appspot.com/10511021/diff/4002/content/content_bro...
content/content_browser.gypi:441: 'browser/power_save_blocker2.h',
On 2012/06/04 23:28:30, rvargas wrote:
> Can we just use the same file as before so that there's no need to rename the
> file at the end?

Done.

[rvargas (doing something else), 2012-06-05 00:30:31]

LGTM

[Avi (use Gerrit), 2012-06-05 00:31:19]

LGTM! Whoo! Ready to implement.

[commit-bot: I haz the power, 2012-06-05 00:39:40]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mdm@chromium.org/10511021/9002

[rvargas (doing something else), 2012-06-05 00:48:50]

https://chromiumcodereview.appspot.com/10511021/diff/9002/content/browser/pow...
File content/browser/power_save_blocker.h (right):

https://chromiumcodereview.appspot.com/10511021/diff/9002/content/browser/pow...
content/browser/power_save_blocker.h:96: scoped_refptr<Delegate> delegate;
nit: this should be delegate_

[Avi (use Gerrit), 2012-06-05 00:57:34]

On 2012/06/05 00:31:19, Avi wrote:
> LGTM! Whoo! Ready to implement.

I see what you did here. You need a shared refptr so you can toss the
implementation objects across to a different thread... clever!

[Mike Mammarella, 2012-06-05 01:02:12]

Yep, that's why I wanted scoped_refptr. :)

https://chromiumcodereview.appspot.com/10511021/diff/9002/content/browser/pow...
File content/browser/power_save_blocker.h (right):

https://chromiumcodereview.appspot.com/10511021/diff/9002/content/browser/pow...
content/browser/power_save_blocker.h:96: scoped_refptr<Delegate> delegate;
On 2012/06/05 00:48:50, rvargas wrote:
> nit: this should be delegate_

Ah yes, so it should. I'll wait for the bots to finish compiling, cancel the CQ,
and commit by hand with the _ here.

[Avi (use Gerrit), 2012-06-05 01:54:03]

On 2012/06/05 01:02:12, Mike Mammarella wrote:
> Yep, that's why I wanted scoped_refptr. :)

And I need it too, for the same reason. You, sir, are more forward-looking than
me.