Service Worker: Remove legacy code for resolving register() to a version

Service Worker: Remove legacy code for resolving register() to a version

register() resolves to a ServiceWorkerRegistration, so we no longer need
to pass a ServiceWorkerVersion around.

This also cleans up a couple TODOs in RegisterJob:
- In ContinueWithRegistrationForSameScriptUrl, the TODO made sense when
register resolved to a version not a registration.
Now, the current behavior seems OK.
- In UpdateAndContinue, the TODO was misleading. We don't have to handle
an existing installing worker within RegisterJob since each job either
succeeds with a waiting/active worker or fails and cleans up the
installing worker.

BUG=406240
Committed: https://crrev.com/da5382d00deafd9909ef02612cd7d4d0b65e4d74
Cr-Commit-Position: refs/heads/master@{#297131}

[falken, 2014-09-26 08:24:04]

falken@chromium.org changed reviewers:
+ nhiroki@chromium.org

[falken, 2014-09-26 08:24:05]

PTAL

[nhiroki, 2014-09-26 09:35:54]

https://codereview.chromium.org/605163002/diff/1/content/browser/service_work...
File content/browser/service_worker/service_worker_register_job.cc (right):

https://codereview.chromium.org/605163002/diff/1/content/browser/service_work...
content/browser/service_worker/service_worker_register_job.cc:297: if
(existing_registration->GetNewestVersion()) {
The CL description says...

> if there is a waiting version, there must be an active version.

Is this correct? I suspect that there is a no-active version case when the
registration is loaded from the storage and it doesn't have an active version.

[falken, 2014-09-29 01:44:21]

> Is this correct? I suspect that there is a no-active version case when the
registration is loaded from the storage and it doesn't have an active version.

I think you're right. I've revised the patch to keep the existing behavior but
still refactor out the TODO.

[nhiroki, 2014-09-29 02:17:34]

LGTM, nice cleanup! :)

[falken, 2014-09-29 02:19:53]

On 2014/09/29 02:17:34, nhiroki wrote:
> LGTM, nice cleanup! :)

thanks!

[falken, 2014-09-29 02:19:57]

The CQ bit was checked by falken@chromium.org

[commit-bot: I haz the power, 2014-09-29 02:20:35]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/605163002/20001

[commit-bot: I haz the power, 2014-09-29 03:09:05]

Committed patchset #2 (id:20001) as 7a1c077d9ae64c7ab4ee7d8f823cf026320f65a6

[commit-bot: I haz the power, 2014-09-29 03:11:48]

Patchset 2 (id:??) landed as
https://crrev.com/da5382d00deafd9909ef02612cd7d4d0b65e4d74
Cr-Commit-Position: refs/heads/master@{#297131}

[falken, 2014-09-29 03:25:31]

On 2014/09/29 03:11:48, I haz the power (commit-bot) wrote:
> Patchset 2 (id:??) landed as
> https://crrev.com/da5382d00deafd9909ef02612cd7d4d0b65e4d74
> Cr-Commit-Position: refs/heads/master@{#297131}

Oops, the bug number should have been http://crbug.com/396400.