Move the `privacy` extension API out of experimental.

Move the `privacy` extension API out of experimental.

BUG=88030
TEST=browser_test (ExtensionAPITest.Preference*)

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=119031

[Mike West, 2012-01-17 17:54:38]

Hello, all. As we discussed way back before I left for vacation in December,
this CL moves `chrome.extension.privacy` out into the wild blue. I'd appreciate
you taking a look.

Aaron: I think we probably need to talk about the permission message. I just
strawmanned something in so that we'd have something to discuss.

Meggin: Can you take a look at the documentation I've added in (privacy.html)?

Patrick, Jordan: FYI.

[Aaron Boodman, 2012-01-17 23:50:12]

LGTM

https://chromiumcodereview.appspot.com/9244005/diff/1048/chrome/app/generated...
File chrome/app/generated_resources.grd (right):

https://chromiumcodereview.appspot.com/9244005/diff/1048/chrome/app/generated...
chrome/app/generated_resources.grd:4062: Settings that affect your privacy
How about "Chrome's privacy settings"?

Not sure how well that localizes though.

[Mike West, 2012-01-18 02:52:51]

>
https://chromiumcodereview.appspot.com/9244005/diff/1048/chrome/app/generated...
> chrome/app/generated_resources.grd:4062: Settings that affect your privacy
> How about "Chrome's privacy settings"?

Not all of the settings are explicitly "privacy" settings. Autofill, for
example, lives in "Personal Stuff". "... affect your privacy" is probably more
likely to cover the maximal extent of the effect the extension can have.

I still don't like it, though. Maybe the privacy team will have brilliant ideas
when they get to the office in ~5 hours.

[Mike West, 2012-01-18 16:19:40]

Meggin: I rebuilt the docs and added them to the CL. The privacy API
documentation is staged at
http://www.corp.google.com/~mkwst/no_crawl/docs/privacy.html

Thanks!

[mkearney, 2012-01-18 16:40:57]

Cool.

On Wed, Jan 18, 2012 at 8:19 AM, <mkwst@chromium.org> wrote:

> Meggin: I rebuilt the docs and added them to the CL. The privacy API
> documentation is staged at
>
http://www.corp.google.com/~**mkwst/no_crawl/docs/privacy.**html<http://www.c...
>
> Thanks!
>
>
http://codereview.chromium.**org/9244005/<http://codereview.chromium.org/9244...
>

[Mike West, 2012-01-19 19:19:26]

On 2012/01/18 16:40:57, mkearney wrote:
> Cool.

"Cool" as in, "It's great! Commit it right away if not sooner!" or "Cool" as in
"Great! It's on my list!"? :)

Thanks!

-Mike

[mkearney, 2012-01-19 19:26:10]

B - Great, it's on my list.

That said, I've already printed out the doc to review-- its sitting on my
desk. I should have the comments before the end of today.

But first I am going to do a second review of the Clear API docs :-)

Meggin

On Thu, Jan 19, 2012 at 11:19 AM, <mkwst@chromium.org> wrote:

> On 2012/01/18 16:40:57, mkearney wrote:
>
>> Cool.
>>
>
> "Cool" as in, "It's great! Commit it right away if not sooner!" or "Cool"
> as in
> "Great! It's on my list!"? :)
>
> Thanks!
>
> -Mike
>
>
http://codereview.chromium.**org/9244005/<http://codereview.chromium.org/9244...
>

[mkearney, 2012-01-19 23:48:25]

And one more question kind of rhetorical question that might be outside the
scope of this doc, but I figured good to get it out there...

How will users be made aware of an extension's ability to control the Chrome
settings that might affect their privacy? 

Should something be included in the API doc about how users are informed? Should
something be included in the Google Chrome and Privacy doc about the Privacy
API?

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
File chrome/common/extensions/docs/static/privacy.html (right):

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:13: gives background detail
regarding the features which this API can control.
This doc doesn't cover all the features - is it worth softening this sentence a
little bit or is that overkill?

The Chrome Privacy Whitepaper describes in detail many of the features which
this API can control.

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:35: to find the property
you're interested in, then you'll call <code>get()</code>
Small - can we add a Chrome here: 'current value of a Chrome setting'?

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:51: setting has been locked
down by enterprise policy, or already been set by
Small grammar fix - 'already been set' is better as 'already set' or else you
could put the whole verbial phrase - 'has already been set'.

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:55: </p>
Question: does one extension have more control over a setting than another
extension or is it first come first serve? Might be worth clarifying this. Also,
it seems a bit fuzzy to me that any one extension might have control over these
settings without some sort of pecking order, especially third-party extensions.

[Mike West, 2012-01-20 11:12:19]

Thanks for your feedback, Meggin. I've rewritten chunks of the documentation,
which I hope you can take a look at soon (this is also targeting M18 :) ).

Thanks!

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
File chrome/common/extensions/docs/static/privacy.html (right):

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:13: gives background detail
regarding the features which this API can control.
On 2012/01/19 23:48:25, mkearney wrote:
> This doc doesn't cover all the features - is it worth softening this sentence
a
> little bit or is that overkill?

There's a new version of the whitepaper in legal/PR/etc. review right now that
does/should cover all the features in this API. I'll send you a link via email
if you're curious.

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:35: to find the property
you're interested in, then you'll call <code>get()</code>
On 2012/01/19 23:48:25, mkearney wrote:
> Small - can we add a Chrome here: 'current value of a Chrome setting'?

Done.

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:51: setting has been locked
down by enterprise policy, or already been set by
On 2012/01/19 23:48:25, mkearney wrote:
> Small grammar fix - 'already been set' is better as 'already set' or else you
> could put the whole verbial phrase - 'has already been set'.

I don't like the way that sounds, honestly. I think of it as "If the setting has
(been locked down by enterprise policy), or (already been set by ...)." The verb
seems essential on both sides of the "or".

I'll just rewrite the section until we're both happy. It could use more detail
anyway.

http://codereview.chromium.org/9244005/diff/11035/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:55: </p>
On 2012/01/19 23:48:25, mkearney wrote:
> Question: does one extension have more control over a setting than another
> extension or is it first come first serve? Might be worth clarifying this.
Also,
> it seems a bit fuzzy to me that any one extension might have control over
these
> settings without some sort of pecking order, especially third-party
extensions.

The pecking order is well-defined by the Content Settings API. It boils down to
first-come-first-served, with the caveat that a more recently installed
extension can grab control. This, I think, makes intuitive sense. If I install
Privacy Protector, and then install Privacy Protector Plus Plus, I'd expect the
latter's setting to take hold. I'll add some documentation around what an
extension's options are in that scenario, and link off to the appropriate
section of the documentation.

[Mike West, 2012-01-20 16:15:09]

Meggin: The updated doc is staged at
http://www.corp.google.com/~mkwst/no_crawl/docs/privacy.html

Aaron, Jordan: The team here has converged on "Privacy-relevant settings." for
the permission message. Is that cool with you folks? If so, I'll land this
whenever the docs are up to snuff.

Thanks!

[Mike West, 2012-01-20 16:16:14]

> Aaron, Jordan: The team here has converged on "Privacy-relevant settings." for

Sorry, "Privacy-related settings."

-Mike

[mkearney, 2012-01-20 16:24:41]

lgtm

[Mike West, 2012-01-20 16:42:14]

On 2012/01/20 16:24:41, mkearney wrote:
> lgtm

Meggin noted that we need to update the permission warning page: that's staged
at http://www.corp.google.com/~mkwst/no_crawl/docs/permission_warnings.html

[Mike West, 2012-01-20 16:54:40]

On 2012/01/20 16:42:14, Mike West (chromium) wrote:
> On 2012/01/20 16:24:41, mkearney wrote:
> > lgtm
> 
> Meggin noted that we need to update the permission warning page: that's staged
> at http://www.corp.google.com/%7Emkwst/no_crawl/docs/permission_warnings.html

And the other permissions warning page at
http://www.corp.google.com/~mkwst/no_crawl/docs/manifest.html

[Mike West, 2012-01-24 12:15:08]

On 2012/01/20 16:16:14, Mike West (chromium) wrote:
> > Aaron, Jordan: The team here has converged on "Privacy-relevant settings."
for
> 
> Sorry, "Privacy-related settings."

Aaron, Jordan: Friendly ping. If you two are fine with this string, then we're
good to go.

[battre, 2012-01-24 12:29:17]

drive-by comments

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
File chrome/common/extensions/docs/static/privacy.html (right):

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:19: <a href="manifest.html">
manifest</a> to use the API. For example:
nit: remove space before manifest

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:51: deny access to a setting
if the setting is either locked to a specific value
technically, it won't deny access but only ignore the written value until it
becomes the most important (policy is revoked; other extension clears its
preference)

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:58: Full details about
extension's ability to control <code>ChromeSetting</code>s
nit: "about *an* extension's" or "about extensions'"

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:67: can grab control over the
setting:
not necessary, see above. It is still good practice warn in case we cannot
control the preference.

[battre, 2012-01-24 12:40:33]

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
File
chrome/common/extensions/docs/examples/api/preferences/enableReferrer/popup.js
(right):

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/examples/api/preferences/enableReferrer/popup.js:6:
var pref = chrome.privacy.websites.referrersEnabled;
If you install this extension, the browser action dialog has a ticked box "Allow
third-party sites to set cookies" but the setting is not active according to
chrome://settings/content. Clicking the check box twice fixes this.

[Mike West, 2012-01-24 13:49:33]

Thanks Dominic.

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
File
chrome/common/extensions/docs/examples/api/preferences/enableReferrer/popup.js
(right):

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/examples/api/preferences/enableReferrer/popup.js:6:
var pref = chrome.privacy.websites.referrersEnabled;
On 2012/01/24 12:40:33, battre wrote:
> If you install this extension, the browser action dialog has a ticked box
"Allow
> third-party sites to set cookies" but the setting is not active according to
> chrome://settings/content. Clicking the check box twice fixes this.

Seems to work for me. Might be a confusion between the phrasing in the extension
("allow") and the phrasing in the content settings ("block"). It's negated, to
make all the settings clearly "xxxEnabled".

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
File chrome/common/extensions/docs/static/privacy.html (right):

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:19: <a href="manifest.html">
manifest</a> to use the API. For example:
On 2012/01/24 12:29:18, battre wrote:
> nit: remove space before manifest

Done.

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:51: deny access to a setting
if the setting is either locked to a specific value
On 2012/01/24 12:29:18, battre wrote:
> technically, it won't deny access but only ignore the written value until it
> becomes the most important (policy is revoked; other extension clears its
> preference)

Gah. Complexity.

Done.

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:58: Full details about
extension's ability to control <code>ChromeSetting</code>s
On 2012/01/24 12:29:18, battre wrote:
> nit: "about *an* extension's" or "about extensions'"

The latter. Done. :)

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/static/privacy.html:67: can grab control over the
setting:
On 2012/01/24 12:29:18, battre wrote:
> not necessary, see above. It is still good practice warn in case we cannot
> control the preference.

Done.

[battre, 2012-01-24 14:06:42]

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
File
chrome/common/extensions/docs/examples/api/preferences/enableReferrer/popup.js
(right):

http://codereview.chromium.org/9244005/diff/21036/chrome/common/extensions/do...
chrome/common/extensions/docs/examples/api/preferences/enableReferrer/popup.js:6:
var pref = chrome.privacy.websites.referrersEnabled;
On 2012/01/24 13:49:33, Mike West (chromium) wrote:
> On 2012/01/24 12:40:33, battre wrote:
> > If you install this extension, the browser action dialog has a ticked box
> "Allow
> > third-party sites to set cookies" but the setting is not active according to
> > chrome://settings/content. Clicking the check box twice fixes this.
> 
> Seems to work for me. Might be a confusion between the phrasing in the
extension
> ("allow") and the phrasing in the content settings ("block"). It's negated, to
> make all the settings clearly "xxxEnabled".

Sorry, let me rephrase this. This is how you can reproduce the problem.

1. Uninstall the extension
2. Reinstall the extension
3. Observe that in chrome://settings/content the entry "Block third-party
cookies and site data" can be toggled by the user.
4. Toggle the checkbox of the browser extension twice.
5. Observe that in chrome://settings/content the entry mentioned above cannot be
toggled by the user any more.

I have to admit that I tested this on a pretty recent ToT, not on your CL.

[jhurwich, 2012-01-24 16:44:48]

"Privacy-related settings" SGTM

On 2012/01/24 12:15:08, Mike West (chromium) wrote:
> On 2012/01/20 16:16:14, Mike West (chromium) wrote:
> > > Aaron, Jordan: The team here has converged on "Privacy-relevant settings."
> for
> > 
> > Sorry, "Privacy-related settings."
> 
> Aaron, Jordan: Friendly ping. If you two are fine with this string, then we're
> good to go.

[commit-bot: I haz the power, 2012-01-24 21:28:36]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mkwst@chromium.org/9244005/27002

[commit-bot: I haz the power, 2012-01-24 21:28:38]

Can't process patch for file
chrome/test/data/extensions/api_test/preference/standard/manifest.json.
File's status is None, patchset upload is incomplete.

[Mike West, 2012-01-24 22:58:16]

On 2012/01/24 21:28:38, I haz the power (commit-bot) wrote:
> Can't process patch for file
> chrome/test/data/extensions/api_test/preference/standard/manifest.json.
> File's status is None, patchset upload is incomplete.

Codereview apparently ate patchset #8.

I spent the last hour or so rebasing and building DumpRenderTree here at home,
and I'll run the result through trybots before landing manually. CQ won't work
regardless because of the zip files. Bleh. :(