Nan and I had an offline chat about burn-in risk. Given that, LGTM to
extend from 133 to 135 inclusive.
However, I still have some concerns about successfully turning this off
and hope the team can present a timeline or approximate plan for
expiring the experiment, as an artifact of "progress", should they want
to extend past 135. It may be interesting to consider pausing the
experiment for a few days/1 week to flush out any code that's assuming
navigator.cookieDeprecationLabel /et al./ will exist forever, or moving
to an Origin Trial-based experiment.
On 2/1/25 1:38 PM, Nan Lin wrote:
Hi Mike,
On Sat, Feb 1, 2025 at 12:50 PM Mike Taylor <miketa...@chromium.org>
wrote:
On 1/31/25 4:39 PM, Nan Lin wrote:
On Fri, Jan 31, 2025 at 4:35 PM Mike Taylor
<miketa...@chromium.org> wrote:
On 1/31/25 3:03 PM, Nan Lin wrote:
Hi Mike,
Thanks for the response.
On Fri, Jan 31, 2025 at 11:00 AM Mike Taylor
<miketa...@chromium.org> wrote:
Hey Nan,
On 1/24/25 6:29 PM, Nan Lin wrote:
Contact emails
lin...@chromium.org <mailto:lin...@chromium.org>,
wanderv...@chromium.org <mailto:wanderv...@chromium.org>
Explainer
https://github.com/privacysandbox/tpcd-labeling/blob/main/cookie_deprecation_labeling_explainer.md
<https://github.com/privacysandbox/tpcd-labeling/blob/main/cookie_deprecation_labeling_explainer.md>
https://developer.chrome.com/en/docs/privacy-sandbox/chrome-testing
<https://developer.chrome.com/en/docs/privacy-sandbox/chrome-testing>
Summary
The cookie deprecation labels are useful for developers
to evaluate and optimize deployments of the Privacy
Sandbox APIs prior to any changes in the number of
browsers which support third-party cookies, so we are
asking to extend the current set of labels
<https://developers.google.com/privacy-sandbox/relevance/setup/web/chrome-facilitated-testing>for
three more milestones.
This is a non-standard experiment, so the areas to
demonstrate progress in
https://www.chromium.org/blink/launching-features/#origin-trials
don't cleanly apply. That said, have you received any
useful feedback from developers who are using these labels?
Also, when do you expect this experiment to outlive it's
usefulness?
We've heard from developers using the APIs that the current
implementation of labels remains a useful way to coordinate
while there is traffic where Chrome has disabled third-party
cookies.
Storage Access Headers
<https://groups.google.com/a/chromium.org/g/blink-dev/c/gERgwZfN_-E>will
ship in M133, allowing developers to determine if they
have access to unpartitioned cookies via the
Sec-Fetch-Storage-Access header instead of labels.
Extending this experiment gives time for developers to
make use of this upcoming API as a signal for cookie
access.
My understanding of this experiment was to allow for A/B
testing analysis - but it sounds like it can be replaced
with a signal of "has 3P cookies" (like
navigator.cookieEnabled). Does that fully satisfy the
needs of developers trying to understand PS APIs? Or do
I misunderstand?
The goal of this experiment is to allow ad-techs to run
server side A/B testing from the browser provided treatment
and control groups, and evaluate the impact of third party
cookie phase out.
It allows ad-techs to continue to test Privacy Sandbox APIs
on some traffic without population issues.
Thanks - perhaps my question wasn't super clear. Does
Sec-Fetch-Storage-Access fully replace this experiment to
allow for A/B testing by ad tech companies?
Thanks Mike for clarifying the question. I don't think
Sec-Fetch-Storage-Access can fully replace this experiment at
this stage, as it doesn't tell whether the third party cookies
are disabled by Chrome or not.
We would still need these labels to allow downstream A/B testing
on the traffic slice with and without third party cookies.
I see, thanks. I would still like to understand the plan
for/answer to my first question:
> Also, when do you expect this experiment to outlive it's usefulness?
I can imagine some ad techs would be happy to receive the labels
forever, if they have some utility today.
In terms of burn-in risk, do we have any use counter or UMA
metrics to understand its usage (since it's been in the wild for
more than a year at this point)?
The labels are only sent on the traffic slice where Chrome enables
mode A (label only) and mode B (third party disabled) experiments, and
overall this is <8.5% of browsers (these labels are also subject to a
number of eligibility requirements).
The labels will not be sent forever, and will be deprecated when
Chrome stops the experiments and there is no traffic where Chrome
disables third party cookies.
Hope that clarifies!
Link to “Intent to Experiment” blink-dev discussion
https://groups.google.com/a/chromium.org/g/blink-dev/c/0_dR-ffA2LA/m/ZgmMhK-XAQAJ
<https://groups.google.com/a/chromium.org/g/blink-dev/c/0_dR-ffA2LA/m/ZgmMhK-XAQAJ>
https://groups.google.com/a/chromium.org/g/blink-dev/c/3escBQGtIpM/m/ntcytva5BgAJ
<https://groups.google.com/a/chromium.org/g/blink-dev/c/3escBQGtIpM/m/ntcytva5BgAJ>
https://groups.google.com/u/1/a/chromium.org/g/blink-dev/c/v3PiIzm1M-Y
<https://groups.google.com/u/1/a/chromium.org/g/blink-dev/c/v3PiIzm1M-Y>
Goals for experimentation
Continued deployment and evaluation of Privacy Sandbox
Ads APIs.
Experimental timeline
This feature was previously approved to run up until
Chrome 132.
We would like to extend this for Chrome 133 through
135, inclusive.
Any risks when the experiment finishes?
Minimal, the cookie deprecation labels are only
available for a subset of users and must be requested.
Reason this experiment is being extended
We have received feedback that these labels are useful
for ad tech companies to evaluate and optimize the APIs
in preparation for changes to third party cookie
availability.
Ongoing technical constraints
None
Will this feature be supported on all five Blink
platforms supported by Origin Trials (Windows, Mac,
Linux, Chrome OS, and Android)?
No, not supported on webview.
Link to entry on the feature dashboard
https://chromestatus.com/feature/5189079788683264
<https://chromestatus.com/feature/5189079788683264>
--
You received this message because you are subscribed to
the Google Groups "blink-dev" group.
To unsubscribe from this group and stop receiving
emails from it, send an email to
blink-dev+unsubscr...@chromium.org.
To view this discussion visit
https://groups.google.com/a/chromium.org/d/msgid/blink-dev/849254f2-1020-46d2-b42a-b7dd02db5e85n%40chromium.org
<https://groups.google.com/a/chromium.org/d/msgid/blink-dev/849254f2-1020-46d2-b42a-b7dd02db5e85n%40chromium.org?utm_medium=email&utm_source=footer>.
--
You received this message because you are subscribed to the Google Groups
"blink-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to blink-dev+unsubscr...@chromium.org.
To view this discussion visit
https://groups.google.com/a/chromium.org/d/msgid/blink-dev/2bb4aaf6-4ac5-4995-b979-5218936c86c7%40chromium.org.
