LGTM1 On Tue, Jul 25, 2023 at 5:31 PM Paul Jensen <pauljen...@chromium.org> wrote:
> Contact emails > > pauljen...@chromium.org > > Explainer > > > https://github.com/WICG/turtledove/pull/639/files?short_path=d65ba97#diff-d65ba9778fe3af46de3edfce2266b5b035192f8869280ec07179963b81f4e624 > > > https://github.com/WICG/turtledove/pull/486/files?short_path=d65ba97#diff-d65ba9778fe3af46de3edfce2266b5b035192f8869280ec07179963b81f4e624 > > Specification > > https://github.com/WICG/turtledove/pull/711 > > https://github.com/WICG/turtledove/pull/636 > > Summary > > This I2S covers two features extending Protected Audience: > > Recency: > > The “recency” signal for Protected Audience interest groups indicates how > long ago the user was joined to an interest group, which can be a useful > signal when calculating an ad auction bid (e.g. recently expressed interest > likely indicates more interest). Previously we provided this signal in a > strictly bucketed and noised form to buyers’ win reporting function, > reportWin(). <https://github.com/WICG/turtledove/issues/435> This change > additionally exposes this signal to the buyers’ bidding function, > generateBid(). It can be provided without bucketing or noising to > generateBid() like other signals available in that function. In fact, > Protected Audience already allows developers to calculate this signal (e.g. > by storing join time in the interest group), but developers have asked (see > “Web Developers” section below) to have the browser supply it to > generateBid() to ensure it’s calculated identically to the value supplied > to the reporting function, so that models training on the reported data are > compatible with bidding. > > Rounding bids and scores: > > In Protected Audience, the bid price and desirability score pass from > functions that have access to cross-site data (generateBid() and scoreAd()) > to the reporting worklets that have access to first party data (reportWin() > and reportResult()), so to prevent event-level win reports from > facilitating cross-site identity joins, we need to limit this data as much > as possible. This change limits the information in the bid price and > desirability score by rounding them from 64-bit floating-point numbers to > 16-bit floating point numbers. Previously these numbers were not rounded. > > Blink component > > Blink>InterestGroups > <https://bugs.chromium.org/p/chromium/issues/list?q=component:Blink%3EInterestGroups> > > TAG review > > The parent proposal, Protected Audience, is still pending: > https://github.com/w3ctag/design-reviews/issues/723 > > TAG review status > > Pending > > Risks > Interoperability and Compatibility > > Recency: This is unlikely to break existing sites as it’s only adding a > new field to an object the browser provides to Protected Audience bidding > and scoring scripts. > > Rounding: This is unlikely to break existing sites as it’s only clearing > some of the less significant bits of the bid and score values, while not > changing the most significant bits or where the values flow from and to. > > Gecko & WebKit: No signal on parent proposal, Protected Audience. Asked > here <https://github.com/mozilla/standards-positions/issues/770> and here > <https://github.com/WebKit/standards-positions/issues/158>. > > Web developers: > > Recency: Adtechs asked for the recency feature here > <https://github.com/WICG/turtledove/issues/435#issuecomment-1592103934> > as part of the larger ask <https://github.com/WICG/turtledove/issues/435>. > > Rounding: This has been part of Protected Audience’s plan to accomplish > our privacy goals for some time. We haven’t heard significant resistance. > > Debuggability > > These features affect values provided to Protected Audience scripts > (generateBid(), reportResult(), reportWin()) which are debuggable via > Chrome DevTools. > > Will this feature be supported on all six Blink platforms (Windows, Mac, > Linux, Chrome OS, Android, and Android WebView)? > > It will be supported on all platforms that support Protected Audience, so > all but WebView. > > Is this feature fully tested by web-platform-tests > <https://chromium.googlesource.com/chromium/src/+/main/docs/testing/web_platform_tests.md> > ? > > These are in progress and hope to land before M116 stable release. > > Finch feature name > > FledgeRounding > > Requires code in //chrome? > > False > > Estimated milestones > > Shipping on desktop and Android in M116. > > Anticipated spec changes > > None related to these two features. > > Link to entry on the Chrome Platform Status > > https://chromestatus.com/feature/5084137479733248 > > This intent message was generated by Chrome Platform Status > <https://chromestatus.com/>. > > -- > 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 on the web visit > https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CABQTWr%3DfU5ucEnyYgx6Xg5wpmX4q9MOqBryOjLQN17E7ANafrA%40mail.gmail.com > <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CABQTWr%3DfU5ucEnyYgx6Xg5wpmX4q9MOqBryOjLQN17E7ANafrA%40mail.gmail.com?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 on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAOMQ%2Bw8zTzGiCMFfAn1E1269aQrFEEZuJsBKau64pxVMVo8E8Q%40mail.gmail.com.
