Everyone, Thanks for the feedback. Let's take a step back here. The issue mentioned here is one that *i* opened asking whether we should expose this to iframes or not. I wanted to hear feedback from the community: we heard they want it exposed. There isn't an ongoing discussion.
Now the question is whether we want to put that behind a permission policy. Guarding the JS API behind a permission policy is trivial, the CSS part is not. I'm not against doing the work but we need to take a step back and evaluate why we would guard it behind a permission policy in the first place. The API only exposes two values: continuous and folded. Continuous is returned for *every* devices not a foldable (on all platforms) and when a foldable device is used flat. There is no fingerprinting value in this. When the device is folded there is a possibility of ephemeral fingerprinting (maybe?) where you could determine that it's a folded device across sessions. However, the posture can change (user generated) across browsing session defeating any tracking. What one could determine is if the device is a foldable and folded, however it's easy to determine that a device is a foldable just by using the viewport width/height and various other information. Knowing that the device is folded has no real value per se, at least I couldn't find any. Finally requiring permission policy was not required by the TAG and PING during their respective reviews. We did discuss the issue with PING but after exposing the fact that the CSS would still leak the information and there wasn't any official CSS recommendation how to gate a CSS API then we did not get opposition in the review. On Wed, Oct 9, 2024 at 5:05 AM Yoav Weiss (@Shopify) <yoavwe...@chromium.org> wrote: > +1 to either resolving this before shipping, or shipping a restricted > version (not exposing this to cross-origin iframes at all, or adding a > default `self` permission policy), which we'd be able to relax in the > future if needed. > > As there seems to be active debate happening on the issue, maybe we should > let this discussion reach its conclusions? > > On Wednesday, October 9, 2024 at 3:28:18 AM UTC+2 Domenic Denicola wrote: > >> To be clear, I'm OK if the relevant standards body (i.e. the W3C Devices >> and Sensors WG) has decided that this does not require a permissions >> policy. But I would like it to be resolved one way or another before we >> approve this for shipping. Right now it is listed as an open issue in the >> spec, and it's one that will be hard to change after shipping, so per the >> "Anticipated spec changes" of our I2S template, I'd like to get that >> resolved. >> >> On Wed, Oct 9, 2024 at 3:14 AM Alex Russell <slightly...@chromium.org> >> wrote: >> >>> It's unclear to me that this should have a permission. >>> >>> LGTM1 >>> >>> On Mon, Oct 7, 2024, 7:19 AM Alexis Menard <alexis.men...@intel.com> >>> wrote: >>> >>>> >>>> >>>> On Mon, Oct 7, 2024 at 9:31 AM Ian Clelland <iclell...@chromium.org> >>>> wrote: >>>> >>>>> >>>>> >>>>> On Mon, Oct 7, 2024 at 7:36 AM Alexis Menard <alexis.men...@intel.com> >>>>> wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> Thanks for your approval. >>>>>> >>>>>> I can confirm that the API is exposed to iframe through JS and CSS. >>>>>> >>>>>> Concerning your suggestion I agree that we could put this behind a >>>>>> permission policy but unfortunately if the intent is to limit potential >>>>>> ephemeral fingerprinting then it will not help at all. In Chromium it's >>>>>> easy to gate the JS API behind a permission policy but there is no prior >>>>>> art of a CSS API being gated behind a permission policy (I may be wrong). >>>>>> So the iframe CSS code will still be parsed which would not impede >>>>>> accessing the posture. Finally one can just use JavaScript to query the >>>>>> posture using `matchMedia` so in order for this whole permission to truly >>>>>> block we would need to patch the CSS engine deep down. >>>>>> >>>>> >>>>> We've never shipped any CSS gating based on permissions policy, but we >>>>> have experimented with it; in particular, we've released experimental >>>>> policies to restrict the use of animations on properties which affect >>>>> layout, and to restrict the values which can be used for the font-display >>>>> property. These have since been removed from the code, as we're not >>>>> pursuing those anymore, but the idea of controlling the CSS engine with >>>>> permissions policy has been tried. >>>>> >>>>> I'm not sure if the fact that this API is exposed through media >>>>> queries makes this more complex, but from a spec perspective, as long as >>>>> you can describe the behaviour in terms of what the current document is >>>>> "allowed to use", then you should be able to express the right constraints >>>>> to use permissions policy. >>>>> >>>> >>>> Interesting. I'll try to dig the CL just out of curiosity. >>>> >>>> >>>>> >>>>> Ian >>>>> >>>>> >>>>>> I had this discussion with the PING and they agreed that we don't >>>>>> have any mechanism in place even CSS to support such a thing. There is a >>>>>> discussion which started few weeks ago between the PING and CSS WG. I >>>>>> believe that in the future this use case could come up for some other >>>>>> APIs >>>>>> especially when things are exposed through env variables. So unless there >>>>>> is some idea of a spec or update to permission policy spec I'm not sure >>>>>> if >>>>>> we should start modifying the CSS engine deeply. >>>>>> >>>>>> Coming back to this API, to be honest I think the fingerprinting is >>>>>> very low risk, ephemeral and is going to be less and less relevant as >>>>>> more >>>>>> and more users are using foldables especially in the folded posture >>>>>> (remember that any other device including desktop returns the continuous >>>>>> posture). >>>>>> >>>>>> Thanks. >>>>>> >>>>>> >>>>>> On Mon, Oct 7, 2024, 12:24 AM Domenic Denicola <dome...@chromium.org> >>>>>> wrote: >>>>>> >>>>>>> This looks like a really solid spec that has benefited from years of >>>>>>> iteration and had good TAG review discussion. The fact that you >>>>>>> specified >>>>>>> and are working on WebDriver hooks to emulate posture changes during >>>>>>> testing, and added DevTools integration, are more great signs of >>>>>>> maturity. >>>>>>> I'm excited to approve this. >>>>>>> >>>>>>> The only blocker is that >>>>>>> https://github.com/w3c/device-posture/issues/111 remains open and >>>>>>> changing that after shipping would be a significant change. It sounds >>>>>>> like >>>>>>> your current plan is to expose this information across iframes. Can you >>>>>>> confirm? If so, are you ready to close that issue and lock in the >>>>>>> current >>>>>>> state? >>>>>>> >>>>>>> A more conservative plan would be to not expose the information >>>>>>> across cross-origin iframes. You could then loosen that in the future, >>>>>>> probably by introducing a permissions policy: either with a default >>>>>>> allowlist of '*' to get the current behavior (but allow top frames to >>>>>>> restrict), or a default allowlist of 'self' to keep the restriction by >>>>>>> default (but allow top frames to share). Absent strong use cases for >>>>>>> sharing cross-origin by default, that would be my suggestion. >>>>>>> >>>>>>> On Thu, Oct 3, 2024 at 11:42 PM Alexis Menard < >>>>>>> alexis.men...@intel.com> wrote: >>>>>>> >>>>>>>> Contact emails alexis.men...@intel.com >>>>>>>> >>>>>>>> Explainer https://github.com/w3c/device-posture >>>>>>>> https://www.w3.org/TR/device-posture/#introduction >>>>>>>> >>>>>>>> Specification https://www.w3.org/TR/device-posture >>>>>>>> >>>>>>>> Summary >>>>>>>> >>>>>>>> This API helps developers to detect the current posture of a >>>>>>>> foldable device. The device posture is the physical position in which a >>>>>>>> device holds which may be derived from sensors in addition to the >>>>>>>> angle. >>>>>>>> From enhancing the usability of a website by avoiding the area of a >>>>>>>> fold, >>>>>>>> to enabling innovative use cases for the web, knowing the posture of a >>>>>>>> device can help developers tailor their content to different devices. >>>>>>>> Content can be consumed and browsed even when the device is not flat, >>>>>>>> in >>>>>>>> which case the developer might want to provide a different layout for >>>>>>>> it >>>>>>>> depending on the posture state in which the device is being used. >>>>>>>> >>>>>>>> >>>>>>>> Blink component Blink>FoldableAPIs >>>>>>>> <https://bugs.chromium.org/p/chromium/issues/list?q=component:Blink%3EFoldableAPIs> >>>>>>>> >>>>>>>> TAG review https://github.com/w3ctag/design-reviews/issues/575 >>>>>>>> >>>>>>>> TAG review status Issues addressed >>>>>>>> >>>>>>>> Risks >>>>>>>> >>>>>>>> >>>>>>>> Interoperability and Compatibility >>>>>>>> >>>>>>>> None >>>>>>>> >>>>>>>> >>>>>>>> *Gecko*: No signal ( >>>>>>>> https://github.com/mozilla/standards-positions/issues/882) >>>>>>>> >>>>>>>> *WebKit*: No signal ( >>>>>>>> https://github.com/WebKit/standards-positions/issues/328) >>>>>>>> >>>>>>>> *Web developers*: >>>>>>>> >>>>>>>> https://github.com/w3c/device-posture/issues/111#issuecomment-2363251667 >>>>>>>> >>>>>>>> *Other signals*: >>>>>>>> >>>>>>>> WebView application risks >>>>>>>> >>>>>>>> Does this intent deprecate or change behavior of existing APIs, >>>>>>>> such that it has potentially high risk for Android WebView-based >>>>>>>> applications? >>>>>>>> >>>>>>>> Feature is disabled on WebView for now. See >>>>>>>> https://issues.chromium.org/issues/335314107 for more details. >>>>>>>> >>>>>>>> >>>>>>>> Debuggability >>>>>>>> >>>>>>>> Besides the usual DevTools debugging of the CSS and JavaScript API, >>>>>>>> a specific device has been added into the Device Emulation mode. >>>>>>>> >>>>>>>> >>>>>>>> Will this feature be supported on all six Blink platforms (Windows, >>>>>>>> Mac, Linux, ChromeOS, Android, and Android WebView)? Yes >>>>>>>> >>>>>>>> The API will work on all the platforms but only Android and Windows >>>>>>>> will return posture information (other platforms do not have this >>>>>>>> category >>>>>>>> of devices) >>>>>>>> >>>>>>>> >>>>>>>> Is this feature fully tested by web-platform-tests >>>>>>>> <https://chromium.googlesource.com/chromium/src/+/main/docs/testing/web_platform_tests.md> >>>>>>>> ? Yes >>>>>>>> >>>>>>>> The tests aren't complete yet because we need integration with >>>>>>>> WebDriver to emulate posture changes. It's being worked on. >>>>>>>> https://github.com/web-platform-tests/wpt/tree/master/device-posture >>>>>>>> >>>>>>>> >>>>>>>> Flag name on chrome://flags device-posture >>>>>>>> >>>>>>>> Finch feature name kDevicePosture >>>>>>>> >>>>>>>> Requires code in //chrome? False >>>>>>>> >>>>>>>> Tracking bug >>>>>>>> https://bugs.chromium.org/p/chromium/issues/detail?id=1066842 >>>>>>>> >>>>>>>> Sample links >>>>>>>> https://github.com/foldable-devices >>>>>>>> >>>>>>>> Estimated milestones >>>>>>>> Shipping on desktop 131 >>>>>>>> Origin trial desktop first 125 >>>>>>>> Origin trial desktop last 128 >>>>>>>> DevTrial on desktop 95 >>>>>>>> Shipping on Android 131 >>>>>>>> Origin trial Android first 125 >>>>>>>> Origin trial Android last 128 >>>>>>>> DevTrial on Android 123 >>>>>>>> >>>>>>>> Anticipated spec changes >>>>>>>> >>>>>>>> Open questions about a feature may be a source of future web compat >>>>>>>> or interop issues. Please list open issues (e.g. links to known github >>>>>>>> issues in the project for the feature specification) whose resolution >>>>>>>> may >>>>>>>> introduce web compat/interop risk (e.g., changing to naming or >>>>>>>> structure of >>>>>>>> the API in a non-backward-compatible way). >>>>>>>> None >>>>>>>> >>>>>>>> Link to entry on the Chrome Platform Status >>>>>>>> https://chromestatus.com/feature/5185813744975872?gate=6219681092599808 >>>>>>>> >>>>>>>> Links to previous Intent discussions Intent to Prototype: >>>>>>>> https://groups.google.com/a/chromium.org/g/blink-dev/c/prHGPxF62i4 >>>>>>>> Intent to Experiment: >>>>>>>> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/8c244153-79c4-483e-8449-4aca14b35636%40chromium.org >>>>>>>> >>>>>>>> >>>>>>>> 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/540e383c-1e1c-4918-9f10-c3fb2dd9bc19%40intel.com >>>>>>>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/540e383c-1e1c-4918-9f10-c3fb2dd9bc19%40intel.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/CAM0wra_U0%3DqYJnDGBM8Zm-yLh7XNT1tA1uKt1a6VzuDBHBdDYA%40mail.gmail.com >>>>>>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAM0wra_U0%3DqYJnDGBM8Zm-yLh7XNT1tA1uKt1a6VzuDBHBdDYA%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/CAOaK9AntwL_dhXaSvEHVAfoisf4fexB_tNTidO9BjqiWUxM2vQ%40mail.gmail.com >>>>>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAOaK9AntwL_dhXaSvEHVAfoisf4fexB_tNTidO9BjqiWUxM2vQ%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/CAK_TSXKv4q4Zj%2B-iDr%3DEbdENuZbdpFqxaaNrqXn6ZgdYX%2BGEXw%40mail.gmail.com >>>>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAK_TSXKv4q4Zj%2B-iDr%3DEbdENuZbdpFqxaaNrqXn6ZgdYX%2BGEXw%40mail.gmail.com?utm_medium=email&utm_source=footer> >>>>> . >>>>> >>>> >>>> >>>> -- >>>> Alexis Menard >>>> Software Engineer @ Intel >>>> >>>> -- >>>> 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/CAOaK9Am-Bas35FSfRbiFBcihOtrHYMMi6J_z7qfyjcMa8VQAqg%40mail.gmail.com >>>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAOaK9Am-Bas35FSfRbiFBcihOtrHYMMi6J_z7qfyjcMa8VQAqg%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/a7040733-9c2b-49dd-a638-613d17c5b91dn%40chromium.org > <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/a7040733-9c2b-49dd-a638-613d17c5b91dn%40chromium.org?utm_medium=email&utm_source=footer> > . > -- Alexis Menard Software Engineer @ Intel -- 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/CAOaK9Akdk042qBj19K8YZ70aRGS6RbkDTrecW8Xxv2On83%3D6og%40mail.gmail.com.
