On Mon, Feb 26, 2024 at 7:36 PM Mike Taylor <[email protected]> wrote:
> Thanks for the feedback Noam - would you mind filing a bug at > crbug.com/new that contains some steps to reproduce the breakage, and > possibly some affected codepaths and report back here? Agree that breaking > Excel is not a great outcome. > That's an understatement.. > On 2/26/24 10:01 AM, Noam Helfman wrote: > > Great to see work is being done to get this standardized! > > However, I think it should not be shipped yet. > > We have done some basic testing of this feature with Excel Online and it > breaks lots of critical user scenarios related to our zoom feature. This > will impact many millions of users and regress a major feature. > > We will need to spend time to investigate if there is a simple workaround > that we can use to address this regression. > > Few questions: > 1. What is the expected timeline to ship this? > > Good question. +Yotam Hacohen <[email protected]> may be able to say more. For now, I see that the feature is still not enabled by default <https://source.chromium.org/chromium/chromium/src/+/main:third_party/blink/renderer/platform/runtime_enabled_features.json5;l=3597?q=StandardizedBrowserZoom%20-f:out&ss=chromium> . > 2. Is there an option to programatically determine if the feature is > enabled? (e.g. would *CSS.supports("zoom")* return *true*? ) > 3. Will there be an option to enable/disable it (e.g. release with OT)? > > It might be a good idea to have an OT that turns this feature off, even if it's only for Excel. (although if we missed this breakage, I wonder what other breakage we may have missed) > > Please do not ship this until we can confirm we have a workaround or the > API is adapted in a way that does not regress existing behavior. > > Thanks, > Noam > On Thursday, February 15, 2024 at 12:18:06 PM UTC+2 Daniel Bratell wrote: > >> Same for me. A proprietary long term CSS property is now fully >> standardized and will be interoperable. This is a win for the web, and >> thank you for all who worked to make it happen! >> >> /Daniel >> On 2024-02-14 18:13, Yoav Weiss (@Shopify) wrote: >> >> Just wanted to say that it's exciting to see this standardized after all >> these years. Given the manual inspection, it seems like shipping this to >> 100% with a killswitch is (hopefully) safe enough! >> >> On Wed, Feb 14, 2024 at 6:11 PM Yoav Weiss (@Shopify) < >> [email protected]> wrote: >> >>> LGTM3 >>> >>> On Wed, Feb 14, 2024 at 6:00 PM Philip Jägenstedt <[email protected]> >>> wrote: >>> >>>> LGTM2 >>>> >>>> On Wed, Feb 14, 2024 at 11:53 PM Daniel Bratell <[email protected]> >>>> wrote: >>>> > >>>> > LGTM1 >>>> > >>>> > /Daniel >>>> > >>>> > On 2024-02-09 20:24, 'Yotam Hacohen' via blink-dev wrote: >>>> > >>>> > >>>> > >>>> > On Thursday, February 8, 2024 at 6:46:00 PM UTC-8 Domenic Denicola >>>> wrote: >>>> > >>>> > On Fri, Feb 9, 2024 at 10:55 AM Yotam Hacohen <[email protected]> >>>> wrote: >>>> > >>>> > Hey Dominic and thanks for the input! >>>> > >>>> > On Sunday, February 4, 2024 at 7:34:53 PM UTC-8 Domenic Denicola >>>> wrote: >>>> > >>>> > It's always exciting to move such an old feature from nonstandard to >>>> standardized! >>>> > >>>> > On Sat, Feb 3, 2024 at 4:18 AM 'Yotam Hacohen' via blink-dev < >>>> [email protected]> wrote: >>>> > >>>> > Contact [email protected] >>>> > >>>> > ExplainerNone >>>> > >>>> > >>>> > FWIW, I think the contents of >>>> https://github.com/w3c/csswg-drafts/pull/9699 and >>>> https://drafts.csswg.org/css-viewport/#zoom-property are probably a >>>> good enough explainer. It might be a good idea to update ChromeStatus to >>>> link to them. >>>> > >>>> > Added those. Thanks! >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > Specificationhttps://github.com/w3c/csswg-drafts/pull/9699 >>>> > >>>> > Design docshttps:// >>>> docs.google.com/document/d/1AcnDShjT-kEuRaMchZPm5uaIgNZ4OiYtM4JI9qiV8Po/edit >>>> > >>>> > Summary >>>> > >>>> > Aligns the existing implementation of the previously non-standard CSS >>>> zoom property to align with the new standard. This changes various JS APIs >>>> to align with the spec (see design doc), change zoom to apply to iframes, >>>> and change it to apply to all inherit all length properties (currently it >>>> only changes inherited font-size) >>>> > >>>> > >>>> > Blink componentBlink>Paint >>>> > >>>> > TAG reviewNone >>>> > >>>> > TAG review statusPending >>>> > >>>> > >>>> > Probably this fits under the first exception here. >>>> > >>>> > >>>> > >>>> > >>>> > Risks >>>> > >>>> > Interoperability and Compatibility >>>> > >>>> > There is web compatibility risk for these changes. However, previous >>>> research indicates broken content due to unexpected changes of the JS APIs >>>> is very unlikely, since: * The changes to the JS API simply change the >>>> coordinate space of the responses, not the syntax or what APIs are >>>> available. * Most pages found during the research didn't appear to use CSS >>>> zoom at all and the ones that did only relied on the visual effect, not JS >>>> APIs. It's possible some pages will be broken by the changes to inherited >>>> properties other than font-size, or applying zoom to sub-frames, but based >>>> on previous research, those are very likely to be minor visual changes that >>>> don't break fundamental user interaction with the site. None of the sites >>>> reviewed contained iframes underneath a zoomed ancestor. We will use direct >>>> outreach to avoid any broken features in Office 365 or the Gmail native >>>> mobile app >>>> > >>>> > >>>> > Can you give more quantitative details on this previous research? >>>> E.g. when you say "most pages", is that 3/5 pages? 99/100? >>>> > >>>> > Sampling pages from the doc, I couldn't find even one example of a >>>> page that uses zoom in a way that will change it's behavior (i.e. - calling >>>> GetBoundingClientRect or GetBoundingRects on an element with CSS zoom). I >>>> also compared those sites visually side by side on a stable version of >>>> chrome and a local version with the planned changes in effect, and couldn't >>>> see any change. >>>> > >>>> > >>>> > This sounds like a good sign, but I'd still appreciate some numbers. >>>> So it's zero out of how many? >>>> > >>>> > I checked the first 15 websites in the list on this doc: >>>> https://docs.google.com/document/d/1cmbXpjAcXAht2ufi7bNKy-rbVNveqaf0UzeYg_DIMNA/edit#heading=h.6sz4u73bikbd >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > Regarding the direct outreach targets you mentioned, are they already >>>> fixed, or do they need more time to update? >>>> > >>>> > We have reached out to the relevant people. >>>> > >>>> > >>>> > So, you have contacted them, but they still need more time to update? >>>> Do you have an estimate for when they will be updated? >>>> > >>>> > We already got a response from the gmail team, and everything is ok >>>> there, we even have a jsfiddle example that shows that the visual aspect >>>> doesn't change for them. Still waiting for a response from the Office 365, >>>> if we don't get a response in the next week we will reach out again for a >>>> better defined timeline. >>>> > >>>> > >>>> > >>>> > What is your rollout plan for this change---straight to 100% with a >>>> killswitch, or a gradual rollout, or...? >>>> > >>>> > Our plan is to go straight to 100% with a killswitch. >>>> > >>>> > >>>> > >>>> > >>>> > Gecko: No signal Filed a standard position request: >>>> https://github.com/mozilla/standards-positions/issues/977 >>>> > >>>> > WebKit: No signal Filed a standard position request: >>>> https://github.com/WebKit/standards-positions/issues/311 >>>> > >>>> > Web developers: Positive ( >>>> https://docs.google.com/document/d/1cmbXpjAcXAht2ufi7bNKy-rbVNveqaf0UzeYg_DIMNA/edit#heading=h.6sz4u73bikbd) >>>> Research collected as part of the previous attempt to remove CSS zoom >>>> demonstrated several use cases. >>>> > >>>> > Other signals: >>>> > >>>> > WebView application risks >>>> > >>>> > See Interoperability and Compatibility above >>>> > >>>> > >>>> > >>>> > Debuggability >>>> > >>>> > None >>>> > >>>> > >>>> > Will this feature be supported on all six Blink platforms (Windows, >>>> Mac, Linux, ChromeOS, Android, and Android WebView)?No >>>> > >>>> > Is this feature fully tested by web-platform-tests?Yes >>>> > >>>> > All JS APIs affected by zoom are tested with the following wpt tests: >>>> https://wpt.fyi/results/css/cssom-view/offsetTop-offsetLeft-with-zoom.html?label=master&label=experimental&aligned&q=cssom-view%2FoffsetTop-offsetLeft-with-zoom.html >>>> https://wpt.fyi/results/css/cssom-view/client-props-zoom.html?label=master&label=experimental&aligned >>>> https://wpt.fyi/results/css/cssom-view/getBoundingClientRect-zoom.html?label=master&label=experimental&aligned >>>> https://wpt.fyi/results/css/cssom-view/getClientRects-zoom.html?label=master&label=experimental&aligned >>>> https://wpt.fyi/results/css/cssom-view/scroll-zoom.html?label=master&label=experimental&aligned >>>> https://wpt.fyi/results/intersection-observer/zoom-scaled-target.html?label=experimental&label=master&aligned >>>> > >>>> > >>>> > Are the non-JS aspects of the API also tested? >>>> > >>>> > Yes, the tests also test the cpp code that is affected. >>>> > >>>> > >>>> > My question was about the visual aspects. Are there any, for example, >>>> reftests, which show that zoom has a visual effect? >>>> > >>>> > Yes, there are reftests for zoom in the wpt folder. The behavior of >>>> many aspects of the zoom are not changed (Especially the visual effect of >>>> CSS zoom on most elements, excluding iframes) and those tests stay the >>>> same. We will also add reftests for iframes with CSS zoom withe the patch >>>> adding those changes to iframes. >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > Flag name on chrome://flagsStandardizedBrowserZoom >>>> > >>>> > Finch feature nameStandardizedBrowserZoom >>>> > >>>> > Requires code in //chrome?False >>>> > >>>> > Sample linkshttps://jsbin.com/wasafateko/edit?html,css,js,output >>>> > >>>> > Estimated milestones >>>> > >>>> > No milestones specified >>>> > >>>> > >>>> > Link to entry on the Chrome Platform Statushttps:// >>>> chromestatus.com/feature/5198254868529152 >>>> > >>>> > This intent message was generated by Chrome Platform Status. >>>> > >>>> > -- >>>> > 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 [email protected]. >>>> > To view this discussion on the web visit >>>> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAAOtuiYKjC9Gt%2BgXwWNT_hJneBMa053RizCX5Xj5p_07CVLXkA%40mail.gmail.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 [email protected]. >>>> > To view this discussion on the web visit >>>> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/1879c472-27c5-46f1-858d-125890807771n%40chromium.org >>>> . >>>> > >>>> > -- >>>> > 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 [email protected]. >>>> > To view this discussion on the web visit >>>> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/7efa033c-2a53-4ac0-8c1a-23c0f9d72f40%40gmail.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 [email protected]. >>>> To view this discussion on the web visit >>>> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAARdPYcAahY9vaoH3Pzths-fM-quQuynoQTvRHuqVMuTo5cMKQ%40mail.gmail.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 [email protected]. > To view this discussion on the web visit > https://groups.google.com/a/chromium.org/d/msgid/blink-dev/9fb533af-50e1-4246-98f5-717bced6797cn%40chromium.org > <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/9fb533af-50e1-4246-98f5-717bced6797cn%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 [email protected]. To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAOmohS%2Bx8XEWVoP8Hb3wf7c8FASaQhs%2BQh-XfJ5HAmA96u32Gg%40mail.gmail.com.
