css transform causes incorrect position of child sticky element
Reported by
j...@untapt.com,
Jun 16 2017
|
|||||||
Issue descriptionUserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.104 Safari/537.36 Steps to reproduce the problem: View this codepen: https://codepen.io/telekid/pen/NgbamJ?editors=1100 What is the expected behavior? Element A is nested in Element B. Element B is nested in Element C. If Element A's 'position' property is set to sticky, and Element B's 'overflow' property is set to scroll, and Element B's content overflows, and Element C's 'transform' property is set to 'translateX(30px)', then the left edge of Element A should be 30px from the left edge of the document root, and in line with the left edge of Element B. What went wrong? The left edge of Element A is positioned 60px from the left edge of the document root, and 30px from the left edge of Element B. As a rule, the left edge of Element A will always be 2x farther from the left edge of Element B than expected behavior would dictate. Did this work before? N/A Does this work in other browsers? Yes Chrome version: 59.0.3071.104 Channel: stable OS Version: OS X 10.12.4 Flash Version: As a workaround, developers can set the opacity of Element B to any value other than 1. (See the commented-out statement on line 27 of the attached CodePen for an example workaround.)
,
Jun 20 2017
,
Jun 20 2017
Can repro on Mac 58.0.3029.110 (Official Build) (64-bit) Cannot repro on Mac Canary 61.0.3131.0 (Official Build) canary (64-bit) Requesting a reverse bisect to find when the issue was fixed.
,
Jun 21 2017
Tested the issue on Mac retina using chrome M59 #59.0.3071.109 and issue is reproduced and fixed in latest dev and canary channels. Issue is specific to Mac retina. Issue got fixed in M61 #61.0.3125.0 and hence providng the reverse bisect result, Good build: 61.0.3125.0(Revision: 478135). Bad build: 61.0.3123.0 (Revision: 477506). You are probably looking for a change made after 477880 (known good), but no later than 477881 (first known bad). CHANGELOG URL: The script might not always return single CL as suspect as some perf builds might get missing due to failure. https://chromium.googlesource.com/chromium/src/+log/1723af89f9dca5b35dc342ccdf36970fc49b548b..65784430dea11bf2e12e947c77dd08ce2c543c6f From the CL above, assigning the issue to the concern owner @yigu- Could you please merge the fix to M59 stable. Review-Url: https://codereview.chromium.org/2911463002 Thanks!
,
Jun 21 2017
Merging a non-regression fix to stable seems a little bit dangerous. I think it would be better to let the "fix" be tested through canary-beta before shipping to stable so that we can make sure the "fix" won't cause any new bugs. WDYT?
,
Jun 21 2017
,
Jun 21 2017
,
Dec 4 2017
Close the bug as the fix in 61.0.3132.0 has been rolled out to stable on all platforms as of Dec 4. |
|||||||
►
Sign in to add a comment |
|||||||
Comment 1 by ranjitkan@chromium.org
, Jun 19 2017