We're experiencing an issue where, when we scroll on our site, the scroll position is reset every time a user clicks or hovers over a clickable link. We were able to track down the source of the issue and it seems that setting the body height of a page to 100vh causes this issue. The issue is exclusive to Safari and WebKit and does not occur in other browsers. A demo of this behaviour is available on https://openmoji.org/about.html A further discussion on this issue is available here: https://github.com/hfg-gmuend/openmoji-website/issues/15 https://github.com/hfg-gmuend/openmoji-website/pull/18
<rdar://problem/50130566>
Steps: 1. Open http://openmoji.org/about.html 2. Scroll down 3. Hover over a link (like Twitter Icon below a profile) The site scrolls back to the top However, I can't reproduce; maybe the site was fixed? Can we get a version of the site that shows has the bug?
The site linked is the one causing issues, it hasn't been updated. I'm able to reproduce this issue on multiple machines, I just checked. Here's another URL you could try: https://hfg-gmuend.github.io/openmoji-website/about.html If unavailable, this URL might work: https://hfg-gmuend.github.io/openmoji-website/about Demo Video of the issue: https://youtu.be/noNbdwfXuIQ
This might have been fixed. Could you try Safari Tech Preview (https://developer.apple.com/safari/download/)?
This issue seems fixed in the latest STP. When is a public update for Safari with a release of this bugfix expected? Any ideas for a temporary workaround until that bugfix is published or what might cause this issue? Thx for your help.
This bug persists, I've tested it on multiple Apple devices running safari, as well as the Epiphany browser (which also uses Webkit). I have made a demo here: http://jsfiddle.net/rnzt02cL/embedded/html,css,result and highlighted the lines which contribute to the bug.
I've found more properties that cause the bug and added the to the fiddle: http://jsfiddle.net/totalolage/rnzt02cL/11/embedded/html,css,result
Possibly related to bug 201528.
Filip, what browser/OS are you testing on? The test uses hover, so I assume this is not iOS? I can't reproduce. Pretty sure this is fixed.
(In reply to Simon Fraser (smfr) from comment #9) > Filip, what browser/OS are you testing on? The test uses hover, so I assume > this is not iOS? > > I can't reproduce. Pretty sure this is fixed. I can't reproduce either anymore. I assume it's been fixed.