Transitions start after external stylesheet loads when a form control exists
Reported by
ad...@cleanpaw.com,
Apr 8 2017
|
||||||||||||||||||||
Issue descriptionUserAgent: Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/57.0.2987.133 Safari/537.36 Steps to reproduce the problem: 1. Run the HTML file in Chrome or Opera 2. 3. What is the expected behavior? When the page loads the text grows from small to large. What went wrong? The element <input type="checkbox"> conflicts with the line "transition: .5s;" and causes the page to render incorrectly. If you remove the <input> element the text renders normally. This is valid HTML. In addition you can eliminate the bug if you remove the comments around the <style> tag in the HTML file and use local CSS. The CSS inside the <style> tag is identical to what is found in the external CSS file. The bug only occurs if the CSS is external. Did this work before? N/A Does this work in other browsers? No This bug occurs in Chrome and Opera Chrome version: 57.0.2987.133 Channel: stable OS Version: 10.0 Flash Version: I discovered that if <div class="trainwreck"><a href="#">link.com</a></div> is placed inside a flexbox the error does not occur.
,
Apr 12 2017
Able to reproduce the issue on Windows 10, mac 10.12.4 and Ubuntu 14.04 using chrome reported version #57.0.2987.133 and latest canary #59.0.3068.1. Bisect Information: ===================== Good build: 47.0.2494.0 Revision(345501) Bad Build : 47.0.2495.0 Revision(345750) Change Log URL: https://chromium.googlesource.com/chromium/src/+log/3a55bcecda30c5e3990f1dd783505acde0cce9ba..7e69b9af78de854121bb02c91e7a41526e9c141e The issue is broken in M-47. Hence, providing blink change log which is as follows: Blink Change Log: https://chromium.googlesource.com/chromium/blink/+log/04b637a..8122aac From the above change log suspecting below change Review url: https://codereview.chromium.org/1314923005 cbiesinger@ - Could you please check whether this is caused with respect to your change, if not please help us in assigning it to the right owner. Thanks...!!
,
Apr 13 2017
I highly doubt that https://codereview.chromium.org/1314923005 is the cause of this bug. On inspection this is due to the stylesheet loading after the <a> element has had its style computed causing the next style computation to trigger the transition to start. The <input type="checkbox"> element must be forcing an early style recalc, removing it from the page causes the bug to go away. Input team: Can this forced style recalc be avoided?
,
Apr 13 2017
Expected result: No font-size animation Actual result with Chrome: font-size grows Right? This is due to Autofill code. 3 libblink_web.dylib 0x000000013cca3d8a blink::WebElement::HasNonEmptyLayoutSize() const + 234 4 libchrome_dll.dylib 0x00000001218d5c12 autofill::form_util::IsWebElementVisible(blink::WebElement const&) + 34 5 libchrome_dll.dylib 0x000000012193a682 autofill::(anonymous namespace)::GetPasswordForm(...) + 1490 6 libchrome_dll.dylib 0x000000012193e7e9 autofill::CreatePasswordFormFromUnownedInputElements(...) + 1177 7 libchrome_dll.dylib 0x000000012190f7d7 autofill::PasswordAutofillAgent::SendPasswordForms(bool) + 3735 8 libchrome_dll.dylib 0x0000000121911569 autofill::PasswordAutofillAgent::DidFinishDocumentLoad() + 57 9 libcontent.dylib 0x0000000130453af4 content::RenderFrameImpl::DidFinishDocumentLoad(blink::WebLocalFrame*) + 980 10 libblink_web.dylib 0x000000013cc01332 blink::LocalFrameClientImpl::DispatchDidFinishDocumentLoad() + 162 3 libblink_web.dylib 0x000000013cd2b30f blink::WebNode::IsFocusable() const + 271 4 libchrome_dll.dylib 0x00000001218d5c26 autofill::form_util::IsWebElementVisible(blink::WebElement const&) + 54 5 libchrome_dll.dylib 0x00000001218d5a40 autofill::form_util::IsSomeControlElementVisible(blink::WebVector<blink::WebFormControlElement> const&) + 176 6 libchrome_dll.dylib 0x000000012190f6f2 autofill::PasswordAutofillAgent::SendPasswordForms(bool) + 3506 7 libchrome_dll.dylib 0x00000001219115ba autofill::PasswordAutofillAgent::DidFinishLoad() + 26 8 libcontent.dylib 0x0000000130454da2 content::RenderFrameImpl::DidFinishLoad(blink::WebLocalFrame*) + 1122 9 libblink_web.dylib 0x000000013cd0b79c blink::WebLocalFrameImpl::DidFinish() + 124
,
Apr 13 2017
,
Apr 13 2017
Related/duplicate - Issue 676158 Issue 693088
,
Apr 13 2017
,
Apr 14 2017
,
Apr 14 2017
,
Apr 21 2017
This seems related to rendering, not to Chrome's support with filling forms.
,
Apr 24 2017
vabr@, this is an issue of autofill code. We should not call WebElement::HasNonEmptyLayoutSize() before the first layout.
,
Apr 24 2017
Could this have been introduced by https://codereview.chromium.org/2769023003 ?
,
Apr 25 2017
> Could this have been introduced by https://codereview.chromium.org/2769023003 ? Probably no. This issue is not a recent regression.
,
Apr 26 2017
HasNonEmptyLayoutSize is used to check visibility when the layout is forbidden (https://cs.chromium.org/chromium/src/components/autofill/content/renderer/form_autofill_util.cc?rcl=2369b7f3dec2f78a760ed52bf64badb2240a1968&l=1332). But HasNonEmptyLayoutSize actually raises layout. It is legacy code. tkent@: could you recommend any function that gives approximation of visibility when the layout is forbidden?
,
Apr 27 2017
> tkent@: could you recommend any function that gives approximation of visibility when the layout is forbidden? No such way essentially. Can we postpone anything until the first layout?
,
Apr 27 2017
Maybe we can just return true or false for all elements. I need to discuss it with the team.
,
May 2 2017
(Adding Hotlist-Polish for password manager triaging -- this is a crash, and that is user visible. We use Hotlist-Polish for user visible bugs.)
,
May 12 2017
,
Jun 13 2017
,
Oct 19 2017
Downgrading priority, because M59 has sailed a long ago and this apparently did not need an escalatation.
,
May 1 2018
,
May 7 2018
,
Nov 7
Looks similar/duplicate of issue 332189.
,
Nov 29
vabr going hobby only -> reducing involvement. Please contact me directly in urgent matters. |
||||||||||||||||||||
►
Sign in to add a comment |
||||||||||||||||||||
Comment 1 by ad...@cleanpaw.com
, Apr 8 2017