Hi all,

Is there a way to remove this option yet? The usability is getting worse
for me. Now the ugly highlighting is sticking around for longer.

I don't need Chrome to randomly highlight an arbitrary sentence of some
websites in a different colour. It's distracting and annoying. I'd like to
turn it off.

Kind regards,
Adam


On Tue, 4 May 2021, 20:40 Adam Semenenko, <[email protected]> wrote:

> Hi all,
>
> Do you know if there's a consistent and easy way to disable this yet? It's
> really distracting for me. When I google something and click on a result, I
> like consistent behaviour and to see the whole page from the start. I feel
> disrupted when I'm randomly dropped into the middle of a page with a garish
> colour jumping out at me.
>
> Cheers,
> Adam
>
> On Mon, 26 Apr 2021 at 21:54, 'Grant Wang' via blink-dev <
> [email protected]> wrote:
>
>> Hi all,
>>
>> It’s been roughly nine months since we first utilized Scroll To Text
>> Fragment in featured snippets in Google Search. In that time, we’ve seen
>> that Scroll To Text Fragment links help us towards our goal to get users
>> the information they need.  In particular:
>>
>>    1. We find that Scroll To Text Fragment links increase engagement --
>>    users are less likely to visit a page and then quickly hit the back 
>> button,
>>    because they can more readily understand how relevant the page is to their
>>    search after arriving at the page.
>>
>>    2. In user surveys, we find that users prefer being scrolled to the
>>    relevant section of a page that’s in a featured snippet. Users who were
>>    scrolled to the relevant section preferred the experience at a rate of 
>> 2:1;
>>    even users who were not scrolled in the control group preferred the option
>>    of being scrolled to the relevant section at the same 2:1 rate.
>>
>> Besides their usage on Google Search, we’ve noticed scroll to text
>> fragments links during our crawls of the web.  One of the best use cases
>> has been wikipedia citations.  For instance, citation 9
>> <https://en.wikipedia.org/wiki/Melbourne_Cup_%28greyhounds%29#:~:text=%22How%20the%20Cup%20Was%20Won%22.%20Sandown%20Greyhounds.%20Retrieved%2017%20March%202021.>
>> on this page: https://en.wikipedia.org/wiki/Melbourne_Cup_(greyhounds)
>> provides the detailed attribution to the fastest-ever time at the Melbourne
>> Cup.
>>
>> Cheers,
>> Grant
>> On Thursday, December 12, 2019 at 12:24:40 PM UTC-8 [email protected]
>> wrote:
>>
>>> LGTM4
>>>
>>>
>>> On Thursday, December 12, 2019 at 12:17:49 PM UTC-8, Daniel Bratell
>>> wrote:
>>>>
>>>> LGTM3
>>>>
>>>> /Daniel
>>>>
>>>>
>>>> On Thursday, 12 December 2019 19:45:38 UTC+1, Chris Harrelson wrote:
>>>>>
>>>>> LGTM2
>>>>>
>>>>> On Wed, Dec 11, 2019 at 10:27 PM Yoav Weiss <[email protected]> wrote:
>>>>>
>>>>>> LGTM1
>>>>>>
>>>>>>
>>>>>> On Wed, Dec 11, 2019, 22:03 Nick Burris <[email protected]> wrote:
>>>>>>
>>>>>>> Hi all,
>>>>>>>
>>>>>>> We feel that we're now in good shape for shipping. We have addressed
>>>>>>> all of the shipping blockers that I listed in my previous email, and the
>>>>>>> corresponding implementation changes have landed in Chrome. We're still
>>>>>>> continuing to make improvements to the spec, functionality, and web
>>>>>>> platform tests but we consider the outstanding issues to be minor and
>>>>>>> wouldn't have an effect on interop, so we don't believe they're
>>>>>>> ship-blocking.
>>>>>>>
>>>>>>> We have received positive signal on the feature from Safari, thank
>>>>>>> you Maciej for the reply on webkit-dev
>>>>>>> <https://lists.webkit.org/pipermail/webkit-dev/2019-December/030996.html>!
>>>>>>> Note that we actually do have feature detectability specified 
>>>>>>> implemented
>>>>>>> per my reply
>>>>>>> <https://lists.webkit.org/pipermail/webkit-dev/2019-December/030998.html>.
>>>>>>> My apologies this was not mentioned in the initial intent to ship 
>>>>>>> email, it
>>>>>>> developed a few emails down the line.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Nick
>>>>>>>
>>>>>>> On Thursday, October 31, 2019 at 11:50:21 AM UTC-4, Nick Burris
>>>>>>> wrote:
>>>>>>>>
>>>>>>>> Thanks so much for the detailed feedback! Here's a specific list of
>>>>>>>> blockers, with some comments inline.
>>>>>>>>
>>>>>>>> Specification issues
>>>>>>>>
>>>>>>>>    - #64 <https://github.com/WICG/ScrollToTextFragment/issues/64>
>>>>>>>>    - Prevent invocation from popup
>>>>>>>>    - #66 <https://github.com/WICG/ScrollToTextFragment/issues/66>
>>>>>>>>    - Clarify how scroll to fragment is performed
>>>>>>>>
>>>>>>>> Web platform test cases
>>>>>>>>
>>>>>>>>    - Security restrictions
>>>>>>>>    
>>>>>>>> <https://wicg.github.io/ScrollToTextFragment/#should-allow-text-fragment>
>>>>>>>>    - Setting window.location.fragmentDirective has no effect
>>>>>>>>    - All combinations of optional parameters in text directive
>>>>>>>>    - Matching TextMatchChar
>>>>>>>>    <https://wicg.github.io/ScrollToTextFragment/#textmatchchar>s
>>>>>>>>    and PercentEncodedChar
>>>>>>>>    <https://wicg.github.io/ScrollToTextFragment/#percentencodedchar>s
>>>>>>>>    (in particular the syntactical characters ‘,’ and ‘-’) including 
>>>>>>>> non-ASCII
>>>>>>>>    - Multiple matches in the page (currently we only test 0 or 1
>>>>>>>>    match)
>>>>>>>>    - Cross-whitespace/node matching (i.e. context terms and match
>>>>>>>>    terms can be separated by whitespace and node boundaries)
>>>>>>>>    - Test matching hidden and shadow DOM
>>>>>>>>    - Test horizontal scroll into view
>>>>>>>>
>>>>>>>>
>>>>>>>> On Thursday, October 31, 2019 at 10:17:56 AM UTC-4, Frédéric Wang
>>>>>>>> wrote:
>>>>>>>>>
>>>>>>>>> On 30/10/2019 15:52, Yoav Weiss wrote:
>>>>>>>>>
>>>>>>>>> This intent received a lot of feedback, but some of it more
>>>>>>>>> relevant to the general Blink process in general than to this intent
>>>>>>>>> specifically. So, let me try to sum up where I believe things are and 
>>>>>>>>> what
>>>>>>>>> is and isn't blocking this intent from my perspective.
>>>>>>>>>
>>>>>>>>> While the original intent could have done a better job at
>>>>>>>>> expressing the outreach efforts done, and potentially a better job 
>>>>>>>>> reaching
>>>>>>>>> out to WebKit folks, that *should not block* the current intent.
>>>>>>>>> Official signals from other vendors would be most welcome, but I 
>>>>>>>>> would not
>>>>>>>>> block the intent on getting them. (The Blink process needs to 
>>>>>>>>> establish the
>>>>>>>>> best ways to get feedback from other vendors in reasonable 
>>>>>>>>> timeframes. That
>>>>>>>>> discussion is beyond the scope of this intent)
>>>>>>>>>
>>>>>>>>> A list of blockers for this intent from my perspective:
>>>>>>>>>
>>>>>>>>>    - Anne's security concern
>>>>>>>>>    
>>>>>>>>> <https://github.com/w3ctag/design-reviews/issues/392#issuecomment-510855073>
>>>>>>>>>  seems
>>>>>>>>>    like something we should address in spec. Even if Chrome security 
>>>>>>>>> folks
>>>>>>>>>    don't consider this a blocking issue, assuming Mozilla does, that 
>>>>>>>>> would get
>>>>>>>>>    in their way if they wished to follow us.
>>>>>>>>>    - Daniel's feedback on augmenting the Privacy & Security
>>>>>>>>>    section with feedback from the Chrome security seems valuable, and 
>>>>>>>>> I'd like
>>>>>>>>>    to see it addressed before shipping.
>>>>>>>>>
>>>>>>>>> Forgot to note that David did address this in #62
>>>>>>>> <https://github.com/WICG/ScrollToTextFragment/issues/62>, I
>>>>>>>> believe the security and privacy
>>>>>>>> <https://wicg.github.io/ScrollToTextFragment/#allow-text-fragment-directives>
>>>>>>>> section now details all of the feedback and work we've done here.
>>>>>>>>
>>>>>>>>>
>>>>>>>>>    - Regarding Rego and Fréd's feedback on WPTs - I'd like for us
>>>>>>>>>    to reach agreement on which test cases should be added beyond 
>>>>>>>>> what's
>>>>>>>>>    currently covered in order for the test suite to be considered 
>>>>>>>>> complete.
>>>>>>>>>    Rego/Fréd - do you have such a list of cases in mind? Once we reach
>>>>>>>>>    agreement on what that list should be, we should block shipping 
>>>>>>>>> until the
>>>>>>>>>    test suite is complete.
>>>>>>>>>
>>>>>>>>> People developing the feature probably know better the things to
>>>>>>>>> test. That said, after checking a bit the spec and tests, it looks 
>>>>>>>>> like the
>>>>>>>>> features can be divided into the following categories:
>>>>>>>>>
>>>>>>>>> (1) Fragment directive, IDL interface and TreeWalker navigation
>>>>>>>>>     This is the core of the proposal, so it would probably be a
>>>>>>>>> blocker if it
>>>>>>>>>     is not tested extensively. Exiting tests already cover several
>>>>>>>>> cases, but
>>>>>>>>>     I suspect more can be tested here (e.g. check the actual value
>>>>>>>>>     of window.location.fragmentDirective for different cases,
>>>>>>>>> check that
>>>>>>>>>
>>>>>>>>
>>>>>>>> Note that window.location.fragmentDirective does not actually
>>>>>>>> expose the fragment directive string, for now it is just specified for
>>>>>>>> feature detectability (see #19
>>>>>>>> <https://github.com/WICG/ScrollToTextFragment/issues/19> and spec
>>>>>>>> <https://wicg.github.io/ScrollToTextFragment/#feature-detectability>
>>>>>>>> ).
>>>>>>>>
>>>>>>>>
>>>>>>>>>     setting it has no effect, doing query for all combinations of
>>>>>>>>>     mandatory/optional parameters, TextMatchChar, percent encoding
>>>>>>>>> of special
>>>>>>>>>     characters, non-ascii chars, more complex test pages with 0, 1
>>>>>>>>> or more
>>>>>>>>>     matches, with whitespace, with different locales, etc)
>>>>>>>>>
>>>>>>>>> (2) Security & Privacy
>>>>>>>>>     Apparently people have raised concerns about this so it seems
>>>>>>>>> important to
>>>>>>>>>     tests any mitigation or protection described in the spec, if
>>>>>>>>> any (and if
>>>>>>>>>     possible with the current WPT infrastructure).
>>>>>>>>>
>>>>>>>>> (3) Navigating to a Text Fragment
>>>>>>>>>     It seems that the idea of the proposal is to rely on existing
>>>>>>>>> concepts
>>>>>>>>>     like Range/TreeWalker and APIs similar to
>>>>>>>>> window.find/scrollIntoView.
>>>>>>>>>     I think it would be good to have minimal tests checking that
>>>>>>>>> (scroll
>>>>>>>>>     position actually changed, scroll alignment/behavior, hidden
>>>>>>>>> DOM/CSS, etc)
>>>>>>>>>     but this does not need to be exhaustive, since it is assumed
>>>>>>>>> that these are
>>>>>>>>>     already implemented, tested and inter-operable (See comment
>>>>>>>>> below though).
>>>>>>>>>
>>>>>>>>> (4) Indicating The Text Match
>>>>>>>>>     The spec explicitly says it is UA-defined so it cannot really
>>>>>>>>> be
>>>>>>>>>     tested. I guess one could write a minimal != reftest to check
>>>>>>>>> that highlight
>>>>>>>>>     actually happens but it would be very weak anyway, so I'm not
>>>>>>>>> sure it's
>>>>>>>>>     necessary. These will instead likely be browser-specific tests.
>>>>>>>>>
>>>>>>>>
>>>>>>>> Agreed, I think this should be left as browser-specific; we only
>>>>>>>> want to specify the matching/scroll-into-view behavior and leave it up 
>>>>>>>> to
>>>>>>>> the UA/browser how the specific text is actually indicated.
>>>>>>>>
>>>>>>>>
>>>>>>>>> BTW, I don't think my comment regarding BroadcastChannel is a
>>>>>>>>> blocker, I just believe it would be nice to avoid relying on
>>>>>>>>> non-interoperable APIs.
>>>>>>>>>
>>>>>>>>
>>>>>>>> Though not a shipping blocker I definitely want to fix this, I
>>>>>>>> spoke to some WPT experts and using WPT's Stash
>>>>>>>> <https://web-platform-tests.org/tools/wptserve/docs/stash.html>
>>>>>>>> seems like a viable option.
>>>>>>>>
>>>>>>>>
>>>>>>>>> Besides tests, I share Anne's concern on Mozilla repo regarding
>>>>>>>>> lack of existing primitive for actually performing the scroll to 
>>>>>>>>> text. I
>>>>>>>>> opened https://github.com/WICG/ScrollToTextFragment/issues/66 for
>>>>>>>>> that purpose. Right now it's unclear to me if this is well-specified 
>>>>>>>>> and
>>>>>>>>> tested in the current proposal, and this may be considered a blocker.
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> Frédéric Wang
>>>>>>>>>
>>>>>>>>> --
>>>>>>> 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/1238c06f-dcd1-434c-87b8-97a373fdf735%40chromium.org
>>>>>>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/1238c06f-dcd1-434c-87b8-97a373fdf735%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/CACj%3DBEgFMwT1ArGjrHMcWZ9pKe8%2Bsv%2BJHDLpOd4ofOQss0a-zA%40mail.gmail.com
>>>>>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACj%3DBEgFMwT1ArGjrHMcWZ9pKe8%2Bsv%2BJHDLpOd4ofOQss0a-zA%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>>>>> .
>>>>>>
>>>>> --
>> You received this message because you are subscribed to a topic in the
>> Google Groups "blink-dev" group.
>> To unsubscribe from this topic, visit
>> https://groups.google.com/a/chromium.org/d/topic/blink-dev/zlLSxQ9BA8Y/unsubscribe
>> .
>> To unsubscribe from this group and all its topics, 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/e9c07baa-3c2a-4835-9014-9d5a2b249618n%40chromium.org
>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/e9c07baa-3c2a-4835-9014-9d5a2b249618n%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/CAAyoetRLA%3Da5WScarBrGJnBOpiR9A7eBNJ6nt%3DR4CXDB7czLCg%40mail.gmail.com.

Reply via email to