Re: [cssom-view] scroll-behavior CSS property 'auto' value

On Thu, 02 Oct 2014 09:42:22 +0200, Simon Pieters <simonp@opera.com> wrote:

> On Wed, 01 Oct 2014 20:34:03 +0200, Kip Gilbert <kgilbert@mozilla.com>  
> wrote:
>
>> It is desirable in some circumstances for content to disable all
>> smooth scrolling, including scrolling that is not due to navigation or
>> CSSOM scrolling APIs.  One use case would be a spreadsheet
>> application.  In this case, it might be desirable to snap to a cell
>> boundary without the animation.
>>
>> To enable content authors to explicitly disable all smooth scrolling
>> animation (including the scroll bar smooth scrolling and scroll
>> snapping animations), would the scroll-behavior CSS property be the
>> ideal interface for this?
>
> Does it make sense to disable smooth user scrolling when there are no  
> snap points? What should happen then on e.g. a touch device when the  
> user flicks?

Replies that accidentally were off-list (snipping deep quotes):

On Mon, 06 Oct 2014 21:29:23 +0200, Kip Gilbert <kgilbert@mozilla.com>  
wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On 2014-10-02, 12:42 AM, Simon Pieters wrote:

>> Does it make sense to disable smooth user scrolling when there are
>> no snap points? What should happen then on e.g. a touch device when
>> the user flicks?
>>
>
> One use case for disabling smooth scrolling without scroll snapping
> would be a web application intended for viewing on an e-ink display or
> other high latency monitor.  Ideally a UA could have knowledge of such
> a display and disable flings and smooth scrolling by default, for the
> "auto" scroll-behavior value.
>
> IMHO, interfaces feel more dated when fling gestures are not enabled;
> however, it might not be best for applications such as graphic novel
> readers, CAD, and drawing applications where precise positioning is
> desirable.
>
> Taking an example from Google's Spreadsheet, "Sheets", fling gestures
> can work well when mixed with instant scroll snapping.  Perhaps
> enabling or disabling the fling panning gestures and fling zooming
> gestures would be ideal as a separate attribute that could be set
> independently from scroll-behavior?
>
> I expect that most web content would choose "auto" or "smooth" values
> for the scroll-behavior CSS property unless they have specialized
> requirements such as those above.
>
>
> Thanks again for your feedback.
>
> - - Kearwood "Kip" Gilbert
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1
>
> iQEcBAEBAgAGBQJUMu2TAAoJEHziBT1mB5A7EEQIAJQ12Urks9DHLb/5nL1yuOCC
> 3yfiptRxhWAbsybQLRFrOPptZPC6eVy6iCmdGc7xHxE0mOEHH5a+kV+UuSGPreci
> cuHR/jGEg7WKLefNA4x/xnbLMH+12vHvsTHRuK4TKJtQMHfpZMoqj4V0SDdC30FA
> YxMNji6QJPeUALn3StO3T6+gbpbuG3lS8aLEDp8I+EjMXiGf6TYIP6YWzl5rbJQ4
> s1rMHXapo1gZCRsFxOBxJIFSipkpT50QBaLClM/sO1+xlK+6WhO/Jay1ZUXeFQv3
> K5aT6aFC0c5M77+F2DxgZZO5uyWlqBMLdu1ZtLTIlRSuiMYs9fQxd0ZSW74HmBM=
> =uS1H
> -----END PGP SIGNATURE-----

On Tue, 07 Oct 2014 09:50:26 +0200, Simon Pieters <simonp@opera.com> wrote:

> On Mon, 06 Oct 2014 21:29:23 +0200, Kip Gilbert <kgilbert@mozilla.com>  
> wrote:
>
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> On 2014-10-02, 12:42 AM, Simon Pieters wrote:

>> One use case for disabling smooth scrolling without scroll snapping
>> would be a web application intended for viewing on an e-ink display or
>> other high latency monitor.  Ideally a UA could have knowledge of such
>> a display and disable flings and smooth scrolling by default, for the
>> "auto" scroll-behavior value.
>
> The UA needs to take care of that without help from the author. About  
> 100% of authors aren't going to do anything about it. So I think that's  
> not a valid use case.
>
>> IMHO, interfaces feel more dated when fling gestures are not enabled;
>> however, it might not be best for applications such as graphic novel
>> readers, CAD, and drawing applications where precise positioning is
>> desirable.
>
> Not sure I follow here.
>
>> Taking an example from Google's Spreadsheet, "Sheets", fling gestures
>> can work well when mixed with instant scroll snapping.  Perhaps
>> enabling or disabling the fling panning gestures and fling zooming
>> gestures would be ideal as a separate attribute that could be set
>> independently from scroll-behavior?
>
> Why would you let the author disable fling panning?
>
> Zooming is handled by <meta name=viewport> or @viewport I think.
>
>> I expect that most web content would choose "auto" or "smooth" values
>> for the scroll-behavior CSS property unless they have specialized
>> requirements such as those above.
>
> I don't want to enable something that leads to a bad user experience  
> when authors use it for bogus reasons (copy-paste, reading an old copy  
> of the spec, etc). So if instant scrolling is only useful for  
> spreadsheets with snap points, and worsens user experience without snap  
> points, it seems best for users to make instant only have an effect when  
> scrolling to a snap point.


On Tue, 07 Oct 2014 19:40:01 +0200, Kip Gilbert <kgilbert@mozilla.com>  
wrote:

> On 2014-10-07, 12:50 AM, Simon Pieters wrote:
>
>> I don't want to enable something that leads to a bad user
>> experience when authors use it for bogus reasons (copy-paste,
>> reading an old copy of the spec, etc). So if instant scrolling is
>> only useful for spreadsheets with snap points, and worsens user
>> experience without snap points, it seems best for users to make
>> instant only have an effect when scrolling to a snap point.
>>
>
> I tend to agree.  Beyond the case with snap-points, I don't see many
> use cases for disabling the smooth scrolling animation.  I'll check
> with my peers to see if they have any objection to 'instant' simply
> disabling the animation for snap-points.
>
> Cheers,
>  - Kearwood "Kip" Gilbert

-- 
Simon Pieters
Opera Software

Received on Wednesday, 8 October 2014 08:33:39 UTC