Intent to deprecated and remove: overflow: overlay

[Julien Chaffraix]

Primary eng (and PM) emails:
jchaf...@chromium.org

Summary:
overflow:overlay is a proprietary value that has been around for a
long time [1]. It is similar to overflow: auto except that the
scrollbars don't take up any space in your layout (note that the
concept of overlay scrollbars is orthogonal in Blink which means that
the scrollbars can be always visible even with this property). Tab
Atkins tried to get it standardized but the working group's consensus
was globally negative [2].

Motivation:
I haven't found any online documentation that accurately describe what
it does (most link equate it to 'auto' which is not completely
accurate, see e.g. [3], [4]). The feature is Blink / WebKit only (IE
implements it on top of -ms-overflow-style) that failed at
standardization so the only way forward is deprecation / removal.

Finally while the footprint of the feature is small, I think it would
help simplify our scrolling code a bit.

Compatibility Risk:
This intent is to add a deprecation message and a removal once the
usage is low enough. I don't have any data to say when the removal
will happen. My gut feeling is that very few websites use it but I
prefer to take the safe approach.

What do people think?

Julien

[1] https://bugs.webkit.org/show_bug.cgi?id=32388#1
[2] http://lists.w3.org/Archives/Public/www-style/2013Jan/0386.html
[3] http://help.dottoro.com/lccdtehj.php
[4] http://quirksmode.org/css/css2/overflow.html

[PhistucK]

This feature is so unknown, that I would remove it without any notice.

Almost no one mentions this value and even a Google search yields very few relevant results, which are mostly from the W3C mailing list, or one StackOverflow question about it not working in Firefox (which should have implied to interested developers that it is not cross browser anyway), along with varying degrees of incorrect information.

☆PhistucK

To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.

[Julien Chaffraix]

Re-adding blink-dev

> What does the usecounter data look like?

There is no UseCounter for the feature yet. This email is to add the
deprecation message + UseCounter with a focus on removal once we are
happy with the numbers.

Julien

[Kenneth Rohde Christiansen]

lgtm

> To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.

--
Kenneth Rohde Christiansen
Web Platform Architect, Intel Corporation.
Phone +45 4294 9458 ﹆﹆﹆

[Julien Chaffraix]

> This feature is so unknown, that I would remove it without any notice.
> Almost no one mentions this value and even a Google search yields very few
> relevant results, which are mostly from the W3C mailing list, or one
> StackOverflow question about it not working in Firefox (which should have
> implied to interested developers that it is not cross browser anyway), along
> with varying degrees of incorrect information.

I agree with you that it's not a well known feature and probably could
do with a direct removal. The only reason I didn't push for that is
that I don't have any data to back the removal. Adding a UseCounter
should only defer the removal by 2 months if we are right but we would
be sure we are not breaking anything significant.

Julien

[PhistucK]

Can a UseCounter be merged to Chrome 31 (currently in early beta), or are those types of changes disallowed on beta at this stage?

☆PhistucK

[Julien Chaffraix]

> Can a UseCounter be merged to Chrome 31 (currently in early beta), or are
> those types of changes disallowed on beta at this stage?

I don't really have an answer to your question.

We talked several times among developers of doing such merges but no
one tried to do one so far. We would need to get the product managers
involved in the releases on board if we were to do that. Those merges
are simple enough that we could argue our case but I wouldn't assume
they would be happy with mostly untested code landing on a stabilizing
branch.

Julien

[PhistucK]

So I guess you will not be the first to try, huh?

;)

☆PhistucK

[Tony Chang]

FYI, we use this a lot in the inspector UI.  Search for "overflow-y: overlay" in Source/devtools/front_end/.

[PhistucK]

Do the UseCounters count items used by the Developer Tools?

If so, along with user agent styles, this should be excluded as well.

(And the Developer Tools feature must get rid of this non standard feature quickly...)

☆PhistucK

[Julien Chaffraix]

> Do the UseCounters count items used by the Developer Tools?
> If so, along with user agent styles, this should be excluded as well.

Yes, this is http://crbug.com/293000 and that's a broader issue that
we count all internal pages (e.g. about://flags).

> (And the Developer Tools feature must get rid of this non standard feature
> quickly...)

AFAICT the Dev Tools are the only internal users of the property.

Julien

[Eric Seidel]

"This intent is to add a deprecation message and a removal once the

usage is low enough."

This sounds like an intent to deprecate only.  which SGTM.   I'm supportive of removal too, but that's entirely up to the usage numbers (which we don't have yet).

[godfrey...@prospectsplus.com]

We use this quite a bit in our editor. It is very helpful and I was disappointed that it is not widely supported.

[PhistucK]

Since then, the downloads and settings web user interface pages added more usage.

ChromeStatus does not mention a use counter for it, was it ever added? If not, can someone add it? :(

☆PhistucK

[may...@gmail.com]

I'd like to use the property in my project, but now Chrome just overlays the page with a standard opaque scrollbar (windows machine) instead of a slim one with a transparent track. So it's fairly useless/ugly now. Or was the intention that the scrollbar should be then adjusted via CSS?

[PhistucK]

> but now

As opposed to what behavior and when?

(This was not removed)

☆PhistucK

--
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 blink-dev+...@chromium.org.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/aefbb346-2cec-488c-b5c6-a6e7d6afa70c%40chromium.org.

[Vitaliy Kupets]

In the past years I've seen some websites having the typical overlay scrollbar like on mobile (transparent track and dark thin semitransparent thumb disappearing on inactivity) and I've checked the source code which said they used overflow: overlay. I don't remember checking styling CSS, so it's possible they all did the custom styling (which looked identical everywhere, so I assumed it was in-browser).

[PhistucK]

I do not think mobile browsers (at least on Android and iOS) ever employed a non-overlay scroll bar, regardless of overflow values. If they used overflow: overlay, it might have been to get the same positioning calculations on desktop (with a probably degraded experience, or using custom scroll bars as you suggest).

☆PhistucK

[zuf...@google.com]

I found this thread after seeing `overflow: overlay` recommended at https://stackoverflow.com/a/20014743/7320095.