Intent to Deprecate and Remove: HTMLPreElement.wrap IDL attribute

[Philip Jägenstedt]

Primary eng (and PM) emails

phi...@opera.com

Summary

Remove the HTMLPreElement.wrap IDL attribute, but not the effect the content attribute has on rendering.

Motivation

The IDL attribute is not in the spec, and no browser except WebKit/Blink supports it:

http://www.whatwg.org/specs/web-apps/current-work/multipage/semantics.html#htmlpreelement

http://jsconsole.com/?document.createElement('pre').wrap

However, the content attribute is part of the user agent style sheet, and all browsers except IE support it:

http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#flow-content-3

http://software.hixie.ch/utilities/js/live-dom-viewer/saved/3079

Compatibility Risk

Insignificant.

Alternative implementation suggestion for web developers

Use CSS to control wrapping, and a class if it needs to be toggled from scripts.

Usage information from UseCounter

http://www.chromestatus.com/metrics/feature/timeline/popularity/399

Available in M36, the highest recorded measurement is 0.0002%.

When I added the use counter I wasn't aware that the user agent stylesheet bit was per spec, so I included that as well. Fortunately, the combined usage is low enough that it doesn't matter.

Entry on chromestatus.com, crbug.com, or MDN

None.

Requesting approval to remove too?

Yes.

[Elliott Sprehn]

Yay, please keep up the removals of all this vestigial stuff. :)

[Tab Atkins Jr.]

LGTM as well

On Mon, Jul 28, 2014 at 3:27 PM, Philip Jägenstedt <phi...@opera.com> wrote:

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

[Eric Seidel]

lgtm

[Dimitri Glazkov]

LGTM2

[Jochen Eisinger]

lgtm3

[Erik Arvidsson]

If all browsers support the wrap boolean attribute it seems like reflecting the attribute as a property is reasonable.

--
erik

[Anne van Kesteren]

On Tue, Jul 29, 2014 at 7:36 PM, Erik Arvidsson <a...@chromium.org> wrote:
> If all browsers support the wrap boolean attribute it seems like reflecting
> the attribute as a property is reasonable.

Why should non-WebKit/Blink browsers add this? No requests in Bugzilla.


--
http://annevankesteren.nl/

[Erik Arvidsson]

My concern is that is that we should not have have content attributes that are not reflected. I do not care about wrap in particular, just about consistency.

--
erik

[Philip Jägenstedt]

FWIW, <pre wrap> is invalid.

If we want to reach a consistent, not just interoperable, state I think that dropping pre[wrap] { white-space: pre-wrap; } from the UA style sheet would be the quickest path to success. Lack of support in IE and the use counter data (which included that case) make that change seem very safe. It would also make HTMLPreElement.h/cpp empty like HTMLHeadElement.h/cpp.

If you like, a spec bug seems like the next step.

Philip

[Ian Hickson]

On Tue, 29 Jul 2014, Philip Jägenstedt wrote:
>
> FWIW, <pre wrap> is invalid.
>
> If we want to reach a consistent, not just interoperable, state I think
> that dropping pre[wrap] { white-space: pre-wrap; } from the UA style
> sheet would be the quickest path to success. Lack of support in IE and
> the use counter data (which included that case) make that change seem
> very safe. It would also make HTMLPreElement.h/cpp empty like
> HTMLHeadElement.h/cpp.
>
> If you like, a spec bug seems like the next step.

Removing it (the CSS rule) seems fine to me, but I would expect this to
have effects on the Web. A use counter for the number of times that rule
is hit would provide useful data.

--
Ian Hickson U+1047E )\._.,--....,'``. fL
http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,.
Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'

[Philip Jägenstedt]

On Tue, Jul 29, 2014 at 11:14 PM, Ian Hickson <i...@hixie.ch> wrote:
>
> On Tue, 29 Jul 2014, Philip Jägenstedt wrote:
> >
> > FWIW, <pre wrap> is invalid.
> >
> > If we want to reach a consistent, not just interoperable, state I think
> > that dropping pre[wrap] { white-space: pre-wrap; } from the UA style
> > sheet would be the quickest path to success. Lack of support in IE and
> > the use counter data (which included that case) make that change seem
> > very safe. It would also make HTMLPreElement.h/cpp empty like
> > HTMLHeadElement.h/cpp.
> >
> > If you like, a spec bug seems like the next step.
>
> Removing it (the CSS rule) seems fine to me, but I would expect this to
> have effects on the Web. A use counter for the number of times that rule
> is hit would provide useful data.

http://www.chromestatus.com/metrics/feature/timeline/popularity/399

This use counter includes the CSS bit, because at the time I didn't
realize it was in the spec. To be specific, it counted any time the
wrap attribute case in
HTMLPreElement::collectStyleForPresentationAttribute was hit.

Unless I've missed something, removal would be entirely safe. (Large
amounts of content can't depend on it because of IE.)

Philip