Re: [css3-ui] text-overflow:ellipsis interaction with atomic inline elements is not web-compatible

On Tue, Jan 3, 2012 at 01:42, Florian Rivoal <florianr@opera.com> wrote:
> On Sun, 01 Jan 2012 00:59:53 +0100, Tantek Çelik <tantek@cs.stanford.edu>
> wrote:
>
>> I think IE10's treatment is the most reasonable and consistent, and
>> I'm wondering if other implementations are willing to change their
>> treatment of those test cases to match. That's the direction that
>> we're leaning with Firefox/Gecko.
>
>
> Opera is willing to change our behavior about this. Every browser
> does something different today, and that's not good. We have no
> particular preference for our own implementation in this case,
> so we are in favor of the WG trying to reach a consensus on
> what's supposed to happen, and everybody aligning to it.

Thanks Florian, that is good to hear.

Per the consensus at last week's CSS WG telcon I updated the CSS3-UI
editor's draft to specify that both characters and inline atomic
elements are ellipsed, except for the first such thing on a line,
which is clipped not ellipsed - which is IE's behavior since IE10 (and
apparently earlier as it was asserted that functionality hasn't been
altered in recent history).

http://dev.w3.org/csswg/css3-ui/#text-overflow

Also, I left the subject line as-is for threading from the original:
http://lists.w3.org/Archives/Public/www-style/2011Nov/0537.html
However the assertion in the subject line is imprecise and may be misleading.

Eliding inline atomic elements entirely if they overflow is *exactly*
what implementations do (and is thus required to be web-compatible)
except if an inline atomic element is the first thing on the line.


> As for what the ideal behavior should be, I haven't given it any
> deep though, but IE10's solution does seem reasonable. Firefox's
> isn't bad either, though. If you've already though about that,
> could you share why mozilla thinks IE's solution is better?

Two reasons:

1. It makes more content sense to show (even if clipped) the first
thing on a line than *just* an ellipse on the line. I think the
results of the test case I posted:

https://bug690187.bugzilla.mozilla.org/attachment.cgi?id=583694

In IE10, demonstrate that:
https://bug690187.bugzilla.mozilla.org/attachment.cgi?id=583917


2. Web sites *do* depend on showing the *first* inline atomic element
on the line clipped rather than ellipsing it, and it makes sense to be
consistent with that for characters as well. Here are the specific
examples we found (with the currently shipping Firefox implementation)
:

[1] https://bugzilla.mozilla.org/show_bug.cgi?id=680610
[2] https://bugzilla.mozilla.org/show_bug.cgi?id=689897
[3] https://bugzilla.mozilla.org/show_bug.cgi?id=690131
[4] https://bugzilla.mozilla.org/show_bug.cgi?id=690187
[5] https://bugzilla.mozilla.org/show_bug.cgi?id=703360


Thanks,

Tantek

Received on Wednesday, 11 January 2012 22:42:42 UTC