- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Thu, 2 Apr 2009 14:18:12 -0500
- To: "Philip TAYLOR (Ret'd)" <P.Taylor@rhul.ac.uk>
- Cc: CSS <www-style@w3.org>, Håkon Wium Lie <howcome@opera.com>
On Thu, Apr 2, 2009 at 9:59 AM, Philip TAYLOR (Ret'd)
<P.Taylor@rhul.ac.uk> wrote:
> Håkon Wium Lie wrote:
>> I think it's a common scenario that authors don't want elements to
>> break. No?
>
> To break across pages : yes; to break across columns ?
> I am far less convinced. But if you can adduce an
> example of a situation in which both page breaks /and/
> column breaks should be inhibited, then I'd be very
> interested to see it.
>
> To address the issue from the other perspective
> (should "page" be a part of the value, rather
> than the property) is interesting (and generalises
> rather better). So one could envisage (say)
>
> break-inside: allow-column
> break-inside: inhibit-page
> break-inside: inhibit-all
>
> and so on. But regardless of whether {page|column}
> are properties or values, I remain convinced that
> they are totally different, and that it would be
> a serious error of judgement to conflate the two.
If I'd gotten to my email this morning rather than this afternoon,
this is precisely what I would have suggested. We have two different
constructions that authors may want to control breaking across (and I
can *definitely* see column-breaking being an issue for myself*).
It's within the realm of possibility that future constructions may
have breaking behavior we'd want to control. So just generalize now,
kill the ambiguity, and everyone wins.
* Being able to prevent an element from breaking at all is required.
Being able to prevent an element breaking across pages, while allowing
a column break, would be very nice. Being able to prevent an element
breaking across columns, while allowing a page break, doesn't seem
useful. However, it's likely simpler to allow it if the other two
scenarios are allowed. And I'd hate to build in this sort of nesting
logic (page break are worse than column breaks) in case we *do*
introduce another construction with breaking properties later.
~TJ
Received on Thursday, 2 April 2009 19:18:47 UTC