Re: [css3-break] fragment selectors

> On 20Apr, 2016, at 14:10, Tab Atkins Jr. <jackalmage@gmail.com> wrote:
> 
> On Tue, Apr 19, 2016 at 8:13 PM, Lea Verou <lea@verou.me> wrote:
>> And like I suggested, styling on ::nth-*-fragment() would not affect fragmentation, which alleviates all of these concerns.
> 
> ::nth-fragment not affecting fragmentation is surprising and
> incompatible with how ::first-line works.  ::first-line styling
> explicitly *does* affect fragmentation - it can easily alter where the
> linebreak occurs.  Anything similar *should* act the same,

I would prefer it to act the same too. However, not affecting fragmentation still solves a bunch of use cases, in a way that can actually be implemented.

> which is
> why we're not planning to add them to native CSS. ^_^

Who is we here?

> Like I said, Houdini Custom Layout can handle this reasonably well,
> tho - by taking a holistic approach to the layout, it can determine
> whether things will break or not, and what to do when they do break.
> It can make the somewhat-arbitrary decision about what to do in the
> case of cycles, and can tweak to satisfaction (or offer a config for
> it).  A library handling this is definitely not easy, but it shouldn't
> be ridiculously hard either.
> 
> This is *precisely* what the Custom * stuff is about: addressing
> complex niche cases that are worth solving for people, but not worth
> adding to the language and shipping to everyone.

And this is precisely why I'm worried about Houdini. It's not the first time I see it used as an excuse to avoid adding anything to CSS. This is NOT a niche case in print. Using a library includes a lot of overhead: Finding the best library, managing dependencies, learning its syntax, even thinking to look for a library in the first place. This is not trivial for designers and other kinds of CSS authors with limited technical experience. Houdini is great for polyfills or niche cases like e.g. the Masonry layouts, but this is a basic need in any serious publication.

~Lea

Received on Wednesday, 20 April 2016 19:04:47 UTC