W3C home > Mailing lists > Public > www-style@w3.org > April 2013

Re: [css4-background] use cases for 'border-corner-shape'?

From: Liam R E Quin <liam@w3.org>
Date: Fri, 05 Apr 2013 00:08:52 -0400
To: Lea Verou <lea@w3.org>
Cc: "L. David Baron" <dbaron@dbaron.org>, www-style@w3.org
Message-ID: <1365134932.11833.114.camel@localhost.localdomain>
On Fri, 2013-04-05 at 05:42 +0300, Lea Verou wrote:
> Here are a few more use cases, provided by the community:
[...]

Wow, quite a few links.

> Also, here are a few comments from authors:

> The most often example is arrows of any kind. Right now there are a
> lot of hacks and tricks to make them work, but if you'd need one with
> gradient, border and box-shadows — it would become really cumbersome
> to emulate. [...] Another thing are negative border-radiuses near
> tabs.

Tabs seem to be such a pervasive design paradigm on the Web that one
could easily argue for HTML-level support.

As for borders - we spent some time in another group trying to find a
way to make rectangular borders be a special case of general borders -
e.g. a list of path segments with joins and styles

SVG does not directly meet the use case for liquid layouts (you don't
want the border scaling and taking up more room as the winder gets
larger, e.g. the corner radius doesn't increase on a box when you resize
the window) but could be made to do so.

But even then, people will still want and use more corner shapes.

Negative corner radii might make sense; SVG also has a list of corner
join methods such as mitre and butt that could be used along with
negaative radii. The more complex certificate-shapes are things I know
I'd love to see, but fear we're baking special cases into a spec. Can we
come up with a way to draw the corners that still as the right
scaling/growing behaviour, and add fewer of the special cases?

I don't want to stand in the way of something good - but I don't like
lists of special cases either. Why not a double or quintuple scallop?
Why not an inset with a specified radius? Or a polygon instead of an
arc? They're all useful.

I do like the negative radius, and I think the bevel is simply an
example of line-join style that's already floating around, and I would
be in favour of the others if I didn't think there was a chance of using
SVG in a way that didn't make the lines get thicker when the content box
got bigger. After that, we could pre-define some SVG paths and give them
names, but users could add their own.

Liam

-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Ankh: irc.sorcery.net irc.gnome.org freenode/#xml
Received on Friday, 5 April 2013 04:08:59 UTC

This archive was generated by hypermail 2.3.1 : Friday, 5 April 2013 04:08:59 UTC