W3C home > Mailing lists > Public > www-style@w3.org > October 2016

Re: [motion-1] !important: motion-* rename

From: Shane Stephens <shans@google.com>
Date: Sat, 01 Oct 2016 01:31:16 +0000
Message-ID: <CAGTfzwSpMpsMZ8UPU1bOt2xhT1S8k+AoV7T5bf9VCZJg4p=tZg@mail.gmail.com>
To: Amelia Bellamy-Royds <amelia.bellamy.royds@gmail.com>, "Tab Atkins Jr." <jackalmage@gmail.com>
Cc: "www-style@w3.org" <www-style@w3.org>
I don't think that it's a good idea to make use of the 'transform' word at
all, as it heavily implies full 2D/3D transform capabilities, which are not
being provided with this feature.

I can empathize with Amelia's concerns about a path-shape property on a
<path> element (less so with path-distance, and there wouldn't be a
path-offset). In general, for better or worse, we've normalized on 'path'
meaning 'a piece of 2D geometry that can be used both as a shape and for
positioning' - having a path property that can't be used for a shape is
therefore confusing. The positioning part of this is interesting - when I
try to describe how a path is used for positioning I inevitably end up
coming back to 'time based positioning' (motion) or 'offset positioning'
because you're describing a locus on the path by means of the length of the
segment from the origin to the locus.

We don't make up words for CSS properties, so 'transpath' is out.

Coming back to offset: Amelia says "nor can the net effect be defined as a
single offset from a base position". But .. it can. The net effect is
exactly an offset of offset-distance along the offset-path (which is a path
in the normal CSS+SVG sense of the word path), modified by the
offset-rotation (which is just a rotation). How the offset and rotation
applies to the element is controlled by offset-anchor, and the path itself
is tied back to the global geometry via offset-position.

This name actually makes the most sense of anything we've come up with. It
doesn't step on existing terms (except for 'offset' from Web Animations,
which admittedly is a shame) and it's really descriptive of what the
feature does - without modifying layout or the geometry of an element, it
offsets the position of that element in some well-defined way.

Cheers,
    -Shane



On Sat, Oct 1, 2016 at 5:39 AM Amelia Bellamy-Royds <
amelia.bellamy.royds@gmail.com> wrote:

> On 30 September 2016 at 13:08, Tab Atkins Jr. <jackalmage@gmail.com>
> wrote:
>
> At this point, why not just drop the "-transform" entirely? "path" as
> the shorthand, "path-shape", "path-distance", etc for the longhands.
> No need to have those extra 10 characters cluttering up the names.
>
>
> I would find that very confusing, especially when working with SVG.  If
> there's a property on a <path> element called path-shape, I would assume
> that it actually controlled the shape of the element.  And path-offset and
> path-distance could easily get confused for the stroke dashing properties
> and textPath attributes.
>
> My main suggestion was to have a name that emphasized that these
> properties apply a transform on the element, defined by the specified path
> & other parameters.
>
> If conciseness is a priority, we could always just make up a new word,
> like "transpath". But "path-transform" isn't any longer than many other
> property names that have sub-properties (border-image, text-decoration,
> etc.).
>
> ~ABR
>
Received on Saturday, 1 October 2016 01:31:56 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:09:04 UTC