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

Re: EPUB using CSS vendor prefixing incorrectly?

From: Ojan Vafai <ojan@chromium.org>
Date: Wed, 21 Aug 2013 14:32:14 -0700
Message-ID: <CANMdWTtuY8=t-8Nfid162yr2sXQPdiUPt2_=ckpZuWFDGbypqA@mail.gmail.com>
To: "L. David Baron" <dbaron@dbaron.org>
Cc: Sylvain Galineau <galineau@adobe.com>, James Craig <jcraig@apple.com>, "www-style@w3.org list" <www-style@w3.org>
On Mon, Aug 19, 2013 at 4:25 PM, L. David Baron <dbaron@dbaron.org> wrote:

> On Monday 2013-08-19 16:04 -0700, Sylvain Galineau wrote:
> > On 8/19/13 12:45 PM, "James Craig" <jcraig@apple.com> wrote:
> >
> > >IDPF appears to be specifying a EPUB-specific technology subset prefix
> > >using the syntax for the CSS vendor prefix. This strikes me as both
> > >problematic (precludes vendor prefixing with this) as well as
> unnecessary
> > >(why not just use the unprefixed properties), but perhaps I'm
> > >misunderstanding something.
> > >
> > >      -epub-speak
> > >      -epub-speak-as
> > >
> > >      -epub-text-align-last
> > >      -epub-text-emphasis
> > >      -epub-text-emphasis-color
> > >      -epub-text-emphasis-style
> > >      -epub-word-break
> > >
> > >
> http://www.idpf.org/epub/30/spec/epub30-contentdocs.html#sec-css-voice-cha
> > >racteristics
> > >http://www.idpf.org/epub/30/spec/epub30-contentdocs.html#sec-css-text
> > >
> > >Could someone please explain whether this is a valid use of the vendor
> > >prefix pattern? If it's not, the CSS WG chair should probably reach out
> > >to the IDPF and EPUB groups. If this *is* a valid use CSS vendor
> > >prefixes, are there any known implementations?
> >
> > My understanding is that this is in fact necessary because EPUB releases
> > may depend on specific CSS *drafts*; thus until EPUB version N depends on
> > a REC for a given feature it uses -epub-.
>
> I'm inclined think it's still a bad idea (although I'm really not
> sure); it leads to accumulation of content that uses those vendor
> prefixes.  I think no matter whether future implementors consider
> that content important enough to support, or whether handling the
> content requires behavior different from what would end up in the
> Recommendation, the Web would still be better off with epub using
> these properties without prefixes.  (The Web might end up better off
> because the pressure from existing use prevented unnecessary changes
> in the properties, or because there ended up being less complexity
> in the platform from having fewer variants supported in Web layout
> engines, or because engine developers just didn't worry about the
> epub content.)


+1

FWIW, Blink exposes 9 -epub properties that are all aliased to other
properties. I'll try adding some measurements to see if we can remove all
these epub prefixes from Blink.

-epub-text-orientation alias_for=-webkit-text-orientation
-epub-writing-mode alias_for=-webkit-writing-mode
-epub-caption-side alias_for=caption-side
-epub-text-transform alias_for=text-transform
-epub-word-break alias_for=word-break
-epub-text-combine alias_for=-webkit-text-combine
-epub-text-emphasis alias_for=-webkit-text-emphasis
-epub-text-emphasis-color alias_for=-webkit-text-emphasis-color
-epub-text-emphasis-style alias_for=-webkit-text-emphasis-style
Received on Wednesday, 21 August 2013 21:33:01 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 21 August 2013 21:33:02 UTC