W3C home > Mailing lists > Public > www-style@w3.org > May 2012

Re: [css3-animations] `alternate-reverse` vs `alternate reverse`

From: Kang-Hao (Kenny) Lu <kennyluck@csail.mit.edu>
Date: Mon, 07 May 2012 12:41:32 +0800
Message-ID: <4FA7527C.1040106@csail.mit.edu>
To: Lea Verou <leaverou@gmail.com>
CC: Sylvain Galineau <sylvaing@microsoft.com>, WWW Style <www-style@w3.org>, fantasai <fantasai.lists@inkedblade.net>
(12/05/06 0:28), Lea Verou wrote:
> Consistency-wise, I think existing CSS properties usually allow `normal`
> to be combined with other keywords, as long as disambiguation is
> possible. For example, the following declaration is perfectly valid
> (albeit needlessly verbose), even in CSS1:
> 
> font: normal normal normal 100%/normal normal;

I think it's a bit controversial to say "existing CSS properties usually
allow `normal` to be combined with other keywords" with a shorthand
property like 'font'. For example, although 'border-style' shorthand
allows 'none' to be combined with other keywords, in general for a
longhand property, 'none' cannot be combined with other keywords:

  # 'content'
  #
  # normal | none | [ <string> | <uri> | <counter> | attr(<identifier>)
  # | open-quote | close-quote | no-open-quote | no-close-quote ]+ |
  # inherit
  #
  # 'quotes'
  #
  # [<string> <string>]+ | none | inherit
  #
  # 'text-decoration'
  #
  # none | [ underline || overline || line-through || blink ] | inherit

.

The truth is, there is no longhand property in CSS2.1 that has 'normal'
and two other values that can be combined. (Well, except 'content', but
combining 'normal' with <string> and others don't make much sense.)

If you look at CSS3, then there are several longhand properties that
disallow 'normal' to be combined with other keywords, including all the
font-variant-* (and the 'font-variant' shorthand actually) and
'unicode-bidi'. That is, I think the consistency is in favor of the
other way around.


(12/05/06 3:32), Sylvain Galineau wrote:
> I'm not sure we want or need this consistency here; I think normal
> was just a poor choice as the opposite to alternate. Maybe
> 'continuous' would have been better?

It's probably weird if you think of 'normal' as the opposite to
'alternate'. It's less weird if you think of 'normal' as "not reverse
and not alternate".

Therefore, I prefer Lea's original proposal ( normal | [ alternate ||
reverse ] ).


(12/05/07 11:14), fantasai wrote:
> Switching 'normal' to 'reverse' or 'alternate-reverse' to 'alternate
> reverse' wouldn't affect sites specifying 'alternate'.

I think you meant 'forward' here?

Yeah, if we can make backwards incompatible change, I'll probably prefer
something similar: [ backward | forward ] || round. But if we can't
switch 'normal' to 'forward', I still prefer Lea's original grammar.

Note that naming matters here, if the initial value of 'transform' is
'identity' instead of 'none', you would probably allow it to be combined
with other values.


As a non-English native speaker, may I ask why we use 'reverse' here
instead of 'reversed'? <ol> has a content attribute 'reversed'[1] but I
don't know the difference between these words (nor do I actually know
the difference between 'alternate' and 'alternative' but I guess at
least there's no conflicts with other Web technology).


[1]
http://www.whatwg.org/specs/web-apps/current-work/multipage/grouping-content.html#attr-ol-reversed


Cheers,
Kenny
Received on Monday, 7 May 2012 04:42:05 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:53 GMT