W3C home > Mailing lists > Public > www-style@w3.org > August 2011

RE: [css] Proposal: making Shorthand Hex Colors even shorter (16 grayscale shades)

From: Sylvain Galineau <sylvaing@microsoft.com>
Date: Wed, 3 Aug 2011 15:21:17 +0000
To: Antony Kennedy <antony@silversquid.com>, Alan Gresley <alan@css-class.com>
CC: Tab Atkins Jr. <jackalmage@gmail.com>, Brian Manthos <brianman@microsoft.com>, Markus Bruch <macinfo@arcor.de>, "www-style@w3.org" <www-style@w3.org>
Message-ID: <3C4041FF83E1E04A986B6DC50F017829313A1C@TK5EX14MBXC296.redmond.corp.microsoft.com>
Not all savings are a worthwhile endeavor because they're trade-offs: savings do not come for free. If the goal is to improve minification, some data on the gains achieved on real-world CSS would be helpful. I find it hard to believe there are so many color values that can be reduced to 2 digits in the average stylesheet that it would significantly improve minification, at least by a rate that justifies the additional complexity. Never mind that in practice such savings are only really possible when most deployed browsers support the new syntax.

Supporting alpha in the hex notation does make sense from a consistency standpoint; that it also saves a few extra bytes is a positive side-effect.

From: www-style-request@w3.org [mailto:www-style-request@w3.org] On Behalf Of Antony Kennedy
Sent: Wednesday, August 03, 2011 2:40 AM
To: Alan Gresley
Cc: Tab Atkins Jr.; Brian Manthos; Markus Bruch; www-style@w3.org
Subject: Re: [css] Proposal: making Shorthand Hex Colors even shorter (16 grayscale shades)

While I'm not necessarily opposed to something like this, I question
the utility of saving two characters on a mere 16 colors.  I think the
3-digit hex is very readable and trivial to type.

I respectfully disagree. I think the saving of any amount of characters is a worthwhile endeavour, and aids future minification scripts and methodologies. I think a one or two character hex code is trivial to read and understand.

I am strongly opposed to a 2-digit variant, because it has a different
expansion rule than the 3-digit hex that already exists.

Again, I disagree. The shorthand for, for example, margins varies dramatically according to the amount of properties you pass it. I believe this falls into the same category. The point of a shorthand is to reduce the amount of characters or properties, and give the same result. I don't see an issue with different expansion rules, when the use cases are this simple.

If we wanted to make it easier to write grays, I'd be more in favor of
a gray(<number> | <percentage>) function.  However, it's relatively
easy to use the hsl() function to create grays already, with
"hsl(0,0%,<gray percentage>)".  50% gray is just hsl(0,0%,50%), which
is slightly more verbose than gray(50%), but much better than
rgb(50%,50%,50%).

I'm not against a Gray() or GrayA() function. True, you can achieve the same with HSL, but let's be honest, have you seen anyone in the wild using HSL other than with pre-compilers for calculated values? It's included for completeness, not for clarity.

What should 4 digit expansion do?  Should it expand characters in both ways (same channel and across channels)?  I think we can agree on a strong "No!" at such chaos.

Obviously a 4 digit expansion is nonsensical in this scenario.

I tend to think any benefit in space or typing here is not worth the
potential confusion for those reading style sheets written by
others.  (I'm particularly hesitant to add new syntax that's hard to
search for with search engines -- somebody who doesn't know what
hsl() is can search for it; that's a good bit harder in this case.)

I don't think it is hard to search for "hexadecimal colour shorthand". How would you search for the current incarnation of e.g. #123?

I support an 8-digit hex with an extra two hex places for alpha (#RRGGBBAA). I believe having a shortcut for this is extra confusing an unnecessary.

I too support this. I think it's strange that alpha has been represented as a decimal rather than a hexadecimal value in RGBA, while the other channels are hex. I don't see an issue in having a 4 digit shorthand for this property.

Thanks,

AK

On 3 Aug 2011, at 03:17, Alan Gresley wrote:


On 3/08/2011 9:51 AM, Tab Atkins Jr. wrote:

On Tue, Aug 2, 2011 at 4:44 PM, Alan Gresley<alan@css-class.com<mailto:alan@css-class.com>>  wrote:
I totally agree with Tab here regarding a 2-hexadigit variant. The expansion
rule is different and could be confusing to authors if any method was
spec'd. I'm also against having a 4-hexadigit as a shortcut for 8-hexadigit
(last two digits for alpha).

Bwuh?  Are you against an 8-digit hex variant as well, or just against
the 4-digit variant specifically?  If the latter, why?  4-digit hex is
expanded to 8-digit hex in exactly the same way that 3-digit expands
to 6-digit.

~TJ


I support an 8-digit hex with an extra two hex places for alpha (#RRGGBBAA). I believe having a shortcut for this is extra confusing an unnecessary.




--
Alan Gresley
http://css-3d.org/
http://css-class.com/
Received on Wednesday, 3 August 2011 15:21:48 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:43 GMT