W3C home > Mailing lists > Public > www-style@w3.org > May 2005

Re: Suggestion: Inheritance

From: Hans Meiser <brille1@hotmail.com>
Date: Thu, 19 May 2005 17:15:06 +0000
Message-ID: <BAY107-F31EF56CC3438724070B704E2080@phx.gbl>
To: www-style@w3.org

>From: Mikko Rantalainen <mikko.rantalainen@peda.net>
>To: www-style@w3.org
>Subject: Re: Suggestion: Inheritance
>Date: Thu, 19 May 2005 15:30:25 +0300
>
>Seems pretty limited to me. Is this any better than a single-pass 
>precompiler?

...limited? I guess you don't seem to recognize the full potential of this 
simple yet powerful extension to the standard.

It brings unprecedented flexibility and maintainability to CSS, particularly 
in multi-client environments where design templates could simply create a 
whole new design by just changing some basic rules.

Using values from another rule adds a flexibility which has never been 
available in CSS yet. Using it, if you change one design aspect in an 
existing web design, from now on you won't have to search and replace any 
rule referring to the same aspect. Back-referencing does all the job.

Moreover, back-referencing leverages object oriented design in CSS. Using 
back-references it is now possible to design reusable web controls in 
libraries and have them integrated in any web page, adapting to the current 
design of the hosting page without having the page designer bother with 
implementation details of the web control. It's a completely new branch 
becoming available to CSS design.

No precompiler can possibly be used or applied in scenarios like these.



>In addition, you'll run into problems because you decided to use CSS 
>selector as a key to reference to. For example:
>
>serif, fantasy /* elements 'serif' and 'fantasy' */
>{
>	font: "Arial Unicode MS" 1em;
>}
>
>fantasy.font /* special class here */
>{
>	font: "Garamond" 1em;
>}
>
>.myspecialclass
>{
>	font: serif, fantasy.font; /* is this replaced? */
>}

I suggest we should first basically agree on the advantages of the concept 
and accept it. After that we can discuss implementation details like this... 
I leave your question to the public. In any way, it doesn't change the 
concept at all.

I on my behalf would agree that "fantasy.font" should be replaced. As well 
as: "border: 1px solid body.color".


>Or
>
>p
>{
>	color: blue;
>}
>
>html:nth-of-type(1) > head + body div > table td > p
>{
>	color: red;
>}
>
>html:nth-of-type(1) > head + body div > table td > ul > li
>{
>	color: html:nth-of-type(1) > head + body div > table td > p.color; /* is 
>this supposed to work to get red instead of blue? */
>}

Same.



>It may be pretty easy to implement but if it has no power over a 
>single-pass precompiler, it's not worth it.

Now, this is *your* perspective... I see much more potential in it.

Back-referencing should be added to the standard so that it can be applied 
everywhere, on the fly, without compilation. It has minimum performance 
issues but maximum gain. - It belongs to the standard.


Axel Dahmen
Received on Thursday, 19 May 2005 17:17:21 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:37 GMT