Re: [CSSWG] Minutes and Resolutions 2009-02-04: box-shadow and border-image

On Feb 9, 2009, at 2:17 PM, "Robert O'Callahan" <robert@ocallahan.org>  
wrote:

> On Tue, Feb 10, 2009 at 10:40 AM, fantasai <fantasai.lists@inkedblade.net 
> > wrote:
> The problem with changing names of properties at this late stage is
> that people are already using the existing names in their style  
> sheets.
> This module is so well advanced implementation-wise that you see  
> things
> like
>  -webkit-border-radius
>  -moz-border-radius
>  border-radius
> in tutorials etc. I agree 'corner-radius' would have been a better  
> name.
>
> But no-one's shipped with "box-shadow" or "border-radius" yet, so we  
> could change the names without breaking actual content.
>
> And there are good arguments for 'border-shadow' instead of 'box- 
> shadow'.
> But I'm not convinced that these are strong enough reasons to change  
> the
> name at this point.
>
> I don't mind either way, as long as the Brad's "disable box-shadow  
> when border-image is specified" proposal is not adopted with the  
> name unchanged.

I find this attitude just astoundingly and unreasonably rigid. It's  
certainly nice when the name of something helps you understand it, but  
the most important thing is what you can actually accomplish with the  
thing. And having 'border' in the name doesn't do anything to help me  
understand how this drop-shadow-producing property works. In fact, I  
always thought the name 'border-radius' is confusing, since it can  
round the corners just as well without any borders. But if you changed  
its name to "corner-radius" (as I constantly do in my own head) I  
would not expect it to behave any different. You could change the name  
of "box-shadow" to "susan" if you want, and I would still want it to  
behave as I've indicated. Because ultimately the importance of the  
name is infinitesimally small compared to the importance of having it  
visible or not based on the usefullness of hiding or showing it in  
different circumstances.

Authors are not so easily confused by the variation of the name from  
some esoteric naming standard as you are suggesting. Results are what  
matter. Authors are used to getting the results they want regardless  
of the name. For instance, we commonly use 'zoom:1' to mean 'has- 
layout:true', and 'position:relative' to mean 'ie-bug-mode:alternate'  
or 'control-stacking-order:more', and 'float:left' to mean 'layout- 
mode:column', and 'margin-left: auto;margin-right:auto' to mean 'this- 
block-align:center', and '*HTML' to mean '@media (browser:ie6)'.

Throwing the word 'border' into some property names in order to give  
them some sort of loose logical association in the mind of an  
implementor doesn't actually make things clearer to a novice, who  
might expect such properties to only have an effect on visible  
borders. Whereas shadows and curved corners can be used effectively  
even when the border width is zero or the border style is none. And  
non-novices will learn what it's called, and just want it to work  
usefully.

Please. Picking the right name may have some importance, but making  
the property work in the most useful way is more important, and a  
separate exercise.