- From: Andrew Fedoniouk <news@terrainformatica.com>
- Date: Sun, 19 Oct 2008 12:05:31 -0700
- To: Brad Kemper <brkemper.comcast@gmail.com>
- CC: fantasai <fantasai.lists@inkedblade.net>, Mike Wilson <mikewse@hotmail.com>, www-style@w3.org
Brad Kemper wrote: > > On Oct 18, 2008, at 12:15 PM, Andrew Fedoniouk wrote: > >> "Named constants are in-scope from the point at which they are >> declared until overridden by a later declaration." > .... >> It appears as CSS constants in order to be used in parse time >> must obey following rules: >> >> 1) First seen - first used. >> >> First declaration of the constant establishes its value. >> Later [re]declarations of the same constant must be ignored. > > I agree. Whatever it is, it is not a constant if a later rule can > redefine it. > I think that main problem is in fact that different people have different ideas about design goals of the @const/@var thing. If the goal is to have declarative parametrization of CSS then you need @constant - parse-time entities. And if the goal is to provide runtime parametrization of CSS - so to be able to change value of some named entity at any time then we need @variables. It appears as most of CSS authors need @constants. As I said in previous message, @constant is a first-seen-first-used thing. It is simply not possible to have any other resolution schema in CSS for constants. If we want to specify @variables then we probably need to rethink the whole CSSOM idea. CSSOM and @variables have big area of overlapping use cases. -- Andrew Fedoniouk. http://terrainformatica.com
Received on Sunday, 19 October 2008 19:06:08 UTC