Hi Eric,
I think the original idea was to make the sass properties resemble ruby :symbols, which are often used as keys for a hash. (And a CSS property list is a hash of sorts.)
It’s not really the right syntax though, since there’s no separator between the key and the value like in :ruby => hash_syntax. It’s not like JSON or CSS either, so why reinvent the wheel and make CSS authors (regular designers to whom Sass is alien) learn an arbitrary new syntax?
I know we all got used to it, but I think changing it to be congruent with CSS is a good idea.
Regarding the use of the first character as a “type” identifier, let’s see, we have selectors like:
h2, #header, .hotpink, :first-child
These include both no-prefix for an element name (h2) and the colon-prefix (:first-child), so I don’t think there’s a solid argument for consistency here.
I think the colon after the property is a better idea and it will make it more obvious what’s a property, because nothing else in CSS looks like that.
There’s one or two other points I would add to this syntax debate, which is that mixing colons and equals= (for interpolated values) adds some mild confusion too. If it doesn’t slow things down too much, I’d be happier to see colons used throughout.
String quoting (where to use or not use double "quotes" and how they’re different from single 'quotes' in ruby is confusing too. Off the top of my head, single quotes get output into the CSS source and double quotes don’t, IIRC.
Andrew Vit
(Is the Sass group an ex-group, bereft of life, or just pining for the fjords? Is this discussion more appropriate in the Haml group?)