Re: [css21][css-position] Better name for "pseudo-stacking context"

On 03/04/2013 22:28, Tab Atkins Jr. wrote:
> While trying to edit today's telcon decision into Flexbox, we realized
> that "pseudo-stacking context" isn't actually defined anywhere.  We
> then noted it was a terrible name for the concept.

I introduced the term as a tongue-in-cheek way to draw attention to the
fact that a fairly important concept existed in CSS21 without a name to
refer to it by!  In my treatment of stacking in CSS21 which became the
infamous Issue 60 on the old wiki issues list, I actually used other
terms, my preferred one at the time being "painting context" (so every
stacking context is a painting context, but not vice versa).

(My original request to define this term for CSS21 was rejected,
incidentally ;-)

> After some brainstorming between us and some coworkers, we've come up
> with these suggestions:
> 
> 1. Rename "stacking context" to "z-layer container".  The current term
> "stacking context" isn't very clear - invoking "stacking" doesn't
> suggest what is being stacked in any way.  The new name uses "layer",
> which *is* a term used more often by normal people when talking about
> these concepts, and it includes a reference to "z", because it
> traps/orders things using "z-index".
> 
> 2. Rename "pseudo-stacking context" to "flow-layer container".  It
> only contains flow-level stuff (blocks, floats, inlines), so this
> seems appropriate, and has nice parallel structure to the "z-layer
> container" term.
> 
> The spec should still define "stacking context" as a synonym for
> "z-layer container", so references into the spec still make sense.

I'm not thrilled by any of these, to be honest.  Although the term
"context" is used a fair bit in CSS without there being any concrete
definition of what it's supposed to mean, I suppose we've all developed
a feeling for it.  I do think "context" is better than "container"
because when dealing with painting/stacking contexts its often not the
ancestor-descendant relationship that's interesting; rather it's the
participation.  Which is why phrases like "participates in a stacking
context" are often used.

Cheers,
Anton