Re: [css3-regions][css3-gcpm] Thoughts on Plan A and Plan B

On Feb 15, 2012, at 10:15 AM, Sylvain Galineau wrote:

> While I understand the pushback against a solution that requires scripting to instantiate region elements, I find solutions that implicitly drop dynamic styling and event handling on the floor to be at least equally problematic

We're in complete agreement on this point. My big issue with regions is that scripting and explicit elements are required even for simplistic use cases. Again, this isn't really the fault of the regions spec. It's just that we're missing the page templates piece, and I think it needs to be there to complete the feature. I don't want to ship one piece without the other.

Scripts and explicit elements just shouldn't be the only way to build content using regions.

I think relevant features of a page templates spec would include:

(1) The ability to define categories of pages, e.g., page masters, "header", "two-column page", "figure page", etc.
(2) The ability to select which master to use for a particular page.
(3) The ability to specify the layout of the regions on the page. Options for this include:
	(i) CSS multi-column
	(ii) ASCII art grid layout template
	(iii) Positioned slots (the .ibooks way)
	(iv) Shadow DOM
	(v) Explicit DOM
	I think that connecting to shadow DOM is pointless and that just making the elements explicit would be fine. Because the explicit elements would be building the "outer shell" of a page, i.e., all the containers for actual content, I think there is not much of an issue with them being explicit vs. shadowed. They wouldn't interfere with event handling on stuff inside the slots for example even if explicit, and you avoid a spec dependency on a whole other feature.

I also think it would be worthwhile looking into how to effectively script anonymous containers. We should have a page slot object model that facilitates easy scripting. That would cut down on the need for explicit elements except in the most complicated of use cases.

dave
(hyatt@apple.com)