- From: François REMY <francois.remy.dev@outlook.com>
- Date: Mon, 28 Oct 2013 23:58:47 +0100
- To: "Tab Atkins Jr." <jackalmage@gmail.com>, Håkon Wium Lie <howcome@opera.com>
- Cc: "Alan Stearns" <stearns@adobe.com>, "Johannes Wilm" <johannes@fiduswriter.org>, "www-style list" <www-style@w3.org>
> > On the ladder of abstraction for document design, regions and flows > > are fairly abstract concepts that depend on lower-level stuff like > > fitting content on a line. If you truly want the low-level stuff, a > > better target is TeX's assembly-language-for-documents-like features. > > There are multiple levels of abstraction that one can target. The > fact that it's possible to go lower doesn't mean that it's not > reasonable to expose something at a given level. I've to agree here, with both remarks. Yes, Regions is not the lowest level feature we can imagine. Yes, we still can go deeper. And yes there are use cases for that, too. But, yes again, this is orthogonal to the current discussion. Changes happen to the platform by iterating, and not by suddenly giving you access to all the primitives the browsers kept hidden from you during years. We have to consider browsers as they are: products of an agile development process where new versions comes every X weeks, and that web developers have to work with every day. Planning large and idealistic features is going to bring us nowhere, because their implementation does not fit in a reasonable amount of time, they seem too far away, people will not feel compelled working on them. With low-level primitives, we can set intermediate goals that are appealing because they actually help solving real problems in the mean time, while still bringing us closer to our goals. Only when the basement is solid, we can think about building the house.
Received on Monday, 28 October 2013 22:59:18 UTC