4 ms·
My journey: See this well trafficked "CSS sucks"-type article, Ctrl-F the comments for John Nagle, be tantalized at the hints of how it could be so much better
by CmdrKrool 13y ago
My journey:
See this well trafficked "CSS sucks"-type article, Ctrl-F the comments for John Nagle, be tantalized at the hints of how it could be so much better: http://www.raizlabs.com/blog/2006/09/25/ten-reasons-why-css-sucks/ http://www.raizlabs.com/blog/2006/09/25/ten-reasons-why-css-...
Look up "constraint-based layout" and more specifically "constraint cascading style sheets". You'll find papers such as this one: http://citeseer.ist.psu.edu/viewdoc/summary?doi=10.1.1.101.4819 http://citeseer.ist.psu.edu/viewdoc/summary?doi=10.1.1.101.4... and this one: http://csis.pace.edu/~marchese/CS835/Lec7/mmJournal.pdf http://csis.pace.edu/~marchese/CS835/Lec7/mmJournal.pdf thoroughly describing such a system.
Find "GSS - Grid Style Sheets" at http://gridstylesheets.org/ http://gridstylesheets.org/ , "an implementation of Badros & Borning's Constraint Cascading Style Sheets". This is still half on my todo list, but it seems like the real thing. The catch? Relies on Javascript.
Dream of a world in which this kind of system was built into browsers as the primary way of doing layout <insert Simpsons happy dancing people gif here>.
- cousin_it 13y agoBest comment in this thread IMO. Thanks! Some key parts from John Nagle's comments: It’s not that CSS is hard; it’s that CSS is bad. (...) The worst problem with DIV-based layout is that the layout system is too weak. (...) There’s no way to relate a DIV to anything but its predecessor, its parent, or an absolute position. (...) Table layouts allow table cells which span multiple rows and columns. If all tables could do were simple grids of cells, the CSS approach might make sense, but tables are more general than that.