Re: [whatwg] Knuth and Plass algorithm; boxes & glue & penalties

David Young wrote:

 > I'm wondering if anyone is developing web standards and prototypes  > for
web layout using the Knuth and Plass algorithm?  It seems  > like responsive
designs could be expressed more simply in a  > boxes-glue-penalties frame,
and responsive layout for JavaScript, CSS,  > and other things that you put
in <pre>, now, would be tractable with the  > right HTML/CSS primitives.  If
you have something like this underway,  > please get in touch.


On  Wednesday, April 08, 2015 6:56 AM Håkon Wium Lie replied:
[...]

However, it seems you're not primarily thinking of line-breaking, but box
stacking on a more general basis?

--------
I have discussed such things over the years with the SVG Working Group,
thinking that content like geographical information (like words or numbers)
being drawn to and repelled by associated geometric features (like lakes or
country boundaries) would make a lot of sense in terms of scaling of such
content. 

SVG,  has for now, relied on the simpler CSS box model, though models that
use simple physics are now more computationally feasible. 

Here's a relatively recent excursion into the domain of non rectilinear
layout: http://cs.sru.edu/~ddailey/TGW2014/RectilinearMold2.html 

Cheers
David

Received on Wednesday, 8 April 2015 19:17:35 UTC