- From: Douglas Rand <drand@sgi.com>
- Date: Thu, 31 Oct 1996 13:19:44 -0500
- To: David Perrell <davidp@earthlink.net>
- CC: Style <www-style@w3.org>
David Perrell wrote: >... > Are you sure no algorithms already exist for this? Word allows frames > to be attached to paragraphs in such a way that the frame is positioned > at the bottom of the previous paragraph. Redraws when the column width > is changed is quite rapid. OK... I suppose some do. Now here's the more important question, is there a public algorithm? What I've noticed is that it is almost impossible to find useful algorithms for this kind of thing. Which forces me to invent my own. I don't imagine MS has published how Word works internally :) What happens when the element is taller than the previous paragraph? > It seems to me that the current spec will require this behavior anyway. > Consider that a floated element behaves just as the Word frame. Give a > floated element at the _top_ of its parent enough negative margin and > it will extent up into the previous element. Won't the text in the > previous element need to flow around the floated element while still > maintaining its vertical margin, just as is the case with the Word > frame? This is a little different. The parent container might also just put the next paragraph low enough to render to floated element without it impinging upon the previous element, somewhat as if the first line of the paragraph was very tall. I guess what I'm worried about is having deterministic algorithms. The current stuff always solves as you move forward. Placing things to they are tied to locations but imping upward into what will normally be previously rendered blocks is much more difficult and given that most of the blocks are paragraphs, it is somewhat iterative since the height of the block moves the floating element, which adjusts the size of the block again as it lays itself out again. Doug -- Doug Rand <drand@sgi.com> (508) 567 - 2217 Silicon Graphics/Silicon Desktop http://reality.sgi.com/drand Disclaimer: These are my views, SGI's views are in 3D
Received on Thursday, 31 October 1996 13:16:50 UTC