Gecko/XUL Box Model: Difference between revisions
(Initial box model doc) |
|||
| Line 10: | Line 10: | ||
* need to think about vertical layout and how the same kind of algorithms can be used for horizontally scrolled columns and vertical text. | * need to think about vertical layout and how the same kind of algorithms can be used for horizontally scrolled columns and vertical text. | ||
Testcases should probably be in the form of reftest pairs... in order for this to work, we need to solve pixel-rounding errors that make the current tests unreliable. | Testcases should probably be in the form of reftest pairs... in order for this to work, we need to [https://bugzilla.mozilla.org/show_bug.cgi?id=366865 solve pixel-rounding errors] that make the current tests unreliable. | ||
=== Notes from bsmedberg === | === Notes from bsmedberg === | ||
Latest revision as of 23:19, 23 January 2007
An important aspect of developing the capabilities of the web is implementing a CSS flexible box model much like the XUL box model. Now that the reflow branch has landed, it should be possible to implement a box model that can be used reliably from HTML.
Hixie and Hyatt worked on an original spec, which was expanded and revised by Neil Deakin here.
Notes from dbaron
In order to be implementable, this specification needs significant additions, including:
- large number of testcases
- need to think about vertical layout and how the same kind of algorithms can be used for horizontally scrolled columns and vertical text.
Testcases should probably be in the form of reftest pairs... in order for this to work, we need to solve pixel-rounding errors that make the current tests unreliable.
Notes from bsmedberg
Need better description of how box layout will interact with paged (printed) media.
Notes from brendan
Possible integration with other (similar) technologies. Hand-wavy stuff here ;-)