CSS WG Blog Minutes and Resolutions June 2009 F2F

Minutes and Resolutions June 2009 F2F

MultiCol

• Resolved: Introduce page-break-* properties as syntactic sugar for new break-*.
  Rationale: Adding column-breaking controls to page-break-* doesn’t make much sense, but the WG wants to avoid introducing a second set of properties.
• Resolved: Håkon to add an example of a float intruding into previous columns and wait for implementors to complain.
• Resolved: Fix errors in pseudo-algorithm pointed out by Sylvain.

Paged Media

• WG not going to address problem of zero-height pages or columns, spec should merely suggest that UA make sure there’s a finite number of pages.
• Discussed wording on distinction between margins on forced vs. unforced breaks.
• Discussed whether to move break-* to css3-page

GCPM

• Håkon has revised the GCPM draft and trimmed it down to features that are mostly implemented by two UAs: Prince and Antenna House.
• Chris and Steve object to definition of cmyk() because it is not tied to a color profile and is therefore poorly defined. Chris to review in depth; possible to rename to device-cmyk() to make it clear that its colors are device-dependent.
• Resolved: text-replace is dropped from GCPM.
  Rationale: No WG consensus on keeping it, and much opposition besides.
• Suggestion to rename ‘border-parts’ to ‘border-clip’.
• Resolved: Move super-decimal counter style to CSS Lists Level 3.
  Rationale: That’s the right module for counter styles.
• Resolved: GCPM page-bleed renamed to bleed.
  Rationale: It’s already restricted to be inside @page.
• Resolved: Keep page floats in GCPM WD.

Grid, Templates, and other coarse layout systems

• Alex demonstrates an implementation of grid units and positioning.
• Discussion of interaction of abspos and floating behaviors.
• Discussion of multi-pass layout and problems created by mixing floats and arbitrary positioning.
• Discussion of snap-to-grid concept.
• Discussion of equal-height boxes.
• Discussion of flexible grid layout.
• Discussion of how to get a good report on layout requirements of web pages from designers.

Flexbox

David Baron explains the flexbox model, what parts have been implemented
by Mozilla and Webkit, and what parts are underdefined.

Resolved: Change Ian’s editor status and publish flexbox as FPWD.

CSS Backgrounds and Borders Level 3

• Resolved: Proposal to split border-image into subproperties is accepted.
  Rationale: It is easier to understand and allows independent manipulation of its various values, which is useful for e.g. :hover effects.
• Resolved: Middle part of border-image clipped by default, use ‘fill keyword to keep it.
  Rationale:
  • It’s a border property, should behave like one by default.
  • Gives ability to use formats without transparency as a border.
  • Improves performance by default: author turns off optimization only when needed.
• Resolved: Drop fallback color from background properties.
  Rationale: Replace with more general image() fallback syntax that allows multiple image formats to be specified as fallbacks, and is easier to understand.
• no-clip keyword dropped from background-clip.

text-overflow

• Resolved: text-overflow in case of vertical overflow still only applies to individual lines that overflow horizontally.
  Rationale: This solves a valid use case, so if a marker is needed to indicate vertical overflow, that will have to be a separate feature.

CSS2.1 Issues

• Discussed various ways of reconciling implemented @margin-box syntax with CSS2.1 error-handling rules:
• Resolved: @page takes a new construct that mixes declarations and @rules.
  Rationale: This limits the scope of parsing changes while still compatible with already-implemented (and deployed) @margin-box parsing.
• Resolved: css3-page to recommend that @margin-box rules come after all declarations in the page context.
  Rationale: Placing @margin-box rules after all declarations is compatible with existing error-handling rules.
• Discussed Steve’s testcases for ambiguities in vertical-align, Steve to write up two proposals for discussion.

Fonts

• Long comparison of bolder-lighter algorithms.
• Discussion of format()
• Resolved: Publish new WD of css3-fonts.
• John Daggett considers fill and stroke properties; fantasai and ChrisL suggest reusing the SVG fill and stroke properties.

SVG Coordination

• Discussion of SVG’s proposal to encode some of its attributes as CSS properties; suggested as a topic for TPAC.
• SVG comments on Transitions deferred to spec maintainer.

image-fit and image-position

• Tentatively Resolved: Merge image-fit and image-position into single fit property.
  Rationale:“image-fit” and “image-position” are not appropriate names for what SVG wants to use them for, and nobody had a better proposal.

Test Suites

• Various bits of infrastructure not set up, many confused about where to find unreviewed tests and what to do with them.
• Desire for zip file of tests.
• Need to set up cron job to build all tests, not just reviewed ones.
• Need to either organize or document svn repo, it’s a mess.
• People complaining about documentation being out-of-date (?) or hard-to-find.
• Various wishlist items proposed.
• Target date for adequate coverage set to Sept 15th.
• Discussion of Selectors test suite changes, new version published.
• Agreement to make svn repo (rather than W3C CVS) the master repository, W3C to make backups daily.

Reftests

Mozilla is proposing a new format for CSS tests. It can’t be used to test everything, but can make testing many things much simpler and easy to automate. The basic principle is that each test consists of two files, which must be rendered identically. A complex test can then be compared against a simpler page that has the same result. This is better than comparing against an image because it equalizes font availability, anti-aliasing, and other platform- and UA-specific aspects of CSS rendering.

Resolved: we accept tests in reftest format as well, as long as they have the required metadata
Rationale: Many tests easier to write, and especially, to run this way.

Molly’s Role and the CSSWG Website

• Discussion of Bert’s proposed homepage redesign based on Jason’s design.
• General agreement to assign liaison role to Molly; she will be responsible for demonstrating new CSS features, educating the public on developments in CSS, and collecting feedback from designers.

IPTV

Discussed IPTV liaison statement and our response to it.

Upcoming Meetings

• TPAC meeting is 2-3 Nov 2009
• Targetting 22-24 March, hosted by Apple in Bay Area
• Targetting 18-20 August, hosted by Opera in Oslo

Disposition of Comments Format

Discussion of formatting for Disposition of Comments document. Various authors have adopted different style, Daniel wants to consolidate.

Abandoned Working Drafts

• Resolved: CSS Math dropped, another WG is working on something related.
• Resolved: keep Style Attribute Syntax spec, but strip it down to only currently-valid syntax.
• Resolved: drop Hyperlink Presentation module, publish note to EOL it.
• Resolved: Replace CSS Introduction with Snapshot.
• Resolved: drop Reader Media Type module, publish note to EOL it.
• Resolved: Publish a new WD of CSS OM View.
• Resolved: drop Aural style sheets from list of work items.
• Resolved: drop CSS Scoping form list of work items.
• Resolved: Ruby spec on hold until editor is found.
• Resolved: CSSOM not dropped, just abandoned temporarily.
• Resolved: CSS Presentation Levels discontinued, publish note containing entire content of the current document.

Full Minutes

1. Wed Part I: Testing
2. Wed Part II: Scheduling, Website, text-overflow, DoC format
3. Wed Part III: Multi-col, Abandoned Drafts, Splitting border-image
4. Thurs Part I: GCPM, Flexbox
5. Thurs Part II: vertical-align ambiguities, Paged Media
6. Thurs Part III: CSS2.1 Grammar Issues, IPTV liaison
7. Fri Part I: SVG Properties, Fonts
8. Fri Part II: Grid, Templates, and other coarse layout systems
9. Fri Part III: CSS3 Backgrounds and Borders

« Previous article Next article »