Resolutions 2008-03 San Diego Part II: Test Suites and Pending Publications
Test Suites
-
Resolved: When we get to a point where the CSSWG agrees the test suite is “good enough” for entering PR, we use that version (snapshot) of the test suite to enter PR, then continue to work on the test suite and errata.
Rationale: Question was “when do we stop working on the test suite?” We will always find deep technical issues. At some point we have to stop, publish the REC, and use the errata system from there. Some people argued that there should be concrete criteria for when the test suite is “done”, but no one offered any usable criteria. We have some tools to index the tests, but ultimately it will need to be a judgement call. It was further argued that work on the test suite should not stop just because we want to enter PR: the test suite will continue to be useful for driving interoperability and uncovering spec problems even after REC, and use of the test suite for PR transition should not prevent other uses of the test suite.
-
Once all of the tests written by Microsoft and by Hixie are reviewed and checked in, we will have a much better view of what remains to be done for the CSS2.1 test suite.
-
Noted: Specs should include a link to the official test suite, whether it is complete or not.
-
Resolved: The intent of the CSSWG is to post two copies of the test suite, one with the W3C Document license and one with the BSD 3-clause license. The copy with the W3C Document License will be the official conformance test suite.
Rationale: While the W3C Document License is fine for a test suite whose sole purpose is checking conformance claims, the amount of work that goes into this test suite does not justify limiting its use to that. Several major contributors do not want to contribute tests if these tests are not freed for other uses. These tests would be very useful in particular during the implementation development process, and the members of the CSSWG would like to make the tests in the test suite accessible for such uses. This requires placing them under a liberal open source license, and the BSD 3-clause license was chosen by consensus.
Mobile Profile
-
Resolved: Mobile Profile can move to CR as long as marquee and overflow are tagged as at-risk.
Resolved: The marquee and overflow properties are in the Box Module, which is still only a working draft. As long as they have not progressed to CR they block the Mobile Profile from reaching PR.
Media Queries
-
Resolved: em and ex units in media queries must be relative to the initial values of the font properties, not to values on the root element (since style sheets can change those).
-
Resolved: Media Queries shall be published as a Last Call Working Draft.
-
Resolved: The editor’s draft of Media Queries to be moved to dev.w3.org.
CSS3 Color
No resolutions were recorded. 🙁 See minutes for discussion; we went through dbaron’s issues list. Assuming dbaron kept track of the discussion, any WG resolutions should appear in the issues list.
This spec should be ready for Last Call soon.
CSS Namespaces
The chairs chose for political reasons not to make any resolutions. The CSSWG hopes to, with the XHTML2WG’s approval, morph the XHTML2WG’s request for removing default namespaces into a request for clarification/guidance on the use of default namespaces in conjunction with Selectors and resolve the issue by adding an XHTML2WG-approved note to CSS Namespaces.
CSS Namespaces issues
Resolutions 2008-03 San Diego Part I: Working Group Operations, Communication, and Charter
Note: The current CSSWG charter expires at the end of June. The new one must be written and submitted to W3C for approval by the end of April.
Note: Daniel Glazman and Peter Linss are co-chairing the CSSWG starting with this face-to-face meeting. Bert Bos remains as the CSS Working Group’s W3C Staff Contact.
Working Group Communications
-
Resolved:That the CSSWG’s IRC channel shall be perpetually and publicly logged. That these logs are for us and for the public to see our work in progress, and that they are NOT for the official record. That any important information in the logs MUST be summarized and forwarded to the mailing list (along with, but not substituted by, a pointer to the IRC logs and/or a pasted snippet) by the people involved in that discussion so that no one needs to track the IRC logs to keep up with the group.
-
Resolved: That the IRC logs shall have a disclaimer explaining that
- None of these comments shall be taken as representing WG decisions or as WG communication in any form: they are merely the WG “thinking out loud”.
- All of these comments shall be taken as “discussion in progress” and should not be taken as official statements even by the people who made the statements. The exact wording of this disclaimer will be decided later.
-
Noted: The CSSWG discussed opening up process-related discussions into a new public mailing list, but there was no consensus on that. The current private process / public technical split will be kept until further notice.
-
Resolved: Given that the IRC logs will be public, the CSSWG minutes shall also be public. (The CSSWG minutes are usually typed into IRC rather than into a text file.)
-
It may take awhile to set up perpetual logging. Public minutes will take effect starting on 2007-03-27.
-
Resolved: CSSWG Telecons have moved from noon in Boston on Tuesdays to noon in Boston on Wednesdays. They still last one hour.
-
Resolved: Members are expected to call in five minutes before the scheduled time. The telecon will start promptly on the hour.
-
Resolved: The agenda for the telecon is due the Monday before the telecon, specifically, before the agenda-writer goes to sleep that night or the sun comes up at his location the next day, whichever comes first.
-
Resolved: If a member cannot make the meeting, then s/he must send regrets by the day before the meeting (which should be after the agenda is sent) or risk bad standing.
-
Resolved: Participation over IRC is acceptable for telecons in exceptional circumstances. Participation over IRC means reading the minutes in IRC and participating back, not merely being online at the right time.
-
Resolved: First topics on agenda must be about resolving issues that require consensus: later topics can be more exploratory discussions.
-
Resolved: Attendance at first part of telecon is mandatory, where the first part is defined by the agenda and expected to take 20 minutes.
-
Noted: When a decision is made, we need to make sure it is recorded.
-
Noted: Issues should come to the telecon with a proposal and a testcase (if appropriate). Issues that need some exploratory discussion before a proposal can be written should be limited to 5-10 minutes discussion.
-
Noted: WG discussion should be kept at the technical level. We are not marketers, lawyers, etc. Our job is to push CSS forward.
Module Prioritization
-
Resolved: To stay in the charter, a module must have both
- an advocate who is willing to drive the module’s progress
- two implementors who report “strong interest” on the scale of “no interest”, “some interest”, “strong interest” in implementing the module
-
Resolved: The chairs shall prepare a prioritization report that explains our priorities for each module, indicates which modules have been cut from the charter, and explains the rationale for these decisions. The report shall include for each module whether it has an advocate and how many implementors have no/some/strong interest in implementing the module. This report will be posted to www-style for feedback.
-
Noted: Items may be added to our work list only if
- The CSSWG members unanimously agree that they are in scope for the current charter or
- The CSSWG agrees that they should be added and the charter is amended to include them.
-
Resolved: That Apple’s proposals for animations, transitions, and filters are in scope for CSS and may be added to the next charter if they pass the same criteria all modules must pass to stay in the charter.
-
Resolved: Add Web fonts to charter as coordinated effort with SVG group
Candidate Recommendation
-
Noted: The “at-risk” feature of CR status should be used more often.
-
Resolved: Publicly-available releases, whether of shipped, beta, or nightly status, may be used — at the CSSWG’s discretion — to qualify a spec for PR. Unofficial releases (betas, nightlies, etc) may only be used a minimum of one month after the build has shipped to give testers time to find any major compatibility problems due to the features implemented.
Primers
-
Noted: The CSSWG discussed and rejected the idea of creating a separate “primer” document parallel to the spec to explain rationale and give examples. Instead the CSSWG feels that the current practice of combining explanations, rationale, and examples into the spec itself benefits implementors, authors, and spec writers alike.
Rationale: Splitting this information into two documents
- makes both documents less clear
- is a lot of extra work that the CSSWG doesn’t need to take on
- forces the reader to cross-reference
- and risks the information getting out-of-sync.
Keeping the informative rationale and explanations in the spec
- makes the spec clearer to authors, educators, and implementors
- helps catch mismatches between the normative text and the spec-writer’s intention
- and by showing how a feature can be useful gives implementors a greater incentive to implement.
-
Noted: The above should not be taken to mean that the CSSWG wants less distinction between normative and informative content: the CSSWG encourages sharp distinctions between these two types of spec content.
Web Conferences
-
Noted: The CSSWG discussed the idea of scheduling a 2009 F2F in conjunction with a web conference. Several members expressed opposition to the idea on the grounds that aligning a CSSWG F2F with a web conference lends that conference an implicit association with the W3C and the W3C should be a neutral party. It was suggested that conferences could instead invite some individual members of the CSSWG, such as the chairs or the staff contact, to represent the group instead.
CSSWG Website Update
Jason Cranford Teague is working on a redesign of the CSS homepage. He presented a work-in-progress at the F2F.
Resolved: Jason will run a contest for a new CSS logo. Jason’s “no missing squares” CSS logo will be used until a new one is chosen.
Trackbacks Off
Trackback spam just jumped to one message every 3 minutes, and there’s no mass-delete interface on this thing, so I’m turning off trackbacks for now. (We’ve only gotten 3 non-spam trackbacks in the past 9 months anyway.) Please post links to any on-topic articles on the Soapbox instead.
Resolutions 2008-03-18
Multiple Text Shadows
-
Resolved: The listing of multiple text shadows shall be layered top to bottom (i.e. the first shadow is on top). This is a change from REC-CSS2.
Rationale: Changing the order to top-to-bottom makes text-shadow consistent with font, background, and the proposed content fallback extension. Multiple text shadow support is not widely implemented (we only know of Konqueror shipping support at the moment), and in most designs it won’t make much of a difference if the order is switched.
CSS2.1
-
Resolved: Issue 1 (clarify ‘width’ on columns and column groups) closed with No Change.
-
Resolved: Issue 10 proposal accepted to define handling of images without intrinsic sizes for background-image, list-style-image, and cursor.
-
Resolved: Issue 22 proposal accepted to clarify that display: none applies to non-visual media.
-
Resolved: Issue 34 proposal accepted to clarify that auto margins on absolutely-positioned replaced elements are not set to zero when calculating the width and height.
Other Topics
- Selectors, expressions, and whitespace was also (very) briefly discussed, but no resolution was recorded in the minutes.
- F2F topics
- test suite status
- css3-page status
Resolutions 2008-03-04, 2008-03-11
CSS2.1
Selectors
-
Resolved: White space is allowed in functional pseudos just after ( and just before ).
-
Preliminary conclusion on whitespace inside 2n+1 is no white space allowed; this may be modified if feedback indicates it will be a usability problem.
Other Topics Discussed
- Using Tracker to track issues and action items
- exit criteria for Mobile Profile CR
- JLTF‘s request for feedback
- rotation syntax conclusions from last F2F
Resolutions 2008-02-26
Selectors Level 3
-
Resolved: Proposal to normalize language strings associated with the document to 3066 before matching against :lang() accepted.
-
Resolved: Attribute substring matches against empty string (specifically, [attr~=""], [attr^=""], [attr$=""], [attr*=""]) considered invalid. Note: This resolution was later retracted.
-
Resolved: Spec needs to clarify that ::first-letter applies to content including generated content.
-
Resolved: Selectors must define that pseudo-class names are case-insensitive within the ASCII range (i.e. [a-z] and [A-Z] match each other, but match nothing outside that range).
CSS2.1
-
Resolved: absolutely-positioned replaced elements with auto width/height, both offsets specified, and no intrinsic size take their size from the constraint equation.
-
No resolution yet on what happens to replaced elements that have an intrinsic size; behavior here is more consistent, so need to check with other implementors (e.g. Mozilla).
New CSS module: CSSOM View Module
The CSS WG published the first draft of a new module, CSSOM View, with an API for inspecting the sizes and positions of rendered boxes.
CSS Namespaces enters Last Call
The CSS Namespaces module has been published as a Last Call Working Draft. The module defines the @namespace rule and makes the namespace features of Selectors possible to use in CSS.
Resolutions 2008-01-29, 2008-02-05
CSS2.1
-
Resolved: CSS2.1 Issue 27 proposal accepted to clarify interaction of scrollbars with width and height calculations.
-
Resolved: CSS2.1 Issue 33 proposal accepted to make intrinsic percentage widths consistent with specified percentage widths.
-
Discussed CSS2.1 Issue 31 about exactly when misplaced @import rules are ignored, next need proposed text.
CSS 2007 Snapshot (css-beijing) Status Update
Print and Paged Media Status Updates
- CSS3 Paged Media is close to done, expect a CR later this year.
- CSS Print Profile’s status is unknown, need to ask previous editor.
Resolutions 2008-01-15
CSS2.1
Resolved: that comma-delimited and space-delimited syntax for rect() may not be combined, see CSS2.1 Issue 28.
Media Queries
We discussed several media queries issues, including:
- syntax, where there seemed to be agreement to use CSS syntax minus escapes
- It was pointed out that media queries even in HTML are not always used in conjunction iwth CSS or by a UA with a CSS parser.
- It was pointed out that CSS has robust forward-compatible parsing rules that are useful to keep (e.g. matching brackets and handling quoted strings)
- It was pointed out that the same rules will need to apply to Selectors
- No one seemed to care whether comments were kept or not
calc(), where there was agreement that it shouldn’t be in media queries and several ideas were floated on how to do this:
- explicitly list which length units are allowed in media queries (awkward, hard to update)
- exclude media queries in the clac() definition in css3-values
- say in Media Queries that only simple lengths are allowed, not expressions
- Some other issues (but not all) are recorded on the wiki
Other Notes
![[Photo: group photo of the CSS working group in San Francisco]](/Style/CSS/csswg-2016.jpg)
Contact: