[CSSWG] Minutes and Resolutions 2011-09-14

Summary:
   - Discussed margin collapsing behavior of column-spanning elements;
     proposal A is to wrap an anonymous BFC around consecutive spanning
     elements with the same 'column-span' value so that the margins among
     them (e.g. between <h1> and <h2>) can collapse. Alternative is no
     collapsing.
   - Administrative topics: collecting availability data for Paris F2F,
     Shepherd feedback, blog transition to WordPress, spec publishing
     checklist, Anton Prowse has now joined CSSWG as Invited Expert

====== Full minutes below ======

Present:
   C�sar Acebal
   Bert Bos
   Arron Eicholz
   Elika Etemad
   Simon Fraser
   Daniel Glazman
   Koji Ishii
   Brad Kemper
   H�kon Wium Lie
   Alex Mogilevsky
   Edward O'Connor
   Florian Rivoal
   David Singer
   Alan Stearns
   Daniel Weck
   Steve Zilles
   <Zakim> +Andrew_Kirkpatrick

<RRSAgent> logging to http://www.w3.org/2011/09/14-css-irc

Scribe: arronei

Paris F2F
---------

   <glazou> http://www.doodle.com/7r9kee8rt5sqwqfv#table
   glazou: f2f meeting in paris
   glazou: please fill in the doodle
   glazou: need a good head count
   glazou: the best date for the f2f in Paris is between the 6th and 8th of Feb
   <dsinger> Week of 23rd is out I assume?
   <glazou> yes
   <dsinger> Bother. 3gpp and mpeg the two doodle weeks

column-span and margin collapsing
---------------------------------

   howcome: issue is how margin collapsing works on spanner
   howcome: current spec in CR does not say anything about that
   howcome: I suggest to add language for that
   howcome: it should happen if two consecutive spanning elements but not with
            other elements
   howcome: spanning lets are usually wider
   howcome: collapsing should be messy
   howcome: http://lists.w3.org/Archives/Public/www-style/2011Aug/0457.html
   fantasai: seems ok
   howcome: two h elements, they should collapse
   alexmog: spanning elements are kind of like floats
   alexmog: floats don't collapse margins
   florian: I see it differently
   florian: you intend things to collapse when there is only 1 column
   florian: they should collapse then
   howcome: yep, least surprise
   howcome: but what will happen column-span integer value
   alexmog: we are trying to optimize for cases that are rare
   alexmog: use cases that don't really exist
   alexmog: make sure elements don't have margins ?
   howcome: what if you column span all and margin 10em, does not look like
            it multicolor at all ; no collapsing then?
   howcome: very natural to have collapsing happen then
   howcome: not really insisting on this edge case
   <Bert> (Typical case in favor of collapsing: <div><h1><h2.subtitle><p>...</div>
           with 'div {columns: 20em} h1, h2 {span: all}')
   fantasai: I agree with howcome that would make more sense to have them collapse
   fantasia: two things make sense to me, implementation-wise
   fantasai: one is for each column-spanning element to be a formatting context
             root individually, and not collapse its margins with anything.
   fantasai: the other is to take all consecutive elements that have same column
             span value and wrap them in anonymous formatting context root, such
             that sibling spanning elements' margins collapse
   fantasai: and a spanning element's children would collapse margin with
             column spanning parent
   howcome: reasonable but what if there's a column-span: all followed by a
             column-span: 2, and the all winds up spanning two columns as well?
   fantasai: they don't collapse
   fantasai: it would be inconsistent for them to not collapse when the window
             is wide and collapse only when it's narrowed.
   fantasai: also, this way whether or not they collapse is calculated only from
             style data; no need to do layout
   howcome: fine solution to the problem
   howcome: but I would like to have alexmog with me on that
   howcome: msft has not done it that way
   alexmog: I did not have a chance to think about it before
   howcome: if we can avoid margin collapsing it would be nice here
   howcome: no decision yet
   howcome: need a week to review it?
   alexmog: need to get in touch with implementors
   howcome: fine
   glazou: add to next week's agenda?
   stevez: technical question: where there is only 1 column, do spaning elements'
           margin collapse with non spanning ones?
   fantasai: no, same reasons as above
   fantasai: default value of column-span?
   Bert: none
   howcome: yeah
   howcome: used to be 1 but changed
   fantasai: probably because of the margin collapsing issue
   <Bert> (I wonder if that means that we can replace 'overflow: auto' with
           'column-span: all' now in order to contain child floats in the
           element's height...)

Administrative
--------------

   glazou: Shepherd, please send feedback
   glazou: checklist for WD publications
   glazou: comments?
   http://wiki.csswg.org/spec/publish
   Bert: comment about the blog
   Bert: some people don't have a blog account
   Bert: new blog system next week, all accounts will be renewed
   Bert: if you don't have an account now, you'll have one next week
   Bert: if you have no account and have something to blog about, please
          wait until next week and the new system
   Bert: termporary state...
   fantasai: Anton as IE?
   glazou: Anton joined a few days ago as IE

Received on Thursday, 15 September 2011 00:11:43 UTC