RE: [css3-multicol] new editor's draft

I assume you are refering to this post from Tab:

http://lists.w3.org/Archives/Public/www-style/2010Oct/0670.html

Afaics, he didn't specify the multicol element, so I can not answer his
question with one answer, because I don't know what the height and width
constraints are, if any.

I don't see any constraints if we go back to Håkon's first example (which
btw did not have a {align:center} on the span), except that he constrained
it to 3 columns, so I don't understand why we are putting any thing in the
inline direction for overflow?

http://lists.w3.org/Archives/Public/www-style/2010Oct/0591.html

I can't even see what the spec will do when height is contrained, and
columns count is contrained, and then there is overflow?  It can't go into
inline direction because no more columns are allowed on that row. We can
not release this CR without block overflow.  We have to remove that inline
overflow always from section 8.2.


below...

Sylvain Galineau wrote:
> So I ask for disambiguation between two answers and now I get five based
> on this, that, the other thing and what if we put this here and that
> there...

I tried to answer you with examples that would show the variants better,
because you asked me why they shouldn't go in source order and I think
they are in source order, so I tried to show you with more examples how
they are in source order.

> What is your answer to Tab's question given the current design as you
> Understand it.

Nobody can with one answer, because it is underspecified.

> I'm not asking what you think it should be, why, or what
> it would do if the spec did this other things or if that element was
> instead
> before this other one, or if we expand this element like so...etc etc.

You stated that I was moving the span out of source order.  So in my mind,
you did implicitly ask for more examples, so that you could see I was not
moving it out of source order.

> What *is* your answer to Tab's question ?
>
>
>
>> From: Shelby Moore [mailto:shelby@coolpage.com]
>> Sent: Tuesday, October 26, 2010 5:44 PM
>> To: Sylvain Galineau
>> Cc: "Håkon Wium Lie"; Alex Mogilevsky; Tab Atkins Jr.; www-style@w3.org
>> Subject: RE: [css3-multicol] new editor's draft
>>
>> Sylvain Galineau wrote:
>> >> From: Shelby Moore
>> >>
>> >> So in my and I think the designer's mind, the correct result is:
>> >>
>> >> Inline overflow:
>> >>  baz----------------------->
>> >>  foo | foo | foo | bar | qux
>> >>
>> >> Block overflow:
>> >>  baz------------->
>> >>  foo | foo | foo |
>> >>  bar | qux
>> >>
>> >> Now isn't that a lot more sane and consist?
>>
>> Typo, 'consist' -> 'consistent'.
>>
>> > I sure don't know since I can't tell what your answer is. I see two
>> > different
>> > renderings ?
>>
>> One is for when overflow is in the inline direction, and the other is
>> for
>> when overflow is in the block direction.  Currently the spec only
>> allows
>> overflow in the inline direction, but I assume the spec will be
>> generalized in the future, so I showed both cases.
>>
>> > I'll assume, however, that in both cases you layout baz on top of all
>> the
>> > columns.
>>
>> Note I had an error for the block overflow case, and I was writing my
>> correction while you were replying:
>>
>> http://lists.w3.org/Archives/Public/www-style/2010Oct/0780.html
>>
>> Actually it needs to be corrected more, as baz should only span the 2
>> columns in the new column row:
>>
>> Block overflow:
>>  foo | foo | foo
>>  baz----->
>>  bar | qux
>>
>> > How and why does is that correct per the designer's mind ? Why would
>> he
>> > expect all
>> > the other elements to flow in source order except that one ?
>>
>> The spanned element is also flowing in source order.
>>
>> Pehaps what is confusing is in general foo, bar, and qux be spannable
>> lines.
>>
>> Let's assume instead with monospace font:
>>
>> foo = fff ooo
>> bar = bbb arr
>> qux = qqq uxx
>>
>> Inline overflow:
>>  baz----------------------->
>>  fff | fff | fff | bbb | qqq
>>  ooo | ooo | ooo | arr | uxx
>>
>> Block overflow:
>>  fff | fff | fff
>>  ooo | ooo | ooo
>>  baz----->
>>  bbb | qqq
>>  arr | uxx
>>
>>
>> Now assume bar = bar and is before baz.
>>
>> Inline overflow:
>>  fff | fff | fff | bar | uxx
>>  baz----------------------->
>>  ooo | ooo | ooo | qqq
>>
>> Block overflow:
>>  fff | fff | fff
>>  ooo | ooo | ooo
>>  baz----->
>>  bar | uxx
>>  qqq
>>
>>
>> >> Note it is incorrect to assume baz is centered,
>> >> that is why I drew it
>> >> as I
>> >> did on the left side with an arrow showing its extent.
>> >
>> > Uh ? The element's style attribute says text-align:center.
>>
>> I must have lost that from the original example. Apologies. Then yes it
>> should be centered. In any case, I am adding the arrow to show the
>> precise
>> extent of the span.
>>
>> > How do we
>> > ignore
>> > properties explicitly specified by the designer and claim the result
>> > aligns
>> > with his expectations ?
>>
>> I would not argue for that. Was merely not seeing the center attribute.
>> Let's try to keep examples simple, not need to include an extraneous
>> center attribute as it is not germane/relevant to this. Sorry I just
>> missed it.
>
>
>
>

Received on Wednesday, 27 October 2010 01:28:14 UTC