W3C home > Mailing lists > Public > www-style@w3.org > February 2015

Re: [css-flexbox] Should "max-width" influence the resolved flex base size? (from default "flex-basis:auto;width:auto")

From: Daniel Holbert <dholbert@mozilla.com>
Date: Mon, 23 Feb 2015 16:57:14 -0800
Message-ID: <54EBCC6A.5070005@mozilla.com>
To: Alex Mogilevsky <alexmog@microsoft.com>, fantasai <fantasai.lists@inkedblade.net>, "www-style@w3.org" <www-style@w3.org>, Rossen Atanassov <Rossen.Atanassov@microsoft.com>
Hi Alex -- thanks for your input! I think you might be addressing
something that's different from what this thread is actually about,
though, and the spec's current requirements are not quite as clear as
your message implied.  I'd be curious for your thoughts, after taking
the following into consideration:

On 02/23/2015 03:38 PM, Alex Mogilevsky wrote:
> We went back and forth on whether max-width is applied once
> to the final result or if it is also honored when calculating
> hypothetical main size.

Sorry for being a language pedant (& hooray for term-renamings) -- I
have to clarify -- "hypothetical main size" is different from what I'm
asking about here.

I'm talking about the *flex base size*  (i.e. the length that we start
our flexing from, when we have an auto-sized flex item with
flex-basis:auto). I'm wondering whether we should ignore min/max-width
*when doing layout to compute this value*.

In contrast, the "hypothetical main size" (the term you used) is an
*explicitly-clamped* version of the flex base size, and it's only used
for a few things like determining where we should wrap.  Importantly,
it's *not* where we start our flexing from -- that's the "flex base size".

> March 2012 draft had max-size explicitly excluded from
> hypothetical size calculation (that's what I implemented
> in IE10). June 2012 draft and current spec have it affect
> hypothetical size.

Right, but the change isn't as significant as you're suggesting  See my
second message on this thread:

Basically, the relevant change between those two drafts was:
And that change simply renamed the old "hypothetical size" to "flex base
size", and it redefined "hypothetical size" to be an explicitly-clamped
value. (and to *no longer* be where we start our flexing)

The question is, should any min/max clamping happen when doing layout to
establish the "flex base size"? The spec doesn't say. (It used to,
before that hg revision, but the prohibition was removed there.) I think
that removal was accidental.

Received on Tuesday, 24 February 2015 00:57:47 UTC

This archive was generated by hypermail 2.4.0 : Monday, 23 January 2023 02:14:48 UTC