W3C home > Mailing lists > Public > www-style@w3.org > April 2011

Re: [css3-flexbox] Best way to denote flexible lengths

From: Andrew Fedoniouk <andrew.fedoniouk@live.com>
Date: Fri, 15 Apr 2011 08:54:56 -0700
Message-ID: <BLU159-ds11CB3115858B19B8B39714F8AC0@phx.gbl>
To: "Alex Mogilevsky" <alexmog@microsoft.com>, "Brad Kemper" <brad.kemper@gmail.com>, "Tab Atkins Jr." <jackalmage@gmail.com>
Cc: "www-style list" <www-style@w3.org>
I've asked couple of times to provide practical examples for negative flexes
(a.k.a. base relative flexes). So far we saw one and very doubtful example 
from
Mozilla team. And that is it. I thought that I do not understand something
and I've implemented them. See [1], 2.2. "Relative flexes", the fx() 
function.

Developers use flex units a lot because of their very simple, intuitive and
predictable physical model see: 
http://www.terrainformatica.com/w3/flex-layout/images/flex-springs.png
But I do not see so far *any* practical use for the fx() (relative flexes).
(Observation is made on around 2000 developers that deal with flexes in
active development at the moment).

In any case as soon as you have flexes already then you can add that fx(), 
fr() or
flex() function any time later.

The most critical thing at the moment as far I can see is to establish 
common
infrastructure for flex units. So other layout modules can be built on this 
common
ground. In current circumstances any implementation of any two modules that 
use
different mechanisms of flexes is not that wise (with my polite Canadian hat 
on).

[1] http://www.terrainformatica.com/w3/flex-layout/flex-layout.htm

-- 
Andrew Fedoniouk

http://terrainformatica.com


-----Original Message----- 
From: Alex Mogilevsky
Sent: Thursday, April 14, 2011 10:57 PM
To: Andrew Fedoniouk ; Brad Kemper ; Tab Atkins Jr.
Cc: www-style list
Subject: RE: [css3-flexbox] Best way to denote flexible lengths

Flexbox spec draft always had negative flexibility. However it was defined 
same as positive (that is whatever grows faster also shrinks faster) which 
doesn't make sense usually.

I wouldn't agree that negative flex is less. It is dealing with cases of not 
enough space, and it is always difficult, so anything that helps create 
better experience in small space is valuable.

And of course for implementation dealing with negative flexibility is no 
more difficult than positive - same formulas, different numbers. So why not 
define it properly?

-----Original Message-----
From: Andrew Fedoniouk [mailto:andrew.fedoniouk@live.com]
Sent: Thursday, April 14, 2011 8:43 PM
To: Alex Mogilevsky; Brad Kemper; Tab Atkins Jr.
Cc: www-style list
Subject: Re: [css3-flexbox] Best way to denote flexible lengths

"Negative flex" you say?  Interesting. I'd like to try that too ...

Seriously: such simple concept as flex (portion of free space) shall not be 
that complex. And it isn't.

Can we just focus on simple flex units for now? And postpone this
nightmare(preferred,negative) for CSS4.
If someone will really ask for it?

--
Andrew Fedoniouk

http://terrainformatica.com




-----Original Message-----
From: Alex Mogilevsky
Sent: Thursday, April 14, 2011 12:25 AM
To: Brad Kemper ; Tab Atkins Jr.
Cc: www-style list
Subject: RE: [css3-flexbox] Best way to denote flexible lengths

Elsewhere I believe the order is not important if it is not ambiguous. Since 
we have to numbers they have to be in order - positive flex then negative 
flex if we want reasonable default.

Having good defaults is key here (just as it is with 'background' property)

flex(1) means (preferred=auto, positive-flex=1, negative-flex=0)
flex(auto) means (preferred=auto, positive-flex=1, negative-flex=0)
flex(0) means (preferred=auto, no flexibility)

if this is really the way to go, it should probably accept unitless zero 
length at certain priority (first or last... I think last)

-----Original Message-----
From: Brad Kemper [mailto:brad.kemper@gmail.com]
Sent: Wednesday, April 13, 2011 11:15 PM
To: Tab Atkins Jr.
Cc: Alex Mogilevsky; www-style list
Subject: Re: [css3-flexbox] Best way to denote flexible lengths


On Apr 13, 2011, at 5:29 PM, Tab Atkins Jr. wrote:

> On Wed, Apr 13, 2011 at 5:22 PM, Alex Mogilevsky
> <alexmog@microsoft.com>
> wrote:
>> I don't like the idea of width/height taking space-separated lists. I
>> would rather have a flex function.
>>
>> Given a choice between flex(1,0,auto) with commas and fixed set of
>> arguments and flex(auto 1 0) with space separated arbitrary order, I
>> think I would clearly prefer any-order version...
>
> Okay, then I'll change the draft to accept the 'fr' unit and the
> 'flex()' function with space-separated any-order arguments.  Sound
> good?

I'm confused. 'flex(auto 1 0)' is the same as 'flex(auto 0 1)' or 'flex(1 
auto 0)'?
Received on Friday, 15 April 2011 15:55:28 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:39 GMT