Re: [css3-flexbox] intuitivity and width computation rules

2011-01-13 09:29 EEST: Daniel Glazman:
> I can find at least ten web pages demo'ing flexbox with
> <div class="box">1</div> and showing three horiz boxes
> same height with 1/2 1/4 1/4 widths. The first thing all
> authors will try is replacing "1" by "lorem ipsum blabla"
> and that will fail...
> 
> Über-techies will say it's normal it fails. All the others will
> ask why. That's what we have to fix.
> And I also say "width: 0" is not understandable to specify the
> desired behaviour here.

I guess the basic problem is that authors of those ten demo web pages
have totally misunderstood the flexbox feature. Obviously I cannot speak
for those authors but I'd guess that flexbox works for what it's meant
for, but the specification fails to explain the intended usage to the
reader.

If a demo is trying to display three boxes with 1/2, 1/4, 1/4
distribution then such a demo should use normal boxes with percentage
widths. A demo that tries to explain flexbox usage should use example
content similar to UI widgets, as used in example by L. David Baron
(Message-ID: <20110113062439.GA22581@pickering.dbaron.org>) in this very
same thread.

I feel that flexbox is mostly suitable for web application UI
implementations and not so much for normal page content. However, this
definitely does not mean that flexbox should not be supported by major
user agents.

-- 
Mikko

Received on Thursday, 13 January 2011 09:04:48 UTC