W3C home > Mailing lists > Public > www-style@w3.org > March 2009

The flow demo, Was: Re: Template Layout, implementation, some comments.

From: Andrew Fedoniouk <news@terrainformatica.com>
Date: Sat, 21 Mar 2009 18:20:35 -0700
Message-ID: <49C59263.80505@terrainformatica.com>
To: Bert Bos <bert@w3.org>
CC: www-style@w3.org

I've assembled small sample that demonstrates various
layout managers that 'flow' supports now. Namely:

flow: horizontal
       | vertical
       | horizontal-flow
       | vertical-flow
       | grid
       | "template";

Sample is located here:


File is of 787 kb in size (787,389 bytes on disk).

It has Windows executable, htmlayout.dll (the rendering engine)
and html sample files.

Unzip and run .exe from there. Button in the middle will load
demo suite that located in /content/ subfolder.

This demonstrates the flow including your "template" and flexes.
Nothing fancy but pretty cool, if you would ask me.

Bert Bos wrote:
> On Friday 20 March 2009, Andrew Fedoniouk wrote:
>> Here is a screenshot of the Template Layout in action:
>> http://www.terrainformatica.com/w3/template-layout.png
> That's quick!

I already had flow:grid in place. That is the same layout but with
different way of defining it.  So it was a matter to write parser for
your template "expression".

Please pay attention that template strings in my implementation
require white-spaces to separate placeholders.
This is made for cases when 26 lowercase and 26 uppercase letters is not
enough. So it is possible to use placeholders of two and more letters:

flow: "he he he"
       "lb ct rb"
       "lb ct rb";

> The 'border-spacing' of -1px seems a bit of a hack... :-)

:) If it is a hack then what are these negative margins then?

In principle border-spacing can be generalized for use in any container.
It establishes minimum value of margin between its children. Pretty 

> Normally, 'position' overrides 'float'. Is that still the case in your 
> implementation? Or would 'position: absolute; float: a' cause 
> the 'position' to be ignored?

float: "a"; and
position: relative | absolute;
can coexist peacefully.

Technically even float:right; and say position:relative; should work. 
Why not?

> Some of the things you ran into have been improved in the internal draft 
> that I maintain. But so far there seemed to be too little interest to 
> spent time on publishing a new version. For some reason that has 
> suddenly changed...
> Next week is a bad week for publishing working drafts (there is a 
> publishing moratorium because the W3C communications team is 
> unavailable) but I'll try to publish a new WD around a week from now.
> B.t.w., the WG discussed the "matrix" idea and its relation to the other 
> layout models a bit, but decided that the best way to discuss them 
> really was around a whiteboard. The urgency seemed low enough that it 
> could wait until the next opportunity, which is early June.
> So whatever comes out of these discussions, maybe a merger of the 
> various proposals, or a new proposal with the best features of all of 
> them in a new syntax, it's unlikely to be written up in a WD before 
> June.
> Bert

Andrew Fedoniouk.

Received on Sunday, 22 March 2009 01:21:12 UTC

This archive was generated by hypermail 2.4.0 : Monday, 23 January 2023 02:13:35 UTC