W3C home > Mailing lists > Public > public-css-testsuite@w3.org > April 2012

Re: CSS3 Flex Box & Multi-Column Test Case

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Sat, 14 Apr 2012 15:41:47 -0400
Message-ID: <35c29ad1138d9c839bd0956674fe4c6c.squirrel@ed-sh-cp3.entirelydigital.com>
To: "John Bilicki III" <jabcreations@hotmail.com>
Cc: public-css-testsuite@w3.org

Le Sam 14 avril 2012 0:53, John Bilicki III a écrit :
> Hello,
> The attached file's mime is application/xhtml+xml. It is a test case for
> Flex Box and Multi-Column Layout. The desired rendering occurs only in
> Firefox 13+. Opera 12 does not (at least currently) support Flex box.
> WebKit browsers (e.g. Safari and Chrome) have a bug
> (https://bugs.webkit.org/show_bug.cgi?id=83971) preventing scrolling.
> IE10 Preview 5(?) currently renders a horizontal scrollbar instead of a
> vertical scrollbar like Firefox 12 and older.
> I give permission to the W3C to use this test case as they see fit.
> I do not participate in emails due to spam concerns,


everyone and anyone would and should be concerned about spam. Always.

> a public forum
> should be made available in place of this which would add greater
> structure.

The current system, a mailing list, has features, search capabilities,
and it will arrange discussions in threads. I've used all those before.

Everything is always perfectible, of course. But I am not able to figure
out what you may mean with "greater structure".

I looked at your test and I would have many suggestions to your test.

line 7 * {margin: 0px; padding: 0px;}
line 8 body {background-color: #444;}
line 9 p {margin: 16px;}

A lot of CSS files reset a huge amount of properties.. just because they
do not know how these work... and then they re-reset those same
properties back to the values they already had by default in the user
agent stylesheet! Using the universal selector is also over-excessive
powering in the stylesheet and it may often have unwanted, unforeseen
side effects.
Choosing a dark background-color for body does not help the test or
achieve anything in the test.
By default, in the user agent stylesheet of all current mainstream
browsers, p elements use the following rule:

p {margin: 1em 0em;}

Furthermore, we can see that the main part of the test is given by
#image_manager which is absolutely positioned/anchored at 40px offsets
on all 4 sides: so, there is no need to "zero" margin and padding.


The stylesheet of your test has 13 declarations using vendor prefixes
(-webkit-, -ms-, -moz-): the CSS flex box module do not want this. You
say "Opera 12 does not (at least currently) support Flex box."  ... but
you have not use the -o- prefix either.

There is already a W3C test suite on multi-column layout module:


and it does not use any vendor prefixes.


line 69 <div id="image_manager">
line 71 <div id="moz_fixer">
line 72 <div id="image_browse">
line 89 </div>
line 90 </div>
line 91 </div>

>From experience, I'd say this looks like over-declaring containers
unneedlessly when they can be merged into 1. Although I must say I'm not
familiar with the properties involved in your test.


> The desired rendering occurs only

We prefer tests which are self-describing tests. Ideally, a test starts
with a sentence stating what are the pass/fail conditions of the test.


> If someone needs or wants to contact me please use the
> contact form on my site at jabcreations.com and I'll be happy to get in
> touch with them. Thanks!

Unfortunately, we do not work like this. The individuals who participate
in this mailing list do not have individual conditions and individual
restrictions about how/where to reach them, reply to them. If you post
something in a mailing list (discussion newsgroups, usenet, etc) that
can be viewed by all participants of a mailing list, then you should
expect replies to appear in such mailing list.

regards, Gérard
Contributions to the CSS 2.1 test suite:

CSS 2.1 Test suite RC6, March 23rd 2011:

CSS 2.1 test suite harness:

Contributing to to CSS 2.1 test suite:
Received on Saturday, 14 April 2012 19:42:20 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:13:24 UTC