Re: [WSG] [OT] Google search/index/webmaster help

2009-11-01 Thread Philippe Wittenbergh


On Nov 2, 2009, at 6:58 AM, Adam Smith wrote:

Swami; I'll hazard a guess here and assume you're using Firefox; and  
you've done what I did and gone tohttp://maps.unimelb.edu.au/sitemap.xml 
, seen a mass of test on screen


Because that file is being served as 'text/html' instead of 'text/xml'  
as it should. That is server misconfiguration. I'm not surprised  
Googlebot doesn't pick it up.

Safari shows the same issue.

Philippe
---
Philippe Wittenbergh
http://l-c-n.com/







***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: memberh...@webstandardsgroup.org
***



Re: [WSG] how to set table column widths with CSS

2008-01-11 Thread Philippe Wittenbergh


On Jan 11, 2008, at 11:51 PM, Alastair Campbell wrote:


On Jan 11, 2008 7:29 AM, Philippe Wittenbergh [EMAIL PROTECTED] wrote:

col:first-child {width:10em;}
col:first-child+col {width: 5em;}


From memory, I don't think col has children as such, it's quite a  
headache:

https://bugzilla.mozilla.org/show_bug.cgi?id=915


I'm styling the col element, not a descendant or child of col  
(there are none, anyway).

(col:first-child applies to the first column, child of colgroup)

width applies perfectly to the col element.
http://www.w3.org/TR/CSS21/tables.html#columns


To anyone who mentioned that IE (6) doesn't support these selectors,  
I **know**.
And that was also the point of my post. The OP wanted a solution  
without the use of classes (or a minimal use of them).


She could attach  class to each col element, eventually.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] how to set table column widths with CSS

2008-01-10 Thread Philippe Wittenbergh


On Jan 11, 2008, at 2:59 PM, Katrina wrote:


Gday,

Can someone please remind me how to set the width on a simple table  
column without suffering classitis?


Doctype: HTML4.01 strict. Must validate.


What is wrong with classes ?

else:
col:first-child {width:10em;}
col:first-child+col {width: 5em;}
...

or
td:first-child {width:10em;}
td:first-child+td {width: 5em;}
...

But: congratulations! your stylesheet just took a beating as far as  
performance goes.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Float-less layouts

2008-01-07 Thread Philippe Wittenbergh


On Jan 8, 2008, at 1:48 AM, Thierry Koblentz wrote:


I'd appreciate any comment that would help me improve this article:
http://tjkdesign.com/articles/float-less_css_layouts.asp

Demo:
http://tjkdesign.com/articles/css-layout/ 
no_div_no_float_no_clear_no_hack_no

_joke.asp


Nice write-up.
One of the issues with this technique: you can't use the 'columns' as  
a containing block for absolute positioned elements.
Another issue: width on a 'table-cell' is more like 'min-width' than  
'width'. The cell can expand in width if it contains e.g. long  
unbreakable text strings (or strings of text with white-space:pre).  
This can eventually be controlled by wrapping such content in a  
'overflow:auto' wrapper, but not always.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Site Check

2007-12-20 Thread Philippe Wittenbergh


On Dec 21, 2007, at 7:30 AM, CK wrote:


http://working.bushidodeep.com/kevon/index.html


The footer (#header) that covers the scrollbar is quite disturbing.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Browser test: Construct

2007-12-09 Thread Philippe Wittenbergh


On Dec 9, 2007, at 4:37 PM, Christian Montoya wrote:


Is
there any way at all to prevent the FAYT behavior? If I apply return
false, will that work?


I don't know :-) Nobody ever managed to block FAYT, seen from here.

To set FAYT to work automatically, add
user_pref(accessibility.typeaheadfind.autostart, true);
in your user.js
(or use about:config).

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Browser test: Construct

2007-12-09 Thread Philippe Wittenbergh


On Dec 10, 2007, at 3:36 AM, Christian Montoya wrote:


To set FAYT to work automatically, add
user_pref(accessibility.typeaheadfind.autostart, true);
in your user.js
(or use about:config).


OK, I added return false; to the keybindings and successfully
repressed FAYT in FF on WinVista, but I don't know if that works in
all gecko browsers.


Gecko is Gecko :-)
I checked your page with Fx 2 and the nightly builds of both Camino  
and Minefield: Fayt is not triggered.
Camino 1.5 appears not to 'get' it here, Fayt is triggered anyway.  
I'll check if there is a known issue/limitation around that. But if I  
were you, I wouldn't worry about it.


There are a few known bugs with keybindings in current nightlies on  
Mac, but those should be fixed by the time Gecko 1.9 is released.  
None affected your test page.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Browser test: Construct

2007-12-08 Thread Philippe Wittenbergh


On Dec 9, 2007, at 11:49 AM, Christian Montoya wrote:

I just created a layout tool: http://lab.christianmontoya.com/ 
construct/


[...]

If I could get some tests on older browsers and other operating
systems, that would be great. Thanks in advance!


Gecko nightly builds (Camino  Minefield) seem to do what you expect.
Of course, using 'J', 'K', 'L' triggers FAYT (Find as you type) when  
set to start automatically. Most people reasonably versed in the use  
of Gecko browsers have it set that way.


Safari 3 doesn't get your keyboard input.

OS X 10.4.11.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] padding on the bottom ;-)

2007-12-07 Thread Philippe Wittenbergh


On Dec 7, 2007, at 4:31 PM, Taco Fleur wrote:


Having the same issue with another layout, but in FireFox 
http://www.clickfind.com.au/ the 3 panels at the bottom of the page.


The issue is also visible in Safari and Opera 9.5b, if what you mean  
a a kind of vertical lines under those 3 boxes.


That comes from the margin on the p that is collapsing through the  
divs
(div#box-1-c p, div#box-2-c p, div#box-3-c p {... margin-bottom:  
0.5em;})

Set the margin to 0 and that issue disappears.

That is the correct behaviour, btw.



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Font-size-adjust (was: RE: Disabling Fonts in Font Stacks)

2007-12-02 Thread Philippe Wittenbergh


On Dec 3, 2007, at 6:05 AM, Terrence Wood wrote:


If, in laymans terms, font-size-adjust allows you to specify the
font-size based on the x-height of a preferred font-family, how is a
rendering engine supposed to deal with this if said font is missing?


Font-size-adjust works based on the first font specified in the list  
for font-family, doesn't matter if the font exists on the system or not.


example.
p {font-family: 'Lucida Grande', Arial, sans-serif; font-size-adjust: 
0.54;}
On Win OS, Lucida Grande is not present (OS X only font), the size of  
the text will nevertheless be adjusted (slightly enlarged) as Arial  
has an aspect ratio of 0.52. This is interesting in the case of a  
stylesheet that specifies 'Verdana, Arial, sans-serif', and where the  
page author bases all his work on the size of Verdana (we all know  
that this is big font, with rather large aspect ratio). Verdana at  
12px maybe or not be readable to the user, Arial at 12px looks and  
feels much smaller, and will probably be harder to read for the user.



And how does it resolve line-height issues for fonts that have a low
aspect ratio?


Could you clarify ?
Line-height depends on the used value for font-family, basically.



Personally, I would like to see some decent column support before
trying to exert this degree of control on font-sizing.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] RE: Disabling Fonts in Font Stacks

2007-11-30 Thread Philippe Wittenbergh


On Nov 30, 2007, at 10:42 PM, Frank Palinkas wrote:


How can we encourage the other (non-Mozilla) browser manufacturers
to add support for this very useful feature?


If I understand your question, please download, install, and have a  
look at

Opera (release version 9.4).

9.24 ? or 9.5b ?


The tool I think you're looking for can be found by going to:
Tools/Preferences/Advanced/Fonts. This opens the extensive Desired  
font

selection dialog box. Please have a look?

Apple menu  PreferencesAdvanced tab. Not everybody uses windoze.
That is about user preferred settings (aka preferences). You can  
change those in any browser worth it weight in bytes. It won't  
necessarily override the fonts set in the user stylesheet.


The question I answered to was about font-size-adjust, a css-3 property.

BTW - On the (original) subject of testing the effect of various  
fonts, on OS X this is very easy: Open Fontbook.app and just disable/ 
enable the fonts you want. For some browsers, you'll need to restart  
before the changes take effect.



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] RE: Disabling Fonts in Font Stacks

2007-11-30 Thread Philippe Wittenbergh


On Nov 30, 2007, at 5:31 AM, Kroon.Kurtis wrote:


I'm also late to the party, but I wanted to add something that no one
else seems to have mentioned -- font-size-adjust (from CSS 2, but
removed -- alas! -- from CSS 2.1 due to lack of support).

It is part of CSS 3
http://www.w3.org/TR/css3-fonts/#font-size-props


...

Now, the (rather substantial) disadvantage:

As far as I can tell[3], font-size-adjust is only supported in
recent Gecko browsers (e.g. Firefox 2).
And for Firefox 2.0.0.x, only on Windows. But fear not, It is  
supported on all platforms in the upcoming Firefox 3.0. Cool.


It is a very useful feature to produce decent quality text. Not only  
for the described case above, but also for example for the case where  
you have different fonts used within the same block of text (e.g. a  
paragraph with some code in).
(decent quality text: because typography on the web is still a far  
cry from what it should be).


Which leads me to my last question (on this topic, anyway):

How can we encourage the other (non-Mozilla) browser manufacturers
to add support for this very useful feature?

1. By using it ?
2. File bugs against your favourite Rendering Engine
(I don't need to, It is already supported in my browser...:-))

for WebKit, it is bug 15257
http://bugs.webkit.org/show_bug.cgi?id=15257
To my surprise, nobody had requested that when I filed that bug.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] RE: Disabling Fonts in Font Stacks

2007-11-30 Thread Philippe Wittenbergh


I wrote:


Apple menu  PreferencesAdvanced tab. Not everybody uses windoze.

Darn stupid mistake
Opera menu  PreferencesAdvanced tab.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Input tag - closing tag optional?

2007-11-25 Thread Philippe Wittenbergh


On Nov 26, 2007, at 8:00 AM, Dusan Smolnikar wrote:

I'm afraid browser don't agree with this, though. I'm not sure  
about input but I'm
positive that div/div is not the same as div / as far as  
browser rendering

goes.


Actually as far as XML (and consequently XHTML) is concerned:

input type=text name=a value=a/input

Is the same as:

input type=text name=a value=a /


Depending on the mime type there is a huge difference.
text/html -- sgml parser in use
application/xhtml+xml -- xml parser is in use.
And no, slamming an xhtml doctype at the top of the page has no  
effect (doesn't affect the mime-type).


Note: the upcoming Firefox 3 and other browsers using the Gecko 1.9  
rendering engine are more strict in handling this.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] opera doesn't play nice with Opacity

2007-11-13 Thread Philippe Wittenbergh


On Nov 13, 2007, at 10:06 PM, James Ellis wrote:

I still find it a bit odd that setting opacity on an outer box  
affects the
opacity on the inner box. For instance when you have nested white  
boxes each
with an opacity of 0.5 over a black background they both end up  
having 50%

black backgrounds (~ #7f7f7f).


Opacity is not a function of background-colour, it affects the whole  
box _and_ its descendants. That is, everything that is inside that  
box will see its opacity lowered.



If they are positioned absolutely over each
other the the opacity works as expected, the top most one having the
expected background colour of #bfbfbf.
That would be the case if none of the boxes are descendants of each  
other.


If you are looking to have a semi-transparent background, you have to  
look for rgba or hsla colours.
These colour values are only supported by WebKit/Safari 3.0beta + and  
Gecko 1.9a (what will be Firefox 3.0/Camino 2.0/SeaMonkey 2.0). I've  
been told that Konqueror 4.0 will also support this (QT4).
You can of course use rgba or hsla colour for any property that takes  
a colorvalue as value: border, background, etc.


Defined in the css3 color module
http://www.w3.org/TR/css3-color/#rgba-color
http://www.w3.org/TR/css3-color/#hsla-color

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] opera doesn't play nice with Opacity

2007-11-12 Thread Philippe Wittenbergh


On Nov 13, 2007, at 10:51 AM, Tee G. Peng wrote:


body {background:#FFB701}

where footer wrap has opacity of 0.92 with a bg color

#ftr_wrap {background:#0d0d0d;
opacity:0.92;}

and 3 columns has

#col1, #col2, #col3 {background:#1a1a1a;}

I think Safari and Firefox got it right, that the 3 columns'  
elements are 'inheriting' the opacity from the #ftr_wrap, which is  
what I wanted.


But Opera doen'st seemed to agree.


Safari and Gecko browsers are doing the right thing.
Opera 9.2 is trying something that looks like, but the resulting  
colour differences between the background-colour of #ftr_wrap and the  
background-colour of the 3 columns is really minimal. After going  
over the whole thing with a colour sampler, I could 'see' the  
difference :-) (that is : the OS X colour-picker saw it, not my eyes).
What Opera 9.5 beta is trying is puzzling. The whole footer has a  
yellow background. No trace of any background-colour as specified in  
the style-sheet, not on #ftr_wrap and not on #col1, #col2, #col3.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Request possible?

2007-11-06 Thread Philippe Wittenbergh


On Nov 6, 2007, at 7:48 PM, Gunlaug Sørtun wrote:


http://www.gunlaug.no/tos/alien/mv/test_07_1106.html

...

Otherwise: no problem.


No problems ? With Fx Mac and Safari 2.04, 3.03beta and the latest  
WebKit builds, the is a slight overlap: the top part of the floated  
block is partly covered by the text outside of it. In Gecko the  
floated block is under that text, in WebKit, the floated block is on  
top.

Depending on your font-size, this is more or less pronounced.

(It is a known issue/bug with that kind of trick, btw; the floated  
block is placed slightly too high in the flow).


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] SilverLight

2007-10-29 Thread Philippe Wittenbergh


On Oct 30, 2007, at 3:21 PM, Michael MD wrote:

The Silverlight build process produces a .dll. You need the  
Silverlight
plug-in to render the resulting html page. Also, from a quick  
test, it will
only render in IE and the Gecko range. Forget Opera, Safari for  
windows, etc.


If they are really serious about getting people to use Silverlight  
they need to provide the plugin for a wide range of popular browers  
and operating systems.. (definately including Safari, Opera, Mac  
browsers, Linux browsers, etc )
...until then I can't see why I should bother with it...  (of did I  
miss something?)


Well, the release version of the plugin actually works in Safari (and  
WebKit builds) and Gecko Mac (SeaMonkey, Camino, even Firefox). iCab  
and Opera (Mac, 9.23 and 9.5 beta) are not supported.


In the beta process, they were doing some flipping browser detection  
**from within the plugin**, and only checked for Safari or Firefox,  
as opposed to check for Gecko.
The demos I've seen still only work half and half on Mac browsers,  
except Firefox 2.0.0.x and Safari.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] skip to content: care of accessibility causing usability

2007-10-27 Thread Philippe Wittenbergh

On Oct 28, 2007, at 11:44 AM, Tee G. Peng wrote:


... so I came out with this technique:
teesworks.com/ (move your mouse to the top to see the result).

Haven't show it to client yet. Been working on this site in the  
last 2 days, I find that I am getting so  annoyed by the surprise'  
everytime the hover pops up. There is no way to miss it everytime I  
move the cursor to the top.


I agree with you: the 'hover' technique is way more annoying, and it  
will annoy way more people.


'skip links' should be visible all the time, as they are useful for  
sighted users  (e.g using the keyboard).
A compromise solution I have used: when a client doesn't want them,  
hide them (position them of screen, not display:none), but make them  
visible when those links are focussed (by pressing the tab key).


a.skiplinks {position:absolute; left: -999em;}
a.skiplinks:focus,
a.skiplinks:active {left: 1em;}
:active state is for iExploder. Add additional styling to taste.

Not really perfect, just a compromise.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] browsers render differently with Optroup

2007-10-24 Thread Philippe Wittenbergh


On Oct 24, 2007, at 3:27 PM, Tee G. Peng wrote:

I am working on a web form that has Optgroup in it, and the first  
time I realized browsers render this attribute differently.


IE Mac displays 'CA' in your 1st example
IE 7 Win displays 'CA' in your 1st example
Opera 9.5 alpha: idem ditto.


I made an example page with markup copied from the above page
http://lotusfromthemud.com/option.html


Under 17.6.1 it says (specifically for label in option):
http://www.w3.org/TR/html401/interact/forms.html#adef-label-OPTION

label = text [CS]
This attribute allows authors to specify a shorter label for an  
option than the content of the OPTION element. When specified, user  
agents should use the value of this attribute rather than the  
content of the OPTION element as the option label.


That sounds, to me, as validating what  Safari, IE Mac, IE Win Camino  
are doing.

Note that Firefox is not wrong by the description given above.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] browsers render differently with Optroup

2007-10-24 Thread Philippe Wittenbergh


On Oct 24, 2007, at 10:37 PM, Tee G. Peng wrote:


Under 17.6.1 it says (specifically for label in option):
http://www.w3.org/TR/html401/interact/forms.html#adef-label-OPTION

label = text [CS]
This attribute allows authors to specify a shorter label for  
an option than the content of the OPTION element. When specified,  
user agents should use the value of this attribute rather than  
the content of the OPTION element as the option label.


That sounds, to me, as validating what  Safari, IE Mac, IE Win  
Camino are doing.

Note that Firefox is not wrong by the description given above.



I am sorry I only understand this thing half.

...

Can 'value' be removed?  Everytime I work on a web form, the same  
qeustion kept throwing back to me. Is there any attribute one can  
saftly remove without causing browsers, screen reader and form  
script suffer?


label and ID are needed, Name also needed for the form script I  
use; this leaves the 'value' which I never able to decide the  
*value* of its usage.


I really find a markup like this is too much
label for=myformMy form /label
input name=myform id=myform value=myform


I think you are confused :-)
There is a difference between label the html _element_ and label  
the _attribute_ on option and optgroup


element:
label for=myformMy form /label
input name=myform id=myform value=myform

attribute
select id=myselectoption label=mylabel value=myvaluestring  
in option/option/select


The 'value' attribute on option (and other form controls) is required  
for back-end processing (so that your script knows what to do with  
all the stuff the form feeds it). The 'label' attribute is always  
optional and _can_ be used by the UA for display purposes. 'name' is  
also required to identify the form control.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Encoded mailto links

2007-10-18 Thread Philippe Wittenbergh


On Oct 19, 2007, at 5:19 AM, Dejan Kozina wrote:

Anybody (Mac  Linux browsers...) wants to take a ride? The thing  
is up
there at http://www.kozina.com/mailtest/ . Let us know of your  
results.


running OS X 10.4.10 + Mail.app

* WebKit latest build + Safari 3.0beta + Safari 2.04 : opens a new  
mail message; page stays where it was (no redirect).


* Camino 2.0a1pre/Minefield + Fx 2.0.0.8pre: open a new message/  
redirect to a page [1].


* Opera 9.5 alpha: bugs me about choosing an email client, but then  
opens a new mail message and stays on the page (no redirect).


* iCab 303: opens a new mail message and stays on the page (no redirect)

* IE 5.23 Mac: opens a new mail message, no redirect.

[1]
http://www.kozina.com/mailtest/example/com/me
Found
The document has moved here.

I don't know enough about 302 code to know if this could be a correct  
behaviour or not.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Catch 22 list problem

2007-10-13 Thread Philippe Wittenbergh


On Oct 13, 2007, at 10:10 PM, Jens Brueckmann wrote:

I prepared a short example at http://lairx.de/071011/numbering- 
lists.html


Triggers strict mode in Firefox - what about MSIE, Opera, Safari,  
etc?


Opera 9.23/Win32+Linux and MSIE 6 SP1 both render in strict mode as  
well.

I do not know about Safari though.


Safari, WebKit, Konqueror 3.5.7 : Strict mode.
iCab: I'm not sure. Its error console says 'unrecognised doctype',  
and my bookmarklet doesn't work in iCab. But I think Strict mode  
(that would make sense for an unrecognised doctype).


IE Mac: Quirks mode.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Catch 22 list problem

2007-10-11 Thread Philippe Wittenbergh


On Oct 12, 2007, at 4:42 AM, Keryx Web wrote:

A. li value=3 is not allowed in strict HTML 4/XHTML 1.0, Spec  
says use CSS.


B. I want to start at 3.

C: CSS has no means to specify a start value!

Pick your poison:

1. Invalid code
2. Use a transitional DOCTYPE
3. Set value with DOM-script


I'll choose: 1. Invalid code. No doubts.

The start attribute also makes a come-back in html 5:
http://www.whatwg.org/specs/web-apps/current-work/multipage/section- 
lists0.html#lists0
http://www.whatwg.org/specs/web-apps/current-work/multipage/section- 
lists0.html#start0
iirc, there was also a discussion in the html-wg to bring the start  
attribute back in the errata's for html 4.01, but obviously nothing  
came to that, and I can't find it in the archives.


List numbers are 'content', after all is said and done. Browser- 
makers are moving towards using CSS 'counters' to generate the list  
numbers as an internal mechanism, but it does not take away that  
those numbers are content.
(and  CSS counters is a very attractive mechanism to control/alter  
the presentation of those numbers).



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Safari problem

2007-09-27 Thread Philippe Wittenbergh


On Sep 28, 2007, at 2:43 AM, Bas V wrote:

So far no problem... however, in Firefox, Internet Explore etc, the  
y-scroll bar shows on PC and Mac but in Safari it doesn't show the  
box's scroll bar, why


Here is the code I am using:

In the style sheet:

#photos {
...
overflow-y:scroll;
overflow-x:hidden;
}


Safari 2.0 and older do _not_ understand overflow-y and overflow-x.  
It is not part of CSS 2.1, but in a draft CSS 3 module.

Afaik, Opera 9.23 and older do not understand that either.

You could try:
#photos {
overflow:auto; /*  providing fall back for older browsers */
overflow-y:scroll;
overflow-x:hidden;
}

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] XHTML+Voice

2007-09-07 Thread Philippe Wittenbergh


On Sep 8, 2007, at 9:32 AM, Tee G. Peng wrote:

It's showing up on Opera site, so I reckon it's supported for 9.50  
Alpha? And Safri Beta 3 too?


http://dev.opera.com/articles/view/xhtml-voice-by-example/


That page doesn't work as described on Safari, seen from here (latest  
WebKit nightly build -OS X) and Safari3.03beta - both OS X and WinXP).


But Mac OS X 10.4 (and esp. Safari) has build in screen reading  
technology: VoiceOver.
Turn VoiceOver on (command F5), and you can have the page read to  
you. See [1] for a quick howto.

No need for special (x)html code.

iCab also has this option (from the View menu).

Firefox and Camino nightly trunk builds will have support for  
VoiceOver in the future (currently you'll need to make a custom build  
from the Gecko 1.9 trunk to enjoy it. Basically, it works, but there  
are still rough edges. In a Camino build, it only works on the chrome  
part at the moment, not on the content part.).


I tried that page in Opera 9.5 alpha, but it didn't seem to make any  
noise. But maybe I'm missing a setting somewhere. I haven't read much  
about that issue, yet.
Opera 9.5a also seem to have basic support for VoiceOver (again,  
chrome only, as far as I could test).


[1] http://www.456bereastreet.com/archive/200505/ 
voiceover_and_safari_screen_reading_on_the_mac/



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Firefox bug on CSS white-space property

2007-08-28 Thread Philippe Wittenbergh


On Aug 29, 2007, at 10:47 AM, Kepler Gelotte wrote:


I don’t think it is a bug. First of all you are not overriding the
white-space property buy assigning it to child elements or parent  
elements.
You just create a conflict. Then precedence takes affect. According  
to the

CSS2 spec:

'white-space'
Value:  normal | pre | nowrap | inherit
Initial:  normal
Applies to:  block-level elements
Inherited: yes
Percentages:  N/A
Media: visual

Since you applied the “white-space: normal” to a span element  
which is not
a block-level element it is ignored. By changing the span to a  
div *or*
adding “display:block” to the span’s style, you will get the effect  
you were

expecting.


But in the testcase, the white-space property is applied to the a  
(in the head: a {white-pace:nowrap})

Firefox 2.0.0.x should _ignore_ that as well.

Anyway, my Gecko trunk builds (Minefield 20070828 and Camino trunk  
build) correctly wrap the text the same way Safari and WebKit wrap  
the text (ignore the white-pace:nowrap).

So Firefox 3.0 will display correctly.

And if the link text were to long for the space allowed by the table  
(e.g. a very very long word), the table should expand, unless the  
table has 'table-layout: fixed' specified.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Firefox bug on CSS white-space property

2007-08-28 Thread Philippe Wittenbergh


On Aug 29, 2007, at 12:43 PM, Kepler Gelotte wrote:


And if the link text were to long for the space allowed by the table
(e.g. a very very long word), the table should expand, unless the
table has 'table-layout: fixed' specified.


The table should expand even though he specified width=175? I learn
something new every day.

Try this:
http://dev.l-c-n.com/_temp/wp.html

And yes that is the expected behaviour. 'Width' on table and td/th is  
more like 'min-width' (idem dito for 'height').


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] safari display issue : help please!

2007-08-23 Thread Philippe Wittenbergh


On Aug 24, 2007, at 9:44 AM, Kit Grose wrote:


I can fix it two ways:
- wrap the form in a div
- remove the display:table; rule on the enclosing div
neither of which are very palatable in the context of the site.



Hi Andrew,

The issue is that you can't technically use the display: table  
property that way.


Yes, you can. Per
http://www.w3.org/TR/CSS21/tables.html#anonymous-boxes
an UA will/should insert a table-row and table-cell element.

But Safari 2.0.x is buggy in this (and  Gecko -Firefox- is buggy the  
other way around, sometimes inserting too many table-rows).


The best solution is
1. use another technique to generate a new block formatting context  
on the div where you currently use display:table. Using  
'overflow:hidden' is a better solution in many cases.

2. or: wrap your form in an extra div.

I'd go for the first solution, as it will avoid potential issues with  
Gecko browsers.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Cross platform line-height?

2007-08-17 Thread Philippe Wittenbergh


On Aug 17, 2007, at 10:07 PM, Paul Collins wrote:


Just wondering if anyone here has ever found a way of achieving a
consistent line-height for Windows and OS X? Been searching for a
while and can't seem to find the answer...


For a given font[size][face] you'll get consistent results using  
length (e.g. 1.5) across a wide range of UA's that support the css  
1 and css 2.1 font-properties.
Consistent doesn't mean that all UA's on all platforms will display a  
line-box at exactly the same pixel size. Different platforms have  
different (raw) font-metrics, different UA's use different methods  
for rounding off numbers, etc.


What exactly is your problem ?

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] setting fontsize in body

2007-08-08 Thread Philippe Wittenbergh


On Aug 8, 2007, at 6:45 PM, Rob Kirton wrote:


However I would have thought it
sufficiently an important feature to dedicated a couple of buttons  
in the
chrome bar to it, maybe a simple + and - (my actual sugestion to  
them).


What makes you think that there won't be 'a couple of buttons' ?
The feature (page zoom) has only been implemented for 2 weeks, and  
there is no GUI for it, yet.

You can track the discussion about the UI aspects:
 https://bugzilla.mozilla.org/show_bug.cgi?id=389628


  [...]To most users it is a moot point
whether or mot these buttons perfrom text resizing or indeed page  
zoom.
That I'm not sure. I, for one, would hate to loose the *text* zoom  
feature.

Page zoom doesn't really work, for me (and I'm not alone).

But that features (and there associated UI) to improve readability of  
a page is important, that we all agree. And those features will  
become ever more important (more 'old' eyes on-line, increasing  
screen resolutions).


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] setting fontsize in body

2007-08-03 Thread Philippe Wittenbergh


On Aug 4, 2007, at 10:45 AM, Gunlaug Sørtun wrote:

Back to Tee's problem with 'body {font-size: 62.5%}' etc in Opera/ 
Mac. It may be caused by the preset value for 'minimum font size'  
in that browser/OS.
If someone can check the preset value for 'minimum font size' in an  
unaltered Opera/Mac, as I set mine to '14px' years ago and have  
since just updated it, and I can't remember the preset value. Now I  
can't even check Tee's problem, because my Mac is off-line.


Unless my copy is sick, the default is 9px.
I did notice sometimes that Opera Mac (and maybe Win) tends to round- 
up decimal percentage points more aggressively than other browsers on  
my Mac(s). That is more often the case in a complex cascade (e.g.  
starting at 85.5% font-size, then a descendant has 90.5%, etc). I've  
seen Opera round the numbers upwards more often (that is not more  
than 1px per step, but can add up for deeply nested elements)


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] (Phillipe) margin problem in Firefox

2007-07-19 Thread Philippe Wittenbergh


On Jul 20, 2007, at 11:27 AM, Tee G. Peng wrote:


 Phillipe wrote in CSS-D:



It is a bug in Gecko. [1]

Here is your list, simplified.
http://dev.l-c-n.com/_temp/moz-egde.html
The top one is wrong, the bottom one is fixed.

li.poster {-moz-float-edge:content-box;}
does all the magic.

(one could argue about the semantics of your list construction,  
but css-d is not the place for that).




For reference, this is the post on CSS-D
http://archivist.incutio.com/viewlist/css-discuss/90611



OK Phillipe, please tell me what I did in-correctly with this  
construction.


(1) The List not being semantical for the comments markup?


A list is perfectly appropriate for comments - I use it myself.
That was not my point, actually.


[...]

(2) The P inside the LI ?
Added P tag for the comment message because there will be people  
leaving more than one paragraph of texts and I don't want a break  
in between.


(3) The numbering should be placed inside the 'poster' class  
instead of giving its own class?
I think placing the numbering inside the  li.poster class is more  
appropriate, but I don't want to do that because adding a new LI is  
more straight forward and easier to position the mustard box :)
Did it with a guilt but decided to go lenient for myself. Now you  
mentioned it, I feel very bad and uneasy even without hearing what  
you are to say


My problem with using a li just for adding a number to the comment.

I understand what you're trying to do - styling the list number for  
the comment, which, out-of-the-box and with current css specs, is not  
possible. (css3 makers/generated contents would allow many  
possibilities).


I would use the following structure:
ol
li
h6span%number/span John Doe on Jul 7, 11:34 AM/h6
pcomment here/p
!-- or more p --
/li

substitute an appropriate level of heading for your needs
float the span, adjust margins to taste

That is about what I do on this now-defunct site :
http://emps.l-c-n.com/articles/94/widgets-for-firefox

(and that uses Texpattern behind the scenes).
Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Shadow validation

2007-07-09 Thread Philippe Wittenbergh


On Jul 10, 2007, at 10:36 AM, Dean Matthews wrote:

I thought I was dense but now I think the W3C is just behind the  
curve ;-)
Not really, just chose the appropriate options (advanced...) when you  
try to validate a file.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] scope, cols and colgroups

2007-06-29 Thread Philippe Wittenbergh


On Jun 28, 2007, at 11:17 PM, Chris Price wrote:

I thought I had scope, cols and colgroups figured out until I built  
a table that clearly showed that I hadn't.


I have built a simplified table that illustrates this at http:// 
choctaw.co.uk/tabletest/


I want all all min columns coloured green and max columns coloured  
yellow but more than that I want to know where my logic has gone  
wrong.

Your page is invalid.
1. a colgroup cannot be nested in another colgroup
http://www.w3.org/TR/html401/struct/tables.html#edef-COLGROUP
is not really clear.
The html5 draft can help:
http://www.whatwg.org/specs/web-apps/current-work/multipage/section- 
tabular.html#the-colgroup

2. you have a xml syntax in there: colgroup /
colgroup is _not_ an empty element, it cannot use self closing syntax.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG]shorthand doesn't work for bg position (was auto-hiding bottom part of bg image when resizing fontsize

2007-06-20 Thread Philippe Wittenbergh


On Jun 20, 2007, at 3:57 PM, Tee G. Peng wrote:

Strange! it really doesn't work for me for my testing in Safair,  
Firefox and Opera.
I tried placing 'fixed' before and after the 'no-repeat', also  
after the x/y positions.


Here is the test page:
http://project.lotusseedsdesign.com/wandahennig/index2.html


There is a fixed background image (a sepia-toned image of a lady) on  
that page.
There are no problems with the order of values in the shorthand  
declaration.

See the syntax in:
http://www.w3.org/TR/CSS21/colors.html#propdef-background

But don't forget that 'fixed' means fixed to the viewport.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Safari 2.0?!

2007-06-19 Thread Philippe Wittenbergh


On Jun 20, 2007, at 5:20 AM, Nick Fitzsimons wrote:


PS - on OS X, of course.


The beta download comes with an uninstall package to roll you back  
to your previous version of Safari. It's on the Safari3Beta.dmg you  
originally installed from.


You can use the nightly WebKit builds to check out your documents in  
what will ultimately be Safari 3.0.

Those nightly builds can co-exist nicely with Safari 2.0.

http://nightly.webkit.org/

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Difference between IE and Firefox - can't figure it out...

2007-06-14 Thread Philippe Wittenbergh


On Jun 15, 2007, at 10:12 AM, Taco Fleur wrote:

I'm hoping someone can see the obvious on the following page and  
tell me

where I've gone wrong;

http://vasco.brucehighway.com/reports-hierarchical-display.htm
in Firefox it displays the totals on the right-hand side exactly  
where I

expect it to be.
In Internet explorer the numbers are aligning to the document and  
not to the

relative positioned elements as in Firefox.
Internet Explorer (6 but I've seen 7 acting up as well) doesn't know  
where 'right' is. It always computes that based on the nearest  
container that 'hasLayout' [1] and is positioned. In your case, I  
think it is body (I haven't gone through all of your stylesheets).


giving the parent p 'layout' _might_ solve the issue.

Why not a table ?

[1] http://www.satzansatz.de/cssd/onhavinglayout.html

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Difference between IE and Firefox - can't figure it out...

2007-06-14 Thread Philippe Wittenbergh


On Jun 15, 2007, at 12:26 PM, Taco Fleur wrote:


I'm not sure what
giving the parent p 'layout' _might_ solve the issue.
means though?

Maybe this translation works better:
Triggering 'hasLayout' on the parent p _might_ solve the issue

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Difference between IE and Firefox - can't figure it out...

2007-06-14 Thread Philippe Wittenbergh


On Jun 15, 2007, at 1:30 PM, Taco Fleur wrote:

What do you mean by 'hasLayout', I don't use that as a class or id  
name and
I have no idea what else it could be? Probably the only person who  
doesn't

get it, sorry...
Read the link in footnote [1] in my first message. It is only 45  
printed pages long.



[1] http://www.satzansatz.de/cssd/onhavinglayout.html



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Mac test please (was Safari now on Windows)

2007-06-12 Thread Philippe Wittenbergh


On Jun 12, 2007, at 5:02 PM, John Faulds wrote:

Well, as you mentioned it: I downloaded Safari for Windows today  
and didn't have any problem with it except that my own site looks  
completely screwed in it. It didn't look like that last time I  
checked with Browsercam and it doesn't look like that using Swift  
so I'm wondering if it might be a Safari 3 issue (and maybe just  
Safari for Windows).


You declare a width of 600px on body.

Yeah, and, you load stylesheets via xml PI. Safari/WebKit doesn't  
recognise media types in that case. It applies all your stylesheets.


(try to convince your boss to by a Mac)

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Mac test please (was Safari now on Windows)

2007-06-12 Thread Philippe Wittenbergh


On Jun 12, 2007, at 9:34 PM, John Faulds wrote:

Yeah, and, you load stylesheets via xml PI. Safari/WebKit doesn't  
recognise media types in that case. It applies all your stylesheets.


Yep, that was it. Thanks for that. But I'm curious why it's only a  
problem in v3 and not earlier versions.
Probably because Safari v2 gets text/html for your site. The same bug  
is present there, though.


- I'm just guessing, I don't have Safari v2 installed (v3 overwrites  
v2) on my PowerBook right now. I suppose you do some detection based  
on q values for application/xhtml+xml or similar. And Safari v2 only  
broadcasts text/html in the headers.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] WCAG Samurai Errata

2007-06-11 Thread Philippe Wittenbergh


On Jun 11, 2007, at 3:12 PM, Tee G.Peng wrote:

Lastly, I did not aware we can do this ? Did a brief reading in CSS  
3 spec and I tried testing it in Safari, Firefox and Opera   
(thought one of the browsers has already supported some of CSS3  
elements) but none of them work.


If images must be used for list bullets, do so only using CSS, as  
with ul { list-style: url(arrow.gif) disc }


That works perfectly fine in all 3 mentioned browsers. Even IE 7  
should work reasonably.
 You'll see an image 'arrow.gif', and if you have display of images  
turned off in your browser, you'll a disc. It is a good practice.


CSS 1/CSS 2.1 syntax btw.
You could also write ul { list-style: disc url('arrow.gif');}

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] layout/font site test - please

2007-06-05 Thread Philippe Wittenbergh


On Jun 5, 2007, at 8:09 PM, Designer wrote:


... the htmlbody is ignored by all except IE6


I hope this is a typo. IE 6 ignores this (and NN4 in case you worry)  
as it doesn't understand the '' selector. All other browsers,  
including IE 7 support the child selector.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Re: Use of Fieldsets other than in form?

2007-06-04 Thread Philippe Wittenbergh


On Jun 5, 2007, at 10:19 AM, Jackie Reid wrote:


Quick question for you lot.

Can fieldsets only to be used in forms or can they be used to group  
any sort of related information together (ie: business name, short  
description, logo and link).


In HTML 4.0, the description of fieldset can be found under the heading
'17.10 Adding structure to forms: the FIELDSET and LEGEND elements'
http://www.w3.org/TR/html401/interact/forms.html#edef-FIELDSET

That makes quite clear that fieldset is intended for forms only

And it is all part of chapter 17, 'Forms'.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] ie positioning help needed

2007-06-02 Thread Philippe Wittenbergh


On Jun 3, 2007, at 3:04 AM, kevin mcmonagle wrote:


heres the url with the styles at the time of the screen shot
http://www.eaf.textdriven.com


In Safari 2.0, WebKit, Gecko nightly builds (Mac), your gradient is  
covering the right hand scrollbar. That makes it impossible to use  
the bottom scroll arrow, or the thumb when one has scrolled to the  
bottom of the div.


It 'works' in Firefox 2.0.0.4, but that is actually a nasty bug with  
Gecko Mac scrollbars - bug that is fixed in nightly builds.


Screenshot:
http://dev.l-c-n.com/_b/kmcm.png

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] safari hack for overflow-x/y

2007-05-31 Thread Philippe Wittenbergh


On May 31, 2007, at 10:04 PM, kevin mcmonagle wrote:


So i want safari to see
#wrapper
{overflow:auto;}

And I want all other browsers to see
#wrapper
{overflow-y:auto;}

Safari hacks seem to be semi-permanent because when they are found  
they are fixed in the lastest version or build of webkit  and  the  
new version of safari is coming out soon.


#wrapper {
	overflow:auto; /* for Safari, older versions of Firefox, Opera,  
other browsers that don'ts support the css 3 syntax */
	overflow-y:auto; /* Firefox 2.0+, IE 5.5+, Safari 3.0-Webkit- 
Konqueror4.0 */

overflow-x:hidden; /* depending on your set up...*/
}

You need to keep them in that order, what comes later in ruleblock  
overrides what comes earlier.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Converting font size from pt to % or em

2007-05-25 Thread Philippe Wittenbergh


On May 26, 2007, at 11:16 AM, Paul Novitski wrote:

Do you mean no elegant way to scale them in a user stylesheet or no  
elegant way to scale them in real time, e.g. with a mouse wheel?
I have my minimum font-size set to 12px [1] (Gecko browser), or  
sometimes 14px (when I'm tired, and really p*** by mouse type and the  
need to zoom in way to often)
* No elegant way to scale the whole thing correctly in a user  
stylesheet, short of rewriting the whole author stylesheet [2]: with  
the 62.5% 'trick', the base for all computation will be 12px in my  
case. Say I reset the font-size to 16px for a particular site (using  
@-moz-document), all scaling in that author style-sheet will be  
oversized, as I thing Christian explained).
* No nice way to zoom out in real time, due to the clash between  
minimum font-size and the author specified miniscule base.


[1] that is my minimum font-size, below which I cannot read text. It  
is _not_ my preferred font-size.
[2] user stylesheets are already a pain for the average user, image  
if they have to rewrite the author stylesheet completely...
(even for me it would be serious nuisance - and I have a 3000 lines  
long user stylesheet)


---
While in theory, I, as a user, should like that method of setting  
font-size - combined with my minimum font-size is should guarantee  
readable text, in practice it is a pain: many more sites break (even  
some where e.g width is set to ems or the like), or quickly become  
way to wide for my preferred window width.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] width of inline lists

2007-05-20 Thread Philippe Wittenbergh


On May 20, 2007, at 9:02 PM, Designer wrote:

Has anyone got a (simple) workaround for the fact that I want to  
specify the width of each li in an inline list?  (which is not  
allowed).


I'm using the simple menu:

div#menu {
font : 13px  Verdana, Arial, sans-serif;
text-align : left;
color: #333;
}

div#menu li{
width : 250px;
display : inline;
list-style-type : none;
}
[...]

and, of course, the 'width : 250px' is just ignored.  I've tried  
setting a width on the a but that doesn't work either.


MUST I use a table?


Obviously:
li {float:left} ?
or li {display:inline-block } ?
The second one won't work in Firefox up to 2.0.0.x (but will work in  
Fx 3)


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] While we're on the subject of validators...

2007-05-01 Thread Philippe Wittenbergh


On May 1, 2007, at 2:40 PM, John Horner wrote:


Well perhaps your second point is valid (and well formed) but this
example:

  body { background: url('image.gif')no-repeat top }

isn't just about cross-browser compatiblity. Surely without the
whitespace, it's actually invalid CSS? The same as it would be if
no-repeat and top appeared without space between?


I don't think it is invalid CSS - the closing parenthesis ')'  
terminates the url value. The white-space is optional. I think.
For keywords, it is of course different 'top right' is not the same  
as topright.

See for more:
http://www.w3.org/TR/CSS21/syndata.html


And as for the logic of HTML tags, it would undeniably be a better
validator if it could warn tag-based declaration doesn't match any
known HTML element, the way Perl's error reporting warns me about
variables I've created but never used, on the basis that it's  
probably a

typo.


The validator doesn't look/care about what you put in a selector.  
What if you are using that css file for some xml document that has an  
element 'navbar' within its namespace ?


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Template Review

2007-04-29 Thread Philippe Wittenbergh


On Apr 30, 2007, at 10:04 AM, CK wrote:


http://working.bushidodeep.com/dom_site/template.php


In random order:
more white-space (everything is like squished together).
as Paul mentioned: why the pointer over the dt/headings ?

the letter-spacing on the dd is annoying. It is a bad idea to letter- 
space body text - font-designers do know their stuff, you know. Might  
work well on your machine, with your fonts/font-size. On my browser  
(Gecko trunk, OS X 10.4.9), with my font-size, I see all characters  
drifting apart.


combining fixed width on a container and then setting a max-width in  
em inside (your dd) is problematic.


use sans-serif fonts for headings and serif fonts for body text (like  
the dd, again) will create some more visual structure.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] strong v's b , em v's i

2007-04-23 Thread Philippe Wittenbergh


On Apr 23, 2007, at 4:23 PM, Mordechai Peller wrote:

The site is wrong, plain and simple. b and i ARE depreciated,  
while strong, em, and blockquote are certainly NOT. Also, the  
size and type attributes are also not depreciated.

deprecated... [1]

And no,  neither b nor i are deprecated; or strong and em.

When in doubt, ask the official docs:
http://www.w3.org/TR/html401/present/graphics.html#edef-B
http://www.w3.org/TR/html401/struct/text.html#edef-STRONG

note also the html5 proposals:
http://www.whatwg.org/specs/web-apps/current-work/#the-b

[1] http://hsivonen.iki.fi/wannabe/

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] making tables/rows/cells in IE lose their tableness via CSS

2007-04-16 Thread Philippe Wittenbergh


On Apr 17, 2007, at 5:21 AM, Patrick H. Lauke wrote:

I'm just at the beginning of a little experimentation, and already  
IE stumps me.


Basically, trying to use a table in markup but hide one column and  
make the other fall apart into individual inline cells. Try as I  
might, though, IE doesn't seem to lose its desire to show the cells  
as a table. Any thoughts?


http://www.splintered.co.uk/experiments/archives/tagcloud_table/ 
index.html
I once tried the same exercise, but using 'inline-block' in stead of  
'inline'.

Failed.
The only thing that would work is using position:absolute on the  
table-cell, but then you'd need to position each one of cells via  
'top' and 'left'.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] opacity

2007-04-11 Thread Philippe Wittenbergh




For a little while, I went down the path of  opacity: 0.5; but
quickly found that nothing could be done to reverse the effect for
children of the object - they all went wishy washy too!


That is the expected behaviour. opacity [1] applies to the whole box  
(and its descendants), it is _not_ about making the background  
transparent/translucent. What you want in that case is using rgba()  
or hsla() colours [2].
rgba() is supported by Safari and I think Konqueror (can't test on my  
set up).
hsla() is supported by Webkit, I think by Konqueror, and by nightly  
builds of Gecko.



actually I've been having problems with opacity with text too...  
fine in Firefox but IE doesn't seem to recognise it.


IE doesn't support the opacity property. It only supports a  
proprietary property


[1] http://www.w3.org/TR/css3-color/#transparency
[2] http://www.w3.org/TR/css3-color/#rgba-color

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Odd behaviour in Opera and Safari

2007-04-09 Thread Philippe Wittenbergh


On Apr 10, 2007, at 11:05 AM, Rachel May wrote:


This site
http://skylight.web1.heliocell.com/professionals--community- 
groups.aspx
(please note still in testing, so e-commerce transactions won't go  
through,
etc.) is having odd behaviour in Opera and Safari - the page  
headings are
going small after the page is loading. This is causing the search  
box to

overlap the navigation (the logo is in an H1 tag).

Also in Opera, the links are going red.

That happens more often when one of the required css files is not found.
In your case, you are linking to
'http://skylight.web1.heliocell.com/niftyCorners.css' (from within  
your script[1]),

but that file returns a 404not found.

The IIS 404 file has some css into it, although its content type is  
text/html
Opera and Safari interpret the file nonetheless and apply the  
stylerules in that 404 page to the real html page, hence: red links  
and smaller type on your heading and who knows what more.
Yes it is a bug in those UA. That is what you get when a UA accept  
whatever content-type one throws at it and try to interpret it.

Gecko based browsers simply ignore that missing file completely.

[1] function AddCss() in script.js
Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] background won't repeat-y

2007-03-04 Thread Philippe Wittenbergh


On Mar 5, 2007, at 2:28 PM, Christian Montoya wrote:


I'm almost certain that you need the positioning in that shorthand to
make it work:

background:url(some image) left top repeat-y;

or center center, or anything really.


No, you don't need.
With shorthand, properties that are not specified assume the default  
(in this case top left or 0 0)


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] playing with layouts, pt2

2007-02-25 Thread Philippe Wittenbergh


On Feb 26, 2007, at 6:50 AM, Gunlaug Sørtun wrote:


Designer wrote:


I don't understand that.  If Safari is capable of handing the correct
 mime type, how can (why would)  it serve 4.01 strict?


Probably because Safari is cheating :-)

Not really.
Safari's Accept Headers: Accept: */*
translated: 'I prefer text/html'
The latest Webkit builds report this:
text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/ 
plain;q=0.8,



1: Don't think Safari has a genuine xml-parser. Suspect Safari to use
the same parser as for 'text/html'.
Don't think so. But just as with other browsers, the xml-parser is  
much weaker and prone to bugs compared to the html parser.


Note also: Safari only knows about the five required named entities.  
If you use other named entities, Safari will report an error. (Gecko  
and Opera know internally about about most named entities). Use  
numerical entities instead.



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com





***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***



Re: [WSG] Odd FireFox div background image behavior

2006-03-18 Thread Philippe Wittenbergh


On Mar 19, 2006, at 6:35 AM, [EMAIL PROTECTED]  
[EMAIL PROTECTED] wrote:



Test site:  www.designbyatfb.com/designs/golfintro.html
The background-image in the
}
#containerDiv{
position: relative;
top: 0px;
width: 800px;
margin-top: auto;
height: 100%;
color: rgb(128, 128, 128);
overflow: visible;
background: transparent url(css-images/swing.gif) no-repeat ;
background-position: bottom;
}
disappears when the page is resized in Fire Fox v 1.0.7 and Mozilla  
1.7.12. It does not even appear at all in Opera 8.02
If page is opended at 800x600 it is fine. If it is opened at  
1024x768 or higher, it is fine.
If you resize after the initial view it disappears until your  
refresh.  I have tried different Doc types, and encodings, nothing  
makes any difference.


it simply never appears on my side (Firefox 1.5, Camino 1.0, Safari  
2.03, Opera 9 TP2). And that is correct.
Your #containerDiv is pretty much empty, because it mainly contains  
two floated blocks,  which by definition are taken out of the flow.

That #containerDiv collapses in height, and hence can't show your image.
And no, your 'height:100%' won't do anything, as it is impossible to  
calculate that height, given that no height has been specified on any  
parent.


It happens to wrok in IE windoze, because declaring a height in that  
browser  gives the element 'layout' [2], and contains the floated  
blocks.


#containerDiv {display:table}
will fix your problem, by establishing a new block formatting context  
[1], which will contain your floated blocks.


[1] http://dev.l-c-n.com/IEW/simulations.php
http://www.w3.org/TR/CSS21/visuren.html#block-formatting

[2] http://www.satzansatz.de/cssd/onhavinglayout.html

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com



**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] CAPS in stylesheets

2006-03-13 Thread Philippe Wittenbergh


On Mar 13, 2006, at 5:14 PM, John Wells wrote:


Can we get a summary of this CAPS lesson?  Are we saying that:

- ONLY when serving XHTML as application/xhtml+xml, stylesheet
selectors must be all lowercase.

?


Some browsers are sensitive to case when XHTML is served as text/ 
html. iCab comes to mind.


Having said that, I'm a bit confused.  I wish I could remember where I
read this, but I thought that *technically* (that's how I remember it
being phrased) stylesheet selectors should not contain characters
other than alphanumeric (and must start with a letter).  Well once I
read that, I moved all of my selector naming to camelCaps...so now I'm
feeling squeezed.  In (properly served) XHTML can I use underscores to
separate my words?  I strive to make my code as readable as possible
(from CSS to PHP), so naming conventions are a bit of a sticking
point.

Why not ask the source ?
http://www.w3.org/TR/CSS21/syndata.html#q6

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com



**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Absolute Positioning-A Naive Question (Maybe)

2006-03-13 Thread Philippe Wittenbergh


On Mar 14, 2006, at 11:35 AM, Paula Petrik wrote:

When I read the W3C specs (not the most riveting exercise on the  
planet), it seems that the developers emphasize absolute  
positioning. For example, they describe using floats to float small  
bits of text or images. It seems, however, that floats have become  
the order of the day. Rather than small bits, whole parts of  
designs are floated about. Was this the W3C's intent? Or, have  
floats become the modern equivalent of tables? Is there some reason  
why absolute positioning has fallen by the wayside? CSS-Discuss's  
wiki describes absolute positioning as capable of simple designs;  
yet, a significant proportion of csszengarden designs are  
absolutely positioned, and I wouldn't term them simple. Just  
wondering what the current wisdom is on this issue.


You read that correctly, and Felix already pointed you to that  
article by David Baron.
The main reason that 'floats' are so popular in current css driven  
design:  they are currently, and very unfortunately, the best tool  
available for the job of creating multicolumn designs.
Absolute positioning has one problematic side effect: as the height  
of the column is usually unknown (depending on the contents), it is  
more difficult to position an element under that absolute positioned  
element. Not impossible, mind you, with a little help of javascript.
An alternative would be the use of {display:table} and {display:table- 
cell}. But that is not supported by IE, which puts serious limits on  
the technique.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com



**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] IE Reverse Indent ?

2006-03-07 Thread Philippe Wittenbergh


On Mar 8, 2006, at 2:51 PM, Jason Turnbull wrote:


http://www.hudsonantiquecarclub.com/index.html
IE6 adds an reverse indent just under the header Monthly Meetings
(Paragraph starting Rudy's...)
Is this to do with the thumbnail to the left?


Known as the 3 pixel text-jog
http://www.positioniseverything.net/explorer/threepxtest.html

Easily fixed by adding #content_home to the IE specific style you have
already:

ul a, ul, li, #content_home {height: 1%;} /* Holly hack fix for IE  
bugs */


That is a bit radical... giving all those element 'layout' can  
possibly cause more problems than it fixes.

In this case:
/* - hide form IE Mac - \*/
* html #content_contain {height:1%}
/* - end hiding - */
ought to do nicely.
(and served only to IE 6 and below. That 3px jog thing is reportedly  
fixed in IE 7 alpha0).


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com



**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Styling Fieldset and Legend Elements

2006-03-07 Thread Philippe Wittenbergh


On Mar 8, 2006, at 2:48 PM, Lachlan Hunt wrote:


Thierry Koblentz wrote:
I think it worth mentioning that the * html hack works in IE *Mac*  
too.


What about conditional comments?  I don't think that really matters  
much anyway, since IE Mac is officially obsolete (i.e. now  
completely unsupported by Microsoft) and ...


It might be unsupported by MS, but it is still quite well used, in  
the **real** world.

That said, no, IE Mac doesn't support CC.

I think some unwanted hacks applying to it would be the least of  
its problems.


Let it see some of the hacks targeted at IE Win can end in a complete  
disaster (i.e. inaccessible content).
The often used 'Holly Hack' (height:1%) causes big problems in IE  
Mac. That browser incorrectly computes % height as '0' instead of  
'auto' as it should per CSS 2.1 (for in-flow elements when no parent  
element has a height specified).



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com



**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Styling Fieldset and Legend Elements

2006-03-06 Thread Philippe Wittenbergh


On Mar 7, 2006, at 2:12 PM, Lachlan Hunt wrote:

  I'm looking for a way to style a legend element with a background  
that spans the whole width of the page.  I originally used a  
regular heading (h3) because I knew the design would have  
difficulty with fieldset/legend, but the after the accessibility  
review it was requested that we use fieldset/legend anyway and I  
need to get the design as close as possible.


By default, legends are only as wide as they need to be to fit the  
text and there seems to be little one can do about this in current  
browsers.  I came up with one solution using a thick top border on  
the fieldset behind the legend element which works in Firefox, but  
it doesn't quite work in IE because it leaves a small white gap  
between the legend and fieldset border.  I currently don't have  
access to any other browsers for testing, but I'd like it to work  
fairly well in at least all major browsers.


The following test case demonstrates how I want it to look and my  
current solution that works in Firefox.  Both examples in the page  
should look roughly identical


http://lachy.id.au/dev/2006/03/fieldset


On OS X 10.4.5
Firefox 1.5, Camino - latest nightly trunk build, iCab 3.0, Safari  
2.03 and WebKit nightly build, Opera 9 tp2,  all look the same.

and so does Konqueror/KDE 3.5 running on top of OS X .

IE 5.2 Mac has a small 1px  white border around the legend, barely  
visible given your light background-color. There is also some white- 
space under the legend (padding on fieldset, you won't get rid of it).


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com



**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] making table cells share equally

2006-02-21 Thread Philippe Wittenbergh


On Feb 22, 2006, at 8:10 AM, Ted Drake wrote:

If I have two or three columns in a table and I want the cells to  
share the
space equally, what would you suggest?  I don't have a width on  
these cells
right now and if one cell has lots of content, it's pushing the  
other column

to the curb.


table {table-layout:fixed; border-collapse:collapse; width:xxx}
http://www.w3.org/TR/CSS21/tables.html#propdef-table-layout

according to the rules, a width on the table need to be specified; IE  
Win will default to 100% if no width is set.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com



**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] TARGET in 4.01 Strict

2006-02-16 Thread Philippe Wittenbergh


On Feb 16, 2006, at 5:58 PM, James Bennett wrote:


I was being somewhat facetious, but every browser I have within arm's
reach (which includes all the popular browsers except Safari -- I
don't have a Mac here at home to refer to) implements some form of
extended Back functionality which displays a list of all the previous
pages for the current window/tab and allows any one of them to be
selected.


Safari has this same functionality: onclick= 1 step back;  
onmousedown= popup menu with your recent widow history. I'm not sure  
how many steps it remembers, I never use Safari except for testing.
Even then, there is the History (Go in some other browsers) menu to  
help the user out.


Philippe

Philippe Wittenbergh
http://emps.l-c-n.com


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Opera Labs and Opera 9 Preview 2

2006-02-08 Thread Philippe Wittenbergh


On Feb 8, 2006, at 4:11 pm, heretic wrote:


Maybe the standards community prefer to ride ponies instead of real
race-horses? ;-)

Must be something to do with keeping nearer the earth. Opera spoils
web developers, and makes Internet Explorer (and Firefox, to a lesser
extent) that much more shocking ;-)


hehehehh ahhh dear, we're the mac users of the browser world aren't we
;) [grabs his fireproof suit]

speaking of macs, i wonder if the new mac version of opera will change
the stuff that angered joe clark?
[http://blog.fawny.org/2005/02/01/opera/] i don't have a mac to try it
out.


Hmm, if they could clone the interface of Omniweb (they already cloned 
a couple of features) or Camino, it would become a very nice browser.
- granted, they did make some serious progress with Opera 9 (tp1 and 
tp2) Part of Joe Clark's issues are fixed. But the UI design still 
feels like poor. The 'looks' are still out of place and Windoze like.


That said, to stay on topic, with the latest release, they've made good 
progress in fixing their rendering bugs. A whole bunch of 
margin-collapse bugs are fixed, quite a few issues with relative 
positioning as well.
And speed is back, on Mac at least (tp 1 felt more like one of those 
big heavy duty horses used in the fields rather than Georg's race 
horse.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Re: Multiple language usability query

2006-02-07 Thread Philippe Wittenbergh


On Feb 7, 2006, at 5:08 pm, Andrew Cunningham wrote:

Also worth noting, with CJK text, in the absence of css specifying a 
font, the browser will chose the font based on langauge of the text. 
If the language of the text is not indicated then IE will try ot use a 
Japanese font to display CJK, Opera uses a Korean font instead, and I 
forget what Firefox does.


Firefox Win takes the fonts as specified in the browser preferences 
(ContentFont colors advanced). If none is available (ie no 
East-Asian language pack installed) then all madness is possible. As I 
said, I've actually never seen what happens on a Win PC that lacks the 
East-Asian language pack ('cause all PC's here in town have it 
installed).


(hence, out of curiosity, if someone would be kind enough to send me a 
screenshot of the file I posted earlier, that would be appreciated).


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Multiple language usability query

2006-02-06 Thread Philippe Wittenbergh


On 7 Feb 2006, at 7:02 am, Sarah Peeke (XERT) wrote:

I am new to multiple language sites. However, I have a client who  
has a
small amount of text on one page of his site referring Japanese  
speaking

users to a different (Japanese) site.

Firefox and Safari on the Mac seem to have a default which allows the
Japanese to be read. However, I noticed that good ol' Windows IE6
doesn't automatically render other languages and that its preferences
needed to be changed etc.


Japanese (and other East-Asian) language support is installed by  
default on OS X.
Not so on Windows side of things (it comes with the install discs as  
an extra package).

I've been told that Firefox/Win tries to display the text nevertheless.

Lachlan wrote
It's probably a font related issue.  IIRC, if none of the fonts  
specified in the font-family property contain the glyphs for those  
characters, IE won't find and display them.  You may be able to  
improve the situation by including the following:


font-family: Arial Unicode MS, Lucida Sans Unicode, sans-serif;


Hmm, not sure (but I can't test those things, I don't have access to  
a PC without East-Asian language support...:-)).

A little test file
http://dev.l-c-n.com/_temp/j-test.php
Includes a screenshot. I've used a whole collection for the font- 
family (first some Japanese font-families, then some fall back stuff  
as mentioned).

Let me know how it looks like (for the curious, it is my name).

Sarah again


My question is this:

Since most web users are using IE, and I imagine most IE browsers  
do not

have multilingual capabilities enabled, then they would see
gobbledegook. So...

How do we overcome this in terms of usability? Is there a standard?
Should I include (in English) a title and/or link above the Japanese
text explaining how to render the text below, or explain, at the very
least, that the strange characters below are in fact in another
language?


What about including the text both in Japanese and English ?

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Display:Table

2006-02-04 Thread Philippe Wittenbergh


On 5 Feb 2006, at 7:13 am, Al Sparber wrote:


The issue:

Imagine that all display values are supported by all browsers as of  
midnight tonight.

LOL - still eleven hours to go on my side ;-) can't wait...

Do you think that using display:table and display:table-cell to  
create multi-column layouts is correct or incorrect - and why?


display:table/display:table-cell is certainly a good way to style  
content-blocks into multiple columns.


Pro: it creates a structured grid that is easily controlled: new  
block formatting context for each column, various properties of table  
(table-layout, border-collapse, empty-cells, horizontal alignment),  
all with a minimum of mark-up; equal-height columns.

Probably closer to the original intent of CSS as a styling language [1]

Cons: bound to source order, some undefined behaviour  
(position:relative on table-cell [2]).


It is already possible to create very consistent things across  
browser-land. Except, that is, in IE (win).

Much less of hack than float based layouts.

@Kenny Graham
CSS's multi-collumns [3] is more about flowing content across  
multiple collumns, not really a layout technique for grid layout and  
grid positioning.


There exist an interesting early draft about grid layouts:
http://www.w3.org/TR/2005/WD-css3-layout-20051215
See Bert Boss message to the www-style list [4]


[1] http://dbaron.org/log/2005-12#e20051228a
[2] http://www.w3.org/TR/CSS21/visuren.html#propdef-position
(pos. relative on table-cell is only supported by iCab 3.0)
[3] http://www.w3.org/TR/css3-multicol/
[4] http://lists.w3.org/Archives/Public/www-style/2005Dec/0081.html

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] IE7 hacks

2006-02-03 Thread Philippe Wittenbergh


On 4 Feb 2006, at 3:25 am, Ted Drake wrote:

I posted a hack to IE7 today. I know I'm not the first one to find  
this, but

thought I'd throw it out there for all to love on.

www.tdrake.net

It's pretty simple. But please, think beyond hacks.


I'll file a bug report for that parsing bug :-)

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Check boxes ticked (UK Law)

2006-01-30 Thread Philippe Wittenbergh


On 31 Jan 2006, at 12:33 am, Paul Collins wrote:

I recall reading somewhere a while back that UK law states you  
can't have a check box ticked on a form


EG - untick this box if you don't want to receive emails would be  
illegal for a UK site.


That would be European Community law, not only UK law.
And yes, I believe this to be correct. You have to make this 'opt- 
in', default being 'opt-out'.




Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Mac CSS issues

2006-01-27 Thread Philippe Wittenbergh


On 28 Jan 2006, at 3:09 am, Veine Vikberg wrote:

Well Georg came with a suggestion that is working for Windows  
machines - now this design is close to what it needs to be on  
Windows IE/Moz/Opera. My issue is now Mac - I have a mac in the  
office (Strawberry iMac os 8.6 IE 5.1) and on that one all layout  
is virtually gone with the wind.


All id and classes seems to be ignored, and they are thrown all  
over the place, stretching the page to a three screen horizonal  
scroll.


I am at a loss to why this happens.

The page is question is:
http://jpfco.com/testdesign/new/


Your iMac runs old sofware...

That site displays fine in Opera 9TP1 Mac, Safari 1.2 and 2.0,  
Firefox 1.5. Camino and Firefox nightly trunk builds have a problem,  
the side test (under the contact data) drops below the image. Nothing  
to worry about, it is a regression from a recent code base mod.


IE mac has some more serious issues, at least one.
It puts #content next to the #header, which is floated. Solutions:  
either add a clearing element between header and content, or just  
float:left  #content.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Article: MIME and Content Negotiation

2006-01-16 Thread Philippe Wittenbergh


On 16 Jan 2006, at 11:53 pm, designer wrote:

However, I remain confused in the particular case of the PHP  
approach.  The PHP looks to see if(stristr($_SERVER 
[HTTP_ACCEPT],application/xhtml+xml)) and on that basis it  
describes the appropriate prolog, mimetype and charset. So the  
final three lines of the php code (where the action takes  
place :-)  are:


   header(Content-Type: $mime;  charset=$charset);
   header(Vary: Accept);
   print $prolog_type;


That sounds very weird. Either your browser is acting up or something  
is not properly configured, or your server is doing something odd on  
the way.


Is the $charset var set up correctly ? (near the top of the code)
It should read:

$charset = utf-8;

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Safari Lightening Entire Background Image

2006-01-12 Thread Philippe Wittenbergh


On 13 Jan 2006, at 9:18 am, Mani Sheriar wrote:

I’m having a slightly maddening problem that appears to be only in  
Safari.


The test page is here: http://www.manisheriar.com/sparkplug/paywi/

On Firefox and IE all seems good, but in Safari the grays of the main
background image are too light and therefore clashing with other  
images

where it is supposed to blend.  See screenshot here:
http://www.manisheriar.com/sparkplug/paywi/_images/paywi.png


There is indeed a huge difference between the rendering of  
bg_home.jpg (the big one in the middle) and bg.jpg. See

http://emps.l-c-n.com/bm/mani.png
(this is on 10.4.4)
I can reproduce this on 3 different machines. Latest Safari on 10.3.9  
shows the same.
Safari has known issues with png files (colour that doesn't match the  
same colour in htmls/css), but that is not the case here.


Are all those images saved as the same compression level ?
Could this be related to background-repeat: repeat or no-repeat ?

I've never seen something like this with jpg files.

PS - the problem only happens with Safari. All other browsers on OS X  
display the page correctly.




Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Jello Mold not working right in IE/Win

2006-01-08 Thread Philippe Wittenbergh


On 9 Jan 2006, at 9:18 am, Seona Bellamy wrote:

I've used the Jello Mold Creator at Position Is Everything to  
handle the
sizing issues for the site I'm redeveloping. Seems to be working  
fine in
Firefox, but when I opened it in IE I found that it was far too  
wide. Ugly
sideways scrolling, content vanishing off the right side of the  
screen, the

works.

I'm pretty sure that I took everything I needed from the Creator,  
so can

someone please have a peek and see if they can figure out why IE isn't
behaving nicely?

Page: http://staging.renovate.com.au/
CSS: http://staging.renovate.com.au/_resources/layout.css


I haven't had a look at your page in IE win but noticed one thing in  
the source code.
In your conditional comment, the width of the expression is set to a  
negative % value. Don't think that is correct.



!--[if IE]
style type=text/css
#sizer {
width:expression(document.body.clientWidth  0 ? 280px : -28% );
}
/style
![endif]--



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] XHTML again

2006-01-07 Thread Philippe Wittenbergh


On 7 Jan 2006, at 3:35 pm, Lynne Pope wrote:


On 1/7/06, Gunlaug Sørtun [EMAIL PROTECTED] wrote:

It is not as easy to hide proprietary and 'not-yet-recommended'  
CSS from
the validator, as it is with all the garbage often needed to make  
IE/win

behave.
OTOH: hiding something in a conditional comment (or in a 'non- 
existent
stylesheet', like I do at times) doesn't make it more valid - just  
hidden.


Conditional statements are not hidden, they just do whatever you code
them to do. They are a valid markup.


Conditional Comments are indeed valid markup. The point here is that,  
it is a comment, and anything within a cc it is hidden from (X)HTML  
compliant browsers. Only IE for Windows can take a peak at it. Thus  
yes, style within a CC is hidden from the validators.


BTW: non-valid CSS doesn't affect HTML/XHTML status/validity at all.


True. However, if we are coding to standards then it pays to be aware
of any coding errors in css. You can't look at each standard in a
vacuum.


Whatever your css contains won't affect the validity of your (X)HTML.
Then, to what Standards are you coding your CSS : CSS 2.0, CSS 2.1  
CSS3 ?
CSS 2.0 is a hopeless mess, CSS 2.1 is a bit better but still full of  
contradictions and undefined behaviour. The only part of the whole  
CSS thing that is pretty stable is the CSS selectors module.

http://www.w3.org/TR/css3-selectors/

And when you validate a stylesheet, against what do you validate ?
In my case you'd need to check against CSS 3, and even then, you  
better check against the respective docs (various versions) because  
the W3C validator doesn't know or support those (namespaces,  
mediaquerries, css-text amongst them).


The text-justify property you mention in a previous message is  
included in one of those (draft) docs and currently only implemented  
by MS IE win.






It does not validate at all as HTML 4.01 Strict - are you sure you
are checking validation against that DOCTYPE?


This sounds a bit strange to me.
Which source-code should be checked as HTML4.01?

Given the fact that the validator is fed an XHTML1.1 page with the
correct MIME-type by default, is it even possible to check that
source-code as HTML? I would think not.

Enforcing the validator wouldn't work - and it shouldn't since the
source-code isn't 'HTML4.01 Anything' when it's served to the  
validator.



And your point is? I made the comment that the site does not validate
as HTML 4.01, did you see me say how I validated it?  Anyone designing
a site to render as one DOCTYPE in some browsers and another DOCTYPE
for other browsers, and who wishes to have the pages validate against
both DOCTYPES  would, I assume, check the validation for both.
This can be done in many ways, such as entering the source code, or in
cases where the person looking at the code is sufficiently
experienced, just looking can show there are errors.


I manually validated 3 random pages from Bobby's site: accessing the  
page with an older browser, copy the source code and pasting that  
into the textarea of my local copy of the W3C validator (@v 0.7.1):  
all validated correctly as html 401 strict.


Georg point is: if you send a url to the validator, it will validate  
against xhtml, because that is the way the data is sent to it (with  
mime type application xhtml+xml).


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Fieldset inclusions and possible Firefox bugs

2006-01-05 Thread Philippe Wittenbergh


On 5 Jan 2006, at 8:30 pm, Nick wrote:

I recently discovered this same problem and it is indeed a bug  
within the
new Firefox. Unfortunately, no one's currently working on it which  
means
that we're stuck using ugly solutions to fix it. I found that  
wraping an
extra clearing container around the original within a fieldset  
works. It's

definitely ugly but seems necessary at this point. So instead of

div class=clear/div

to clear two floating items within a fieldset, its

div class=cleardiv class=clear/div/div

or however you're clearing your elements. Adding that wrapper  
container
seems to do the trick. I did manage to find a bug report of this on  
the

Mozilla site here:

https://bugzilla.mozilla.org/show_bug.cgi?id=309550


Given that the OP use a definition list, there is a simple/clean way  
to fix the problem caused by that bug.

Set the dl to {display:table}
http://dev.l-c-n.com/Gecko/fieldset-clear-309550.php

You may need to set a width on the dl to avoid shrink-wrapping.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] XHTML 1.1 Entities (WAS Claiming compliance when a site doesn't comply)

2006-01-05 Thread Philippe Wittenbergh


On 6 Jan 2006, at 10:50 am, Kenny Graham wrote:


List of XHTML 1.1 entities, served as application/xhtml+xml :
http://www.w3.org/People/mimasa/test/xhtml/entities/entities-11.xhtml

I really hope I'm right, or I'm gonna have to go back to a lot of
sites to fix a lot of ldquo;s and such.


If you want to support Safari (with application/xhtml+xml), I'm  
afraid, you'll have to go back...


Screen shot from the page linked above (Safari 2.02 - latest Webkit  
nightly builds are identical):

http://emps.l-c-n.com/bm/Safari-entities.png

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] XHTML 1.1 Entities (WAS Claiming compliance when a site doesn't comply)

2006-01-05 Thread Philippe Wittenbergh


On 6 Jan 2006, at 1:42 pm, Jason Turnbull wrote:


Kenny Graham wrote:
I really hope I'm right, or I'm gonna have to go back to a lot of
sites to fix a lot of ldquo;s and such.




Philippe wrote:
If you want to support Safari (with application/xhtml+xml), I'm
afraid, you'll have to go back...


If these entities are not allowed when served as application/xhtml+xml
shouldn't the W3C validator pick this up? Or has Safari got it wrong?


Not really wrong. Firefox and recent Opera versions rely on the  
Doctype to parse the document as xml in addition to the mime-type  
[1], while Safari bases its parsing only on the mime-type.


[1] this gives them additional knowledge about those entities.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Mac FF hidden div still shows scrollbars

2005-12-22 Thread Philippe Wittenbergh


On 23 Dec 2005, at 3:21 am, Thomas Livingston wrote:


Hi listers,

We have a div being hidden (visibility:hidden;) and then using  
javascript to show hide layers (sort of a pop-up but not sorta thing).


If we give it a height and apply overflow:scroll; (or auto) it  
looks and works dandy, except for Mac FF (1.5). We are still seeing  
the div's scroll bars when in it's hidden state.


Anyone come across this before and fix it?


https://bugzilla.mozilla.org/show_bug.cgi?id=187435

Try display:none, if that fits in your design.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Setting Up Font Sizes

2005-12-21 Thread Philippe Wittenbergh


On 21 Dec 2005, at 5:25 pm, Lachlan Hunt wrote:


Felix Miata wrote:

Lachlan Hunt wrote:


body { font-size: small; }
is generally acceptable and is approximately the same as 80% of the

Definitely not acceptable to me for content paragraphs. :-(


Why not?  Is it too big or too small for you?  Or is it just not  
precise enough?


If you say it's too small, then I'd accept that.  There are many  
who say anything below 'medium' is too small for body copy.  If you  
say it's too big, then I have to very strongly disagree and say  
that making it any smaller, will make it very difficult or at least  
uncomfortable for many users to read without increasing it.



If you ask me and my tired old eyes:

Depending on which font-family you use, font-size:small is either on  
the lower limit (georgia, which is a *big* font), or really too small  
for me, when used with Roman fonts.


Using the same font-size:small for Japanese fonts/text on the other  
side works pretty well across the board for me, except in Camino/ 
Firefox Mac, which tends to smash down fonts.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Browser Resolutions

2005-12-16 Thread Philippe Wittenbergh


On 16 Dec 2005, at 5:28 pm, Martin Heiden wrote:


Our statistic of last week:


[snipped]

Those numbers refer to the resolution of the monitor, right ?
What would this tell me about the size of the browser window ? That  
is what interests me.


The user may or may not have the browser window open at full width of  
the monitor.
The user may or may not have on of those sidebars open in the  
browser. That side bar may or may not be large.


That makes those numbers pretty meaningless in my eyes.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Justify this

2005-12-14 Thread Philippe Wittenbergh


On 15 Dec 2005, at 2:15 pm, Al Sparber wrote:

If you're into justified text, the only browsers that can do a  
reasonable job are IE5.5x+ Windows, using a proprietary property.  
The programming is pretty darn good. It's nearly as good as the NY  
Times, actually. But it ain't standards and must be deployed via  
Conditional Comment. Take it and hide before somebody yells at us ;-)


!--[if lte IE 6]
style
p {
text-align: justify;
text-justify: newspaper;
}
/style
![endif]--


Those text justification properties from MS are quite nice. I use it  
often for display of Japanese text. It makes Japanese text nearly  
readable on IE (now if those MS fonts weren't so ugly,...)

p {
text-align:justify;
text-justify:inter-ideograph;
}

The text-align:justify also works pretty well in Firefox 1.5, for  
East-Asian text.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/


**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] firefox 1.5 is official

2005-11-29 Thread Philippe Wittenbergh


On 1 Dec 2005, at 12:36 pm, Zach wrote:

Bah, problems already. When you use dead centering in CSS (i.e., 
background: url(picture) center center no-repeat;), it ignores the 
second center, even though it is perfectly valid CSS.


Firefox 1.5 is correct according to css2.0. Firefox 1.07 was buggy in 
this.


See
https://bugzilla.mozilla.org/show_bug.cgi?id=243751#c4

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] shrink-to-fit width in Opera

2005-11-21 Thread Philippe Wittenbergh


On 21 Nov 2005, at 8:33 pm, Andy Budd wrote:

One of my annoyances with Opera is that it calculates the 
shrink-to-fit width of absolutely positioned elements to be the 
minimum width, basically adding a break after each word. I wanted to 
write about how annoying this was, but thought I'd better check the 
specs first, just in case it was actually right.


http://www.w3.org/TR/CSS21/visudet.html#the-width-property



First, which version of Opera are you testing ?
Second, what kind of content goes into that absolute positioned element 
?


If I put an element with just some static text in, Opera 7.5 - 9 prev 
display the same as Firefox (1.6a1 nightly), Camino (1.0b), iCab, 
Safari 1.2+.
However, add a floated block in there, and then Opera is a bit more 
aggressive.[1]
Put multiple floated blocks in the AP element, and Opera 7 behaves 
differently from others. That was  a know bug in that version (not 
limited to AP elements, btw).


The spec doesn't specify exactly what should happen;  here is the key:
quote CSS 2.1 does not define the exact algorithm/quote

What exactly is a line break ? Does the end of a *floated* span 
constitute a line break in this context ?


And if you start to play with more complex constructions, the 
differences between all (decent) browser increase, see [2] as one 
example.


[1] http://dev.l-c-n.com/_temp/AP-shrink-to-fit.php
[2] http://www.satzansatz.de/cssd/tmp/100307.html


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Accessibility: Default placeholders

2005-11-13 Thread Philippe Wittenbergh


On 14 Nov 2005, at 12:22 pm, Geoff Deering wrote:

*Another* thing I see that is happening in design a lot lately is 
that input fields are in greyed background by design, not function.  
What this is telling the user is that that field is *read only*.  
That is the standard way operating systems manage read only data, 
and the same way it is done on web based systems.  It's absolutely 
sending the wrong message to users, when the input field is open to 
accept data input.



The problem with this is obviously that it's difficult to say *how* 
grey an input has to appear before a user thinks it's read only. Do 
we sit down and define a required luminescence/contrast to the 
background? In my mind, hard to quantify other than to say: be 
careful, don't make it too dark/grey, otherwise some users may think 
they can't use them.




I think it is quite simple, don't use any scale of grey at all.  Grey 
is reserved for meaning *read only*.


This makes kind of good argument for *not* styling form inputs at all, 
and leave it to the OS. On most of my OS X browsers, disabled form 
fields are not really greyed out, but rather use opacity reduction to 
indicate read-only.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] hiding legend tag

2005-11-10 Thread Philippe Wittenbergh


On 11 Nov 2005, at 7:47 am, Alexander Jerabek wrote:


So I came up with this:

legend{margin:-1em; font-size:0px;}

In Mozilla the 0px makes the text invisible, but only miniscule in IE 
so

I used margin:-1em; is to push it under the drop down menu in the form.


My question is: does anybody know if this will have adverse affects on
accessibility or if there are any other weird problems with using this
technique?


In my copy of Firefox, your legend will be 12px in size (minimum 
font-size set, us older people really do like that little preference in 
the browser ;-))


visibilty:hidden or text-indent:-1px !important are some options.

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Safari, select, and option attributes.

2005-11-05 Thread Philippe Wittenbergh


On 6 Nov 2005, at 10:54 am, ivanovitch wrote:


Apologies if this has been discussed before, but is it my bad coding,
or does Safari (v2.02, latest build) not honour any select styles?

Further, is there a way to dictate how option can display styled
font attributes across all CSS-compliant browsers? My aim is to make
the drop-down in the same style as the rest of the page, but I'm
having limited success. Firefox and IE have no problems, on WinXP and
OS X (though IE on OS X does odd things when options are selected).


Styling form widgets across browsers is notoriously difficult:
http://www.456bereastreet.com/archive/200410/ 
styling_even_more_form_controls/

(and from a usability perspective, not always the recommended option)

And no, Safari doesn't allow styling of the select and option, the  
same goes for Camino and iCab. Opera has a user option to disable  
author styling of form widgets (something I really like...).

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Is position:absolute legal in a floated div? relates to a Footer problem in NS7

2005-10-24 Thread Philippe Wittenbergh


On 25 Oct 2005, at 11:11 am, Ben Wrighton - StraightForward wrote:

After some testing I isolated the float: right; ( to make content 1st 
in source) and display: inline (which is in there for IE 's double 
margin bug and will be placed in a IE specific stylesheet in due 
course) styles as causing the issue. Float: right is the main culprit.


Which leads to the topic of this post.
'Is position:absolute legal (or advisable) in a floated div?'


Yes, it is completely valid to have an absolute positioned element 
within a floated block that acts as a containing block.
Some older browsers have problems with this, like Netscape 7. You'd 
have to add an additional wrapper inside your floated block to please 
those old boys. Like this

div style=float:right
div style=position: relative
p style=position:absolute


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Set min-width using DOM

2005-10-22 Thread Philippe Wittenbergh


On 22 Oct 2005, at 2:53 pm, Lea de Groot wrote:


On Sat, 22 Oct 2005 00:53:51 -0400, Christian Montoya wrote:

width:740px;
expression( fluid and max-width stuff )


Logically, I would expect this to work.
Set the width to 740, then reset it with the expression.
If the expression is inoperative then voila, its set to 740


Yes, that is the solution I've been using for a while, and seems to 
work fine.

selector {
width:740px;
width:expression(...);
}
(all dumped inside of a conditional comment, of course).

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] BR tag causes odd behaviour ??

2005-10-17 Thread Philippe Wittenbergh


On 17 Oct 2005, at 3:24 pm, Michael Kear wrote:

Can anyone see why the br / is causing the content to drop down 
below the

adjacent floated div in the page
http://afterlifelink.com.au/charges/index.cfm?


Because that is what you tell it to do. At the bottom of
http://afterlifelink.com.au/css/formstyles.css

br{
clear : left;
}


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Counters support?

2005-10-13 Thread Philippe Wittenbergh


On 13 Oct 2005, at 8:34 pm, Joshua Street wrote:


I discovered a page today that's all about counters and printing
things here:  
http://developer.mozilla.org/en/docs/CSS:Getting_Started:Media#Action: 
_Printing_a_document


I didn't think Firefox supported counters, though... I'd read that
Opera did (but later checked that and apparently it's
partial/incorrectly implemented - see
http://nanobox.chipx86.com/browser_support_css.php#support- 
css2propsbasic-content

). Does anyone know if Firefox does/doesn't support this, and, if it
doesn't yet, is it on the list for Deer Park?


Firefox 1.5beta (aka DeerPark) does indeed support counters according  
to the (current) CSS 2.1 draft. Opera 8.5 and (I think) 8.02 supports  
that syntax as well. And iCab 3.0.


Check bug 3247 which implements it in Gecko
https://bugzilla.mozilla.org/show_bug.cgi?id=3247
There was a recent article on ALA where the subject was mentioned more  
in detail in the comments.

http://alistapart.com/articles/multicolumnlists

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Safari FOUC was site check: liquid.rdpdesign.com

2005-10-12 Thread Philippe Wittenbergh


On 12 Oct 2005, at 11:58 am, Steve Ferguson wrote:


It does appear that you've killed the Safari FOUC.

You should document this as appears to be somewhat of a mystery.


I'll second this. I haven't followed the discussion, so it is hard to 
check for a 'before' and 'after'.


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Safari FOUC was site check: liquid.rdpdesign.com

2005-10-12 Thread Philippe Wittenbergh


On 12 Oct 2005, at 3:52 pm, Christian Montoya wrote:


Where would I document this?


Blog ? Own website ?

You could start here as well, or add an item to the CSS-D Wiki:
http://css-discuss.incutio.com/?page=FrontPage

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Firefox caption madness

2005-10-11 Thread Philippe Wittenbergh


On 11 Oct 2005, at 3:49 pm, Jake Badger wrote:



It already has it, but firefox ignores it if the display isn't block,
and if it is it's only as wide as the first cell of the table (as I
just said).


If you set the display value for the caption to 'block', then it is not 
a 'table-caption' [1] anymore..., and Firefox, Opera, Safari do display 
it wherever they feel like, still part of the table anonymous block, 
but they'll attempt to build a tr/td around it according to the rules 
in 17.2 [2].


Next, if you want to positioning something absolute within the 
table-block, you'll bump into an old Gecko bug, where a table-element 
cannot act as a 'nearest positioned ancestor' for the AP element [3].


Plus, a table caption cannot contain block level elements (your h5).

[1] default display-value for caption
[2] http://www.w3.org/TR/CSS21/tables.html#anonymous-boxes
[3] https://bugzilla.mozilla.org/show_bug.cgi?id=63895
Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] :after/:before used for layout

2005-10-11 Thread Philippe Wittenbergh


On 12 Oct 2005, at 8:48 am, Gunlaug Sørtun wrote:


Jan Brasna wrote:
:after generated content cannot receive some CSS properties, 
including 'position', 'float', list properties, and table 
properties.


That's CSS2. Can't find that line in CSS2.1.
All those properties are perfectly fine in CSS 2.1. Unfortunately, 
those are not yet supported in Gecko browsers

https://bugzilla.mozilla.org/show_bug.cgi?id=238072

[...]

Firefox won't even react to 'margin', as exemplified by W3C (CSS21),
while Opera, Safari, and even iCab, are doing fine with simple 'margin'
and absolute positioning. Only Opera seems to be able to handle
'positioned :hover:after'.


?? margin seems to work perfectly fine here. Both on Firefox 1.6a1 
nightly trunk build and Firefox 1.5beta 2

http:dev.l-c-n.com/_temp/after.php
(the word blah in a grey box after each paragraph, and the word 'the 
End' as body:after)



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] :after/:before used for layout

2005-10-11 Thread Philippe Wittenbergh


On 12 Oct 2005, at 1:00 pm, Gunlaug Sørtun wrote:


Philippe Wittenbergh wrote:
?? margin seems to work perfectly fine here. Both on Firefox 1.6a1 
nightly trunk build and Firefox 1.5beta 2 
http:dev.l-c-n.com/_temp/after.php (the word blah in a grey box

after each paragraph, and the word 'the End' as body:after)


Working at my end too - once I got my gray cells cranked up.
- FF need 'display: block' since it doesn't get 'block' through 'AP'.


It works with inline elements as well :-); but vertical margins won't 
affect the flow.



Now, how do one add an empty 'alt-attr' to an image introduced as
'content'. The reason for asking is that Opera show image when images
are off and during slow loading, and that's not quite good enough.


I don't think it is possible with the current CSS 2.1 syntax. IIRC, 
there has been some discussion about this on www-style, without any 
conclusion so far. More like a 'feature request'. Probably something 
for CSS 3.



Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Jello flexible layout

2005-10-05 Thread Philippe Wittenbergh


On 5 Oct 2005, at 10:19 pm, Tom Livingston wrote:

However, I managed to get the designer to go another direction with 
reguards to full height. But full height using Jello Mold is tricky 
because one of the key structural elements must have a height of 0 
within a Holly Hack for IEWin. Traditionally, the hack is used with 
height:1% (for hasLayout I believe) but anything other than 0 breaks 
this layout in IEWIN.


What Jello Mold need is 'hasLayout' [1] in order to paint those 
containers correctly in IE. Hence the use of the 'holly hack', which is 
one the methods to trigger hasLayout.
There are other methods though: you could try 'zoom:1', or 
'display:inline-block'. Then use and additional container/wrapper to 
force the height.


[1] http://www.satzansatz.de/cssd/onhavinglayout.html

Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Flexible Font sizes in tables in ie

2005-10-04 Thread Philippe Wittenbergh


On 5 Oct 2005, at 12:50 pm, Kara O'Halloran - Eduka wrote:

Either way I still have the problem of the table cell text either 
appearing too large in ie, or too small in ff.


Without resorting to setting text size in pixels in my table, I can't 
find any other way to prevent this from happening, I thought perhaps 
there might be a hack out there that will pass 0.8em to ie, but not to 
firefox?




The problem is the table element in IE that doesn't correctly inherits 
the font-size.

What works for me:
body {font-size: 100%} /* or whatever you want, but in percentages */
table {font-size:1em;} /* makes sure the font-size is inherited in IE, 
this has to be explicitly specified */


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



Re: [WSG] Site Review: www.47words.com

2005-10-03 Thread Philippe Wittenbergh


On 3 Oct 2005, at 5:47 pm, Gunlaug Sørtun wrote:


I do have one question that just came up while I was chatting with a
 friend:  Does the hreflang attribute on links do anything, really,
 or is it just a cosmetic thing that no browser does anything with?
 (I'm using it on links to materials in Mandarin/Malay to indicate 
that the content of those sites/pages is in a language other than 
English.)


Not updated for the latest browser-versions, but...
http://www.w3.org/International/tests/results/link-element


That list is -unfortunately maybe- quite up to to date. One could add 
iCab 3.0 which has a similar 'navigation' bar as Opera and Mozilla to 
the list that do something in the UI with some of the link attributes.


The hreflang attribute isn't used by any browser, I think. You can 
style it, though, via css2 attribute selectors. Something like this:

a[hreflang]::after {
content:  [attr(hreflang)];
font-style:italic;
color:#6c767f
}


Philippe
---
Philippe Wittenbergh
http://emps.l-c-n.com/

**
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list  getting help
**



  1   2   3   >