W3C home > Mailing lists > Public > www-style@w3.org > December 2008

Re: Adjusting float height on paged media

From: Anton Prowse <prowse@moonhenge.net>
Date: Mon, 22 Dec 2008 16:09:22 +0100
Message-ID: <494FADA2.8050005@moonhenge.net>
CC: W3C Style List <www-style@w3.org>

Henri Sivonen wrote:
> On Dec 22, 2008, at 15:50, Håkon Wium Lie wrote:
>>  .figure {
>>    float: none, top next page;
>>  }
>>
>> The first value ("none") means that figures are not floated. However,
>> I a figure will not fit on the current page, the next value in the
>> comma-separated list is used instead. So, if there isn't room for the
>> figure on the current page, it will be floated to the top of the next
>> page.
>>
>> [1] http://dev.w3.org/csswg/css3-gcpm/#page-floats
>>
>> When the figure is floated to the next page, it allows the original
>> page to continue to be filled, thereby avoiding unused space.
> 
> This works when the adjacency of text and images is relatively elastic 
> in the opinion of the author. However, in the case I currently have 
> under consideration (case #1 in my previous email; cropping) float: 
> left; and careful placement of each image next to a given paragraph is a 
> relatively inelastic requirement. That is, float: left; is desired even 
> if it causes both the image and text move to the next page.

Then what you want is not really a float, at least not in the 
traditional printed media sense of the word.

Even if you weren't using floats but were using a table row (say), an 
oversize image would cause the same "drop" that you describe, so the 
issue you face is not about floats but about having more control over 
image sizing and cropping.  In the end, it's very difficult to 
programatically control page layout to that level of editorial demand; 
with even the most sophisticated typesetting and page layout software 
the human touch is required at the end to fix issues (particularly with 
regard to the positioning of traditional floats).  Any attempted 
solution in a non-wysiwyg system such as CSS is likely to be at least 
partly unsatisfactory; it's the nature of the beast.

Cheers,
Anton Prowse
http://dev.moonhenge.net
Received on Monday, 22 December 2008 15:10:15 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:55:18 GMT