W3C home > Mailing lists > Public > www-style@w3.org > July 2007

Re: Block Model Additions: Float Positioning

From: James Elmore <James.Elmore@cox.net>
Date: Fri, 20 Jul 2007 16:26:39 -0700
Message-Id: <2B6E683C-1455-4BFF-A4B0-9D2FC3561449@cox.net>
To: www-style@w3.org


On Jul 20, 2007, at 3:30 PM, L. David Baron wrote:

> On Friday 2007-07-20 15:11 -0700, James Elmore wrote:
>> I’m not sure exactly how flowing the text boxes and blocks around the
>> ‘floated’ block would work, especially when the floated element
>> extends into another block. Also, positions where there is limited
>> space (e.g., where only short words would fit on one side of the
>> floated element) might allow some words to flow, skipping across the
>> floated object and leaving other line blocks empty. This is all open
>> to discussion, if the primary proposal is accepted.
>
> I don't see how this proposal could be "accepted" without this key
> part.  A key part of getting the proposal accepted is getting the
> major CSS layout engine implementors to implement it, which requires
> coming up with proposed details for this that are implementable in
> their layout engine architectures and compatible with the rest of
> the specification.  In other words, this is the hard part.

I can foresee some of the problems with flow, but I'm sure I can't  
predict them all. I had hoped that there might be discussion about  
what might and might not work in the flow and float and that some  
people with more experience with the possible problems might offer  
their suggestions. This would be preferable to my making a hard and  
fast declaration of 'how it will work' and then having people tell me  
it would never work that way. If the group prefers to have complete  
proposals from the start, then having something to pick apart, rather  
than the (to my mind) more positive process of working together, I  
could certainly start the ball rolling that way.

Perhaps I am misusing the term 'accepted.' What I hoped was that  
there might be interest in the proposal, even negative interest from  
people willing to be convinced that what I suggested might be  
possible. With interest, enough dialog could work out a completed  
proposal. I was not expecting my email to be incorporated into the  
next revision of the CSS specification, verbatim.

>
> That said, I'm rather skeptical of the idea to begin with, even
> without a detailed proposal.  I tend to think it will (like absolute
> positioning itself) produce fragile layouts that only work over a
> very narrow range of font sizes, viewport sizes, and often even font
> metrics.

You might be right about the 'fragility' of absolute layouts. I need  
to think about it some more before I agree or disagree. My intent was  
to expand what is possible, using pieces of CSS functionality that  
people might already know.

Does anybody out there know if the 'absolute' 'relative' and 'fixed'  
positioning features are being used? If there are significant numbers  
of designers using these features, then they may never go away.  
(Significant numbers may be as low as a few percent of designers.  
Never going away may be a short as a decade.) I have used and taught  
these features because there are some very nice layouts which are  
only possible using them.

> I'd prefer to see the absolute positioning model be a dead
> end that isn't developed further, and see future development go into
> more flexible layout models.
>
> -David
>

I read the CSS Advanced Layout Draft specification. Is that what you  
meant by 'more flexible' layouts?

Some of the Advanced Layout features seem very nice and some seem  
like they will make the designers' jobs harder. But at the end of the  
spec, the authors state that they don't have a consensus about what  
an ideal layout mechanism might be. If my suggestions make someone  
else think of an even better way to position floats, then I will be  
glad to have worked on, and then abandoned, this suggestion. I am  
trying to provoke ideas, not anger. I am also definitely not stating  
that my suggestions are the perfect, only that I see something  
missing from the current feature set and offering one possible way to  
add that feature.

> -- 
> L. David Baron                                 http://dbaron.org/
> Mozilla Corporation                       http://www.mozilla.com/

Since you list mozilla.com as one of your addresses, I expect you  
have far more experience with styles (and programming them) than I  
do. Do you see additions to floats as worthwhile? How would you  
expand what a designer could do with floats? Let's see if we can work  
together and provide something new and wonderful.

James Elmore
Received on Friday, 20 July 2007 23:27:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:51 GMT