pagebreaks (DOCBOOK: Re: Which conversion method should I use?)

["Matt G." <matt_g_ at hotmail dot com>]

> From: Jose <studio@qadram.com>
> To: docbook@lists.oasis-open.org
> Subject: Re: DOCBOOK: Re: Which conversion method should I use?
> Date: Sat, 22 Jun 2002 20:28:43 +0200
>
> For example, in this kind of document:
>
> 1.1 My section
> The following graphic blah, blah, blah...
> [GRAPHIC]
>
> If I render to PDF, the Graphic is big and it doesn't fit in the
> remaining space of the page, so the page is seen on PDF as:
>
> 1.1 My section
> The following graphic blah, blah, blah...
>
>
> [EMPTY SPACE]
>
>
>
>
> --------New page-------------
> [GRAPHIC]
Ah, yes.  First, consider that this is fundamentally a formatting problem.  
It's a relevant consideration, at the DocBook level, because the formatter 
should be given enough information about the logical and structural 
relationships of constructs to do a decent job of formatting.



> I would like to be able to place a pagebreak just before My Section to
> force the presentation display all the content in the next page,
> along with the graphic.
I disagree.  What you want is for the text describing the graphic to be on 
the same page as it.  Manually inserting pagebreaks can be a lot of work, 
and probably needs to be redone, every time content in preceding sections is 
added or removed (or you'd end up with numerous nearly-blank pages).  
However, if you could specify *groupings*, they would only require updating 
when the content or semantics of the enclosed or surrounding elements is 
modified.  In other words, manual pagebreaks could be a maintenance 
nightmare, whereas explicit groupings would be localized and self-contained.

As far as I can tell, it's really an optimization problem, with no "right" 
answer.  For example, would you really want to leave a preceding page nearly 
blank, *just* to get some text and the corresponding image on the same page? 
 At some point, flipping back & forth isn't so bad.  ...and what if the two 
pages are facing (if the output is printed as a book) - do you really need 
the text & figure to be on the same page, then?



> Is this possible? Maybe is a feature in DocBook I don't know cause I'm
> a newbie.
DocBook tries to focus on the meaning and structure of your information - 
not it's presentation.  Sometimes, such as in the case of tables, the 
presentation is coupled to the structure of the information in relationships 
of sufficient complexity and variety that it's not worth trying to maintain 
the abstraction.  However, I don't believe this is a similar case.  I think 
an argument could be made for a grouping element, since it would essentially 
provide metadata that "these elements are more closely coupled to each other 
than the rest of their siblings", and could be used as a layout hint, by 
processing applications.  Its 'role' attribute could even be used to specify 
the nature of the relationship.

BTW, this could be even more powerful, if you consider having a hierarchy of 
grouping elements.  However, something of that sort suggests that you should 
probably introduce more explicit structure (i.e. 'chapter's, 'section's, 
etc.).


Has anyone had a similar idea?  Better yet, has anyone actually used this, 
in some tool, or seen it work?  Is it worth suggesting as a RFE?  If not, 
how do you propose to solve this problem - or why do you think it's not a 
real problem (it's certainly not the first time I've heard about it)?


Matt Gruenke


_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx