Re: Anchoring text to bottom of page

Ken,

> 
> Remember that XSL-FO is a layout language and if there is a layout 
> construct giving you what you want, then ignore what the name of the 
> construct is and use it to do what you need.

It would be really nice if it were so.  It is not, unfortunately. XSL FO
is a mix of lower-level layout elements, presentation elements which have
more or less fixed layout role, as well as of higher-level elements describing
a document's structure.

While your advice is absolutely right for tables or lists -- data put in tables
is not necessary tabular data -- it is less safe to use footnotes for positioning
signatures and disclaimers, unless the signatures and disclaimers ARE footnotes
indeed.

An alternative approach would be to omit the notion of footnote completely and introduce
'bottom-floats' along with 'top-' and 'side-' floats. Since currently XSL FO is a compromise
between a layout assembler and a document structure description language, points where
the language uses semantically loaded constructs should be regarded and the constructs
used according to their intended roles and meanings.

table-caption, footnote, title are structural elements. They should be used in accordance
with their meaning; a somewhat exaggerated example of importance of such use is a speech-based
XSL FO browser; it would interpret this elements according to their intended roles and would be
confused. While an XSL FO reader is mostly a theoretical issue, an XSL FO browser for PDAs
or other special devices may interpret these constructs in a semantically reasonable but
rather unexpected way. 

Besides that, use of footnote as a bottom-float breaks extensibility and forward-compatibility:
unexpected effects can eventually be discovered when bottom-floats are indeed included in the language.

> 
> If you had meant the bottom of every page, check the use of static-content 
> for a footer.

Isn't specifying a different page-master for the last page and static content
just for that page  a better solution?

David Tolpin

Received on Monday, 29 July 2002 15:40:39 UTC