Message-Id: <199312161302.AA07645@mail.cs.tu-berlin.de>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@DHDURZ1.BITNET>
Date:         Thu, 16 Dec 93 12:12:19 GMT
From: P.Taylor@rhbnc.ac.uk
Sender: Mailing list for the LaTeX3 project <LATEX-L@DHDURZ1.BITNET>
To: Multiple Recipients of <LATEX-L@DHDURZ1.BITNET>
Subject:      RE: Preceding-page figure placement
Status: R


Michael --

>> I must admit to puzzlement about when such a placement (figure on a
>> preceding, facing page) would actually occur. Can you give some
>> examples?  If a figure reference occurs 5cm from the bottom of page 59
>> (right-hand page) and the figure is 6cm tall, I don't see how moving
>> the figure to page 58 will help anything: it will most likely push the
>> figure reference to page 60, leaving the figure once again not visible
>> from the point of reference.

Although there are many situations within which your analysis is completely
correct, there are others within which the removal of a troublesome figure to a
preceding page could, in fact, bring about the desired result.  Let us assume
first of all that figures must occur in the order of citation; thus as
references are satisfied, the reference generator will supply (for example)
`See Fig.~3.1', `See Fig.~3.2' and `See Fig.~3.3', and the figures must there-
fore occur in the order 3.1, 3.2 and 3.3 (anything less than this is counter-
intuitive, and will cause the reader frustration in searching for the figure of
interest).  Then suppose that Figures 3.1 and 3.2 have already been tentatively
placed on the recto sheet, and Figure 3.3 is referenced, as you suggest, at a
point close to the bottom of the recto sheet.  Figure 3.3 is also deeper than
the free space remaining on the recto sheet. Now clearly Figure 3.3 won't fit
in situ. But suppose that Figure 3.1 can be moved from the recto to the verso
sheet, and that in so moving it displaces a portion of text from the verso
sheet. We might expect that the depth of the portion of text displaced is equal
to the depth of Figure 3.1, but this is not necessarily the case: in making up
the verso sheet, the makeup system may have forced it to run short, in order to
prevent (for example) a section header occurring in isolation at the bottom of
the page; if we are treating the spread as the basic unit of makeup, then the
matching recto sheet will also be required to run equally short, and thus the
free space on the recto sheet will be less than it would have been had the
verso page run to full length.  Thus we have two facing pages, both of which
have less available space than might otherwise be the case.  Now, if Figure 3.1
is moved to the verso sheet, it will probably displace the offending section
header and possibly allow the page to run to full length.  This will have a
concommitant `knock-on' effect on the makeup criteria for the recto sheet,
which now too should run to full length.  In addition, the textual material
displaced from the bottom of the verso sheet will have a natural depth rather
less than the space which it occupied on the verso sheet, given that in the
scenario outlined we _know_ that the preceding sheet ran light.  Thus we have
potentially gained at least twice the under-run to augment the space available
on the recto sheet within which to place Figure 3.3, and a problem which would
be insoluble were the page to be adopted as the basic unit of makeup becomes
soluble if the spread is adopted instead.

					Philip Taylor, RHBNC.