X-VM-v5-Data: ([nil nil nil nil nil nil nil t nil]
	[nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil])
Date:         Thu, 19 Jul 90 13:58:50 CET
Reply-To: LaTeX-L Mailing list <LATEX-L%DHDURZ1@unido.bitnet>
From: bbeeton <BNB@MATH.AMS.COM>
Subject:      more section heading examples
To: Rainer Schoepf <BK4@DHDURZ1.uucp>
Status: R
X-Status: 
X-Keywords:                  
X-UID: 205

i've had these heading examples hanging around for awhile, and now that
we're back on the subject, i submit them as something that can be used
for testing.  they're not at all easy to handle with latex 2.09.

in one case, the designer specified that a first-level section heading
is to omit section numbers and the heading line is to be followed by a
horizontal rule.  second- and lower-level heads are also to be set on
a line by themselves; the second-level head gets a rule too, but the
third-level doesn't.

the obvious way to implement this, one would think, is to tag all the
headings with \...section*, and test the level in \@ssect to do the
appropriate thing.  there's a problem, though.  \@sect accepts the
counter and level as the first two arguments, plus four more arguments,
but \@ssect accepts only the last four arguments -- no level.

the relevant point here is that the style of a heading may include
something in addition to skip following the heading text, and what
that is (as well as other parts of the style) may depend on the level.

another problem is that t-of-c entries may be wanted for some of these
headings irrespective of the presence of section numbers in the text,
a possibility precluded by the present \...section* definitions.

in another level-specific case, instead of a font change, there was an
instruction to uppercase the third-level heads.  the intent of this is
similar to that of a font change, though, in that it shouldn't affect
the text in any way that makes it necessary for the user to enter a
separate t-of-c text.

clearly these variations should be defined in the document style, not
the kernel, as long as the basic structure has enough flexibilty.  i
think that such level-specific variations have already been accommodated,
but another test example or two shouldn't hurt.
                                                -- bb