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 10:15:00 UT+2 Reply-To: LaTeX-L Mailing list From: POPPELIER@HUTRUU53.bitnet Subject: Re: Proposal for .. (by Rainer) To: Rainer Schoepf Status: R X-Status: X-Keywords: X-UID: 204 Rainer's 'Proposal for a section heading concept' is, in my humble opinion, not an elegant one: there are conventional headings, exceptions to this, and exceptions to exceptions. The scheme is not a 2-level, but a 3-level one. We should try to define a scheme that is as simple as possible. 1. It has been discussed several times that no macro can be designed to handle *all* possible section-heading layouts. Most of the people on this list agree on that, I would say. Solution: the macro that specifies the section-heading layout should reside in the document style -- see Frank's earlier proposal -- and not in LaTeX's kernel, as is the case in version 2.09. 2. It *is* possible, however, to enumerate *all possible positions* of the section heading -- Rainer calls this 'the rest of the cases'. And if you can enumerate them, you can turn the position of the heading into a parameter to the 'generic' macro. | | Let |___________| stand for the preceding paragraph, ****** for the sectional-unit heading, and |-----------| for the next paragraph | | Then the most following 4 positions are most likely two occur | | | | | | | | |___________| |___________| |___________| |___________| ****** |-----------| |******-----| ******|-----------| |-----------|****** | | | | | | | | and maybe also the following 2, although I find these a bit unlikely. | | | | ******|___________| |___________|****** |-----------| |-----------| | | | | Cases 1 and 2 are implemented in \@startsection of version 2.09, with the actual layout hard-wired into the kernel macros. All cases except number 2 can be generated by filling a box with the appropriate contents (document style specifies layout), and putting the box in the appropriate place (parameter to generic section macro). Case number 2 is a matter of letting the title of the sectional unit start the first paragraph. Conclusion: levels 1 and 2 of Rainer's proposal can be made into one. 3. Now the question is: are there indeed exotic layouts that cannot be implemented in this way? Maybe I lack imagination -- not unlikely: I'm not a designer, just a physicist -- but I can't think of any! Neither have I found section-heading layouts here at ESP that do not fit into this scheme. But: for exceptional cases there should be an escape. Nico Poppelier Elsevier Science Publishers, PSED, R&D Department Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands Phone: +(20)5862504. Fax: +(20)5862425. Email: n.poppelier@elsevier.nl