X-VM-VHeader: ("From:" "Sender:" "Resent-From" "To:" "Apparently-To:" "Cc:" "Subject:" "Date:" "Resent-Date:") nil
X-VM-Bookmark: 1
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 "^From:" nil nil nil])
From: Rainer Schoepf <BK4@DHDURZ1.bitnet>
To: LaTeX discussion list <LATEX-L@DHDURZ1.uucp>
Subject: What we want to change in the style file interface
Date:  Mon, 05 Feb 90 11:49:47 CET
Status: R
X-Status: 
X-Keywords:                  
X-UID: 6

I'm forwarding this on behalf of Don Hosek.
   Rainer
----------------------------Original message----------------------------

I think that the first thing that we need to do is determine
exactly how we want to change the style interface. The primary
pieces seem to be:

(1) Interface to sectioning. \@startsection is not general enough
for all applications (for example, it cannot produce section
headers like:

    1.
    _Paradise Lost_ and pseudepigraphical sources

    Section 2     The Origins of the English Sonnet

etc.)

(2) The list and trivlist environments. Are they adequate? This
has not been mentioned on this list at all. However, Rainer
Schoepf and Frank Mittelbach point out the following problems
with these environments (quoted directly from their paper)

    * An actual conceptual bug was the decision to add the value of
    \parskip to all vertical spacing parameters, even when it is
    used in places where no paragraph ends.

    This means that changing this parameter influences the layout
    in unexpected places, which in turn means that other
    parameters must be adjusted unnecessarily to compensate for
    this undesired side effect.

    * There is also the problem that the \parskip parameter is
    user-accessible while the affected parameters are only
    changeable in a style file. The user can change the \topsep
    parameter, for example, but its default value, defined in the
    style file, will be restored later on.

    * The resetting of parameters to their default values if
    nothing else is specified is somewhat arbitrary. Some of the
    important parameters (i.e., the penalty values for page
    breaking before and after the list) get their values from the
    surrounding list which is more than a nuisance for a style
    designer. (As an unpleasant result the LaTeX fleqn style
    option (which causes displayed equations to be typeset flush
    left instead of the usual centering) favours pagebreaks just
    before displayed equations. (This bug is corrected in the
    LaTeX version of 24 May 89.))

    * Another implementation decision makes it impossible to define
    lists with page-wide labels, i.e., with labels placed on a
    line by themselves.  Those lists cannot be nested properly
    see, for example, comments in the article about the
    implementation of the extended theorem environment (TUGboat
    10#3).

    ...

    * There is no possibility of generating a run-in list i.e., a
    list where the first item runs into the preceding text. This
    feature is provided in AmS-TeX.

    * More generally, since the layout of the standard list types
    is fixed by the document style selected, there is no way for the
    user to select different kinds of layouts for the same kinds
    of lists (e.g., enumerated or itemized lists) without
    defining his own environments. It would be better to provide
    a way to specify attributes such as "compact", "stream" (see
    for example "Stream lists and related list types for LaTeX"
    by Reinhard Wonneberger, TUGboat 6#3), "run-in", etc.

    * The vertical space before and after lists is controlled by
    the same parameter

    * The vertical space preceding the first item does not depend
    on the length of the last line of the preceding paragraph (as
    is the case for displayed equations.

(3) Support for easier national language support. As has been
suggested in numerous places at numerous times, all the English
text "hard-wired" into LaTeX should be changed so that it can be
easily changed into a more useful language (e.g., Latin).

(4) I'm sure that there is more than this that we will want to
change, but this makes a good start. I would like to hear from
everybody out there to hear if they can think of anything
missing; I will then post a revised version of this list.