- Instructions for using AMS documentclasses are no longer included
in amsldoc.tex (also known formerly as amslatex.tex). They are
now found in instr-l.tex in the classes subdirectory. This is
a copy of the file instr-l.tex that is found in the
author-info area of e-math.ams.org. That area contains
additional information pertaining specifically to the submission of
or AmS-/ documents to AMS publications.
- As mentioned earlier, a companion document class amsproc is
now provided for books that consist of a collection of articles, such as
the proceedings of a conference.
- The abstract should now be entered before the
maketitle command. This provides maximum control over the position
and vertical spacing when printing the abstract information at different
places, as required by different AMS publications—for example, in some
publications the abstract may be required to fall between certain other
elements in the beginning section of a document, or it may be postponed
to the end of the document, or it might even be omitted entirely. If an
abstract is entered after maketitle, it will still be printed, but
with a warning (and possibly in the wrong place or with wrong spacing).
- As the AMS document classes automatically load the amsmath
package, all changes described above for the amsmath package
affect the AMS classes also. Note in particular that the
character is no longer a special command prefix, so to produce a printed
in an e-mail address you should now write simply instead
of |@@|.
- The nonbreaking dash commands |@-|, |@–|, |@—| are no longer
available because as a special command prefix has disappeared.
An alternative nobreakdash command is provided in the amsmath
package, as described above.
- The AMS document classes now take an option nomath that
suppresses the automatic loading of the amsmath package. This is
sometimes helpful in converting a pre-existing document to AMS style if
you only want to do a quick conversion without attempting to
deal with any of the math.
- The old environments pf, pf* are superseded by a
single proof environment. It takes an optional argument to specify
an alternative heading text.
- Formerly small produced the same typesize as
footnotesize (8pt). Now they produce different sizes, normally 9
and 8 respectively. Furthermore, documentclass options
8pt,
9pt, 11pt, 12pt are now offered in addition to the
default 10pt. Also the range of typesize-changing commands is
filled out to include, below normalsize, the following: small,
Small, SMALL, tiny, Tiny. footnotesize and
scriptsize are retained as synonyms of Small and SMALL
respectively. As it stands tiny no longer produces the same
typesize as it did before (now 6 instead of 5), except in compatibility
mode. This might affect some existing documents if they are updated to
use documentclass instead of documentstyle. Similarly,
large now produces 11pt if the base size is 10pt, where formerly it
produced 12pt. Finally, new commands larger and smaller are
provided for changing the typesize relative to the current size. These
commands take an optional integer argument to specify how many steps to
go up or down: \larger[2] means go up two sizes, and
\larger without an optional argument is the same as
\larger[1]. The list of sizes is a standard progression of type
sizes and associated linespacing values as defined by the documentclass.
If you need finer control over type size or linespacing, consult the
documentation about the fontsize command.
- The qed symbol is now an open square, not a filled black
square, and it is positioned at the right margin, instead of at a fixed
horizontal distance from the preceding text.
- The command rom, for making numbers and punctuation
roman/upright in italic text, has been renamed upn ``upright
punctuation or number''. It will in most cases suffice to leave this
refinement undone until a document is sent to a publisher for final
typesetting, and furthermore, certain publications at the AMS now use
special in-house italic fonts that have upright numbers and punctuation
built in, making the use of upn unnecessary even then. For general
preprint-type use, this refinement is one that most users probably won't
care to bother with.