[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Mathematica symbol fonts

> Having printed the mma symbol fonts, I have the following remarks:
> a) On integrals:

> The mma fonts contain a class of integrals and other big operators larger 
> than \displaystyle. For a moment I played with the thought of an e-TeX
> proposal \bigdisplaystyle (appropriate, if there is a displaystyle fraction 
> in the ``argument'' of a big operator) but discarded it soon. The reason 
> for discarding the thought was, that the argument of a big operator cannot
> be detected automatically.

I guess you're right in that TeX can't make any use of more than two
sizes of big operators, i.e. the only distinction made when setting
big operators is between \textstyle and \displaystyle.  It is possible
that Mathematica uses a wholly different approach, so that it could
make use of three sizes.  Personally, I think that the design of many
of the big operators in Math2 is really ugly and looks much better if
you fake them by using enlarged sizes of normal textstyle operators.

> Maybe setting up the integrals as delimiters can make the largest size 
> available ... I will have to think further about this approach.

As for big integrals as a delimiters, both Adobe Symbol and Math1
contain pieces to construct an extensible integral (top, bottom,
repeatable piece).  I wonder if we should reserve slots for such a
beast in the MX1 font table.  The problem is that it would have to 
be addressed differently than normal integrals.

> c) A special symbol:

> The is a ``mouth'' like symbol near \approx in the mma fonts, consisting of
> smallfrown over smallsmile. What is it used for?

No idea.  It is interesting, however, that `smile' and `frown' are
grouped with the relation symbols, so that you get `smilefrown' and
`frownsimle' glyphs as well as combinations with equal signs.

> d) \de, \ee, \ie

> There are special symbols for the d in differentials, the base of the 
> natural logarithm and the square root of -1. However, Wolfram's designs are 
> just plain ugly and not supported anywhere else (mathroman type being the 
> design of choice).

I think it is a positive aspect that the Mathematica symbol fonts have
reserved dedicated slots for DifferentialD, ExponentialE, ImaginaryI,
ImaginaryJ, and CapitalDifferentalD in Math1 and Math3.  As for their
designs, it is pretty clear that these symbols (double-struck italics
letters) are intended specifically for use with Mathematica and do not
match the usual typesetting conventions applicable in physics.  It is
only a matter of making a distinction between normal letters and those
that have a specific meaning.  BTW, the Mathematica book mentions that
only the lowercase \pi has reserved meaning by default, while all
other Latin and Greek alphabetic symbols don't have one.

> e) Other

> I found the capital Epsilon (like african open E) also in the mma 
> fonts. Worth considering for the encoding, I think.

I think this is meant as a symbol to denote some Euler number.
Doesn't Unicode also have a slot for this in the 0x21xx page?

Cheers, Ulrik.

P.S. About \de, \ee, \ie: I don't think these macros are the best
method for encoding the upright d, e, i (although you're advertising
them in your book, Joerg).  If possible, I'd prefer \d, \e, \i
employing a switching macro like


where \textd, \texti, \textj refer to the encoding specfic text
symbols normally accessed by \d (dot-under accent), \i or \j.

BTW, the following is what I found in various journal classes:

Elsevier: elsart.cls


IOP, LaTeX 2e: iopart.cls


IOP, LaTeX 2.09: ioplppt.sty

\newcommand{\e}{{\mathrm e}}
\renewcommand{\i}{{\mathrm i}}
\renewcommand{\d}{{\mathrm d}}