\varepsilon with mathptmx
Joseph Wright
With the minimal example
\documentclass{article}
\usepackage{mathptmx}
\begin{document}
\( \epsilon \varepsilon \)
\end{document}
the two symbols are the same. This seems very awkward! Is this a known
issue?
--
Joseph Wright
Enrico Gregorio
The two glyphs should never appear together in the same document:
they are only graphic variations of the same symbol. Usually I advise
to say
\renewcommand{\phi}{\varphi}
\renewcommand{\epsilon}{\varepsilon}
if one wants the "continental" shape. It would be easy, then, to change
the symbols into "British" shape, maybe because that's required by the
publisher.
The fact that the two commands choose the same symbol with mathptmx
is easily explained: the font Adobe Symbol lacks an alternative form
for epsilon and rho. It has one for phi, pi and theta.
Ciao
Enrico
Joseph Wright
> The two glyphs should never appear together in the same document:
> they are only graphic variations of the same symbol. Usually I advise
> to say
>
> \renewcommand{\phi}{\varphi}
> \renewcommand{\epsilon}{\varepsilon}
>
> if one wants the "continental" shape. It would be easy, then, to change
> the symbols into "British" shape, maybe because that's required by the
> publisher.
That's not how everyone sees it :-) (My question comes from a user of my
achemso package, where he does want to use the two symbols for different
things.)
> The fact that the two commands choose the same symbol with mathptmx
> is easily explained: the font Adobe Symbol lacks an alternative form
> for epsilon and rho. It has one for phi, pi and theta.
Ah, right. So how do I get the CM one for \varepsilon?
--
Joseph Wright
Enrico Gregorio
> On 25/09/2010 09:13, Enrico Gregorio wrote:
> > The two glyphs should never appear together in the same document:
> > they are only graphic variations of the same symbol. Usually I advise
> > to say
> >
> > \renewcommand{\phi}{\varphi}
> > \renewcommand{\epsilon}{\varepsilon}
> >
> > if one wants the "continental" shape. It would be easy, then, to change
> > the symbols into "British" shape, maybe because that's required by the
> > publisher.
>
> That's not how everyone sees it :-) (My question comes from a user of my
> achemso package, where he does want to use the two symbols for different
> things.)
He shouldn't. :-)
> > The fact that the two commands choose the same symbol with mathptmx
> > is easily explained: the font Adobe Symbol lacks an alternative form
> > for epsilon and rho. It has one for phi, pi and theta.
>
> Ah, right. So how do I get the CM one for \varepsilon?
You need the \epsilon, actually,
\renewcommand{\epsilon}{\text{\usefont{OML}{cmr}{m}{n}\symbol{15}}}
but I'm afraid they don't mix very well. Enclosing it in \text
(requires amsmath or, at least, amstext) ensures that the symbol
changes size in subscripts or superscripts.
Ciao
Enrico
Joseph Wright
> You need the \epsilon, actually,
>
> \renewcommand{\epsilon}{\text{\usefont{OML}{cmr}{m}{n}\symbol{15}}}
>
> but I'm afraid they don't mix very well. Enclosing it in \text
> (requires amsmath or, at least, amstext) ensures that the symbol
> changes size in subscripts or superscripts.
Marvellous, thanks. I'll let my achemso user know.
--
Joseph Wright
Guenter Milde
Is there a warning, at least?
> Ah, right. So how do I get the CM one for \varepsilon?
Why not using a font package that provides the required shapes, e.g.
mbtimes, tgtermes-math, qtxmath, txfonts or wrisym. (Also eventually
fourier, mathdesign or MnSymbol.) A (German) survey of math fonts is
available at
http://milde.users.sourceforge.net/Matheschriften/matheschriften.html
Günter
Enrico Gregorio
> On 2010-09-25, Joseph Wright wrote:
> > On 25/09/2010 09:13, Enrico Gregorio wrote:
>
> >> The fact that the two commands choose the same symbol with mathptmx
> >> is easily explained: the font Adobe Symbol lacks an alternative form
> >> for epsilon and rho. It has one for phi, pi and theta.
>
> Is there a warning, at least?
Why? The two symbols are just graphical variants which should never
get different meanings.
> > Ah, right. So how do I get the CM one for \varepsilon?
>
> Why not using a font package that provides the required shapes, e.g.
> mbtimes, tgtermes-math, qtxmath, txfonts or wrisym. (Also eventually
> fourier, mathdesign or MnSymbol.) A (German) survey of math fonts is
> available at
> http://milde.users.sourceforge.net/Matheschriften/matheschriften.html
Yes, of course. But some classes (IEEEconf, for example) use
specifically mathptmx, probably because they'll use house fonts
for final printing.
Ciao
Enrico
Guenter Milde
> Guenter Milde <mi...@users.berlios.de> wrote:
>> On 2010-09-25, Joseph Wright wrote:
>> > On 25/09/2010 09:13, Enrico Gregorio wrote:
>> >> The fact that the two commands choose the same symbol with mathptmx
>> >> is easily explained: the font Adobe Symbol lacks an alternative form
>> >> for epsilon and rho. It has one for phi, pi and theta.
>> Is there a warning, at least?
> Why? The two symbols are just graphical variants which should never
> get different meanings.
While they are graphical variants in Greek text, they may well have
different meanings in math. This is the reason for two distinct macros in
TeX as well as for two distinct Unicode slots:
03B5 GREEK SMALL LETTER EPSILON
03F5 GREEK LUNATE EPSILON SYMBOL
= straight epsilon
x (small element of - 220A)
# 03B5 greek small letter epsilon
or (in the mathematical alphanumerical symbols) e.g.
U+1D700 MATHEMATICAL ITALIC SMALL EPSILON
U+1D716 MATHEMATICAL ITALIC EPSILON SYMBOL
Günter
Donald Arseneau
> Why? The two symbols are just graphical variants which should never
> get different meanings.
Here is somewhere that unicode and electronic distribution have
changed the game, and TeX is (as usual) based on paper printing.
In a unicode document you should be able to find out
what a glyph *means* based on its code value, even if its
appearance is the same as many other characters (look at the
character tables and see how may repetitions of form there
are).
Even in traditional TeX/LaTeX there was good reason to have
\mathchardef\Epsilon=`E
to make the source file meaningful (think \sum vs \sigma or
\to vs \rightarrow). I don't think anyone expected you to
type "E" in preference to a mathematical Epsilon, but that
capital Epsilon just isn't used in notation (because it looks
just like E).
--
Donald Arseneau as...@triumf.ca