\textasteriskcentered not actually centered

[Lev Bishop]

The \textasteriskcentered is not actually centered. This is a problem
because when you use it as a footnote symbol it gets superscripted
again, ending up with a very small footnote symbol. This happens for
example if you \usepackage{fixltx2e}.

Also, on the topic of asterisks, is there any chance to get
six-pointed asterisks added to MnSymbol? Maybe with a package option
to choose those instead. I personally find the 5-pointed version quite
ugly (it's the only symbol I've used from MinionPro / MnSymbol that I
really wanted to replace -- everything else is quite tasteful). I have
a problem with 5-pointed asterisks in general (they seem too
lightweight somehow) and the Minion ones are especially ugly (to my
mind) because they're not even symmetrical, being slightly rotated to
the right. This applies to the MinionPro asterisk as well as the
MnSymbol $*$, $\oast$, $\bigoast$ etc.

Thanks,

Lev Bishop

[Achim Blumensath]

Hello,

Post by Lev Bishop
The \textasteriskcentered is not actually centered. This is a problem
because when you use it as a footnote symbol it gets superscripted
again, ending up with a very small footnote symbol. This happens for
example if you \usepackage{fixltx2e}.

For a quick fix, you could define

\renewcommand\textasteriskcentered{$*$}

Post by Lev Bishop
Also, on the topic of asterisks, is there any chance to get
six-pointed asterisks added to MnSymbol?

Sorry, no. The five-pointed version was the choice of the designer of
MinionPro. As far as I remember, this is typical for renaissance fonts.

Achim

[Lev Bishop]

Post by Achim Blumensath
Hello,

Post by Lev Bishop
The \textasteriskcentered is not actually centered. This is a problem
because when you use it as a footnote symbol it gets superscripted
again, ending up with a very small footnote symbol. This happens for
example if you \usepackage{fixltx2e}.

For a quick fix, you could define
 \renewcommand\textasteriskcentered{$*$}

Thanks. That works.

Post by Achim Blumensath

Post by Lev Bishop
Also, on the topic of asterisks, is there any chance to get
six-pointed asterisks added to MnSymbol?

Sorry, no. The five-pointed version was the choice of the designer of
MinionPro. As far as I remember, this is typical for renaissance fonts.

I must say that I expected this response :-) But allow me one attempt
to persuade you to reconsider...

I can understand the choice of 5-pointed asterisk for marking
footnotes, but I really think it seems wrong for the mathematical
asterisks to be anything other than six-pointed. I see that at least
one typographer agrees with me on this point:
http://www.tug.org/TUGboat/Articles/tb27-0/kuster.pdf . "...most fonts
already contain an asterisk * (U+002A), but normally the latter is not
suitable for math, which uses a larger version, six-pointed..."

And I see that all the fonts I have that contain the Unicode
mathematical asterisk (U+2217) have it with a 6-pointed form. There
are some fonts which use 5-pointed regular asterisks but six-pointed
mathematical asterisks (for example the fonts from the Lucida family).

Please consider to use the same model as is done with Lucida: keep the
original Slimbach text asterisk but for the mathematical asterisks use
a six-pointed form.

Well, if I failed to change your mind, it's ok. But I thought I should try.

Lev Bishop

[Achim Blumensath]

Hello,

Post by Lev Bishop

Post by Achim Blumensath

Post by Lev Bishop
Also, on the topic of asterisks, is there any chance to get
six-pointed asterisks added to MnSymbol?

Sorry, no. The five-pointed version was the choice of the designer of
MinionPro. As far as I remember, this is typical for renaissance fonts.

I must say that I expected this response :-) But allow me one attempt
to persuade you to reconsider...
I can understand the choice of 5-pointed asterisk for marking
footnotes, but I really think it seems wrong for the mathematical
asterisks to be anything other than six-pointed.

It would help if you could give any reasons supporting this point of
view, instead of just saying that "it seems wrong".

Post by Lev Bishop
http://www.tug.org/TUGboat/Articles/tb27-0/kuster.pdf . "...most fonts
already contain an asterisk * (U+002A), but normally the latter is not
suitable for math, which uses a larger version, six-pointed..."

The way I read that passage, the reason for the unsuitability of the
supplied asterisk is that it is not centred. I don't think the number of
points is that important.

Post by Lev Bishop
Please consider to use the same model as is done with Lucida: keep the
original Slimbach text asterisk but for the mathematical asterisks use
a six-pointed form.

That leads to inconsistency.

Achim

[Lev Bishop]

Post by Achim Blumensath

Post by Lev Bishop
I can understand the choice of 5-pointed asterisk for marking
footnotes, but I really think it seems wrong for the mathematical
asterisks to be anything other than six-pointed.

It would help if you could give any reasons supporting this point of
view, instead of just saying that "it seems wrong".

I understand that you would prefer something more concrete than my gut
feeling, but this is all I have :-)

I suppose the reason is that the five-pointed version seems very
unfamiliar. I don't recall ever seeing a five-pointed mathematical
asterisk in any books or papers I have read. I am not any kind of a
typographer, so I don't have any reference books to look up all
different fonts, but I just checked the mathematical books on my shelf
and they all use six-pointed mathematical asterisks. As I said, even
Lucida uses six-points for mathematics when it uses five-points for
text. And I just found googled for some other mathematical fonts and I
found Bitstream Charter uses 5-points for text but Micropress CH Math
font (which is a mathematical extension of Bitstream Charter) also
uses six points:
http://www.micropress-inc.com/fonts/chmath/chf2.htm
The same thing is true with Micropress HV Math:
http://www.micropress-inc.com/fonts/hvmath/hvmain.htm
And also Micropress PA Math:
http://www.micropress-inc.com/fonts/pamath/pamain.htm

Post by Achim Blumensath

Post by Lev Bishop
http://www.tug.org/TUGboat/Articles/tb27-0/kuster.pdf . "...most fonts
already contain an asterisk * (U+002A), but normally the latter is not
suitable for math, which uses a larger version, six-pointed..."

The way I read that passage, the reason for the unsuitability of the
supplied asterisk is that it is not centred. I don't think the number of
points is that important.

Well, in I found a talk by him which you can listen to here:
http://river-valley.tv/minion-math-a-new-math-font-family/
(it seems he is designing some kind of a competitor to MnSymbol) he is
quite specific (around the 21:30 mark). He says that mathematical
asterisk should always be six-pointed with one vertical stroke and two
cross strokes. His explanation for this is because mathematicians want
the asterisk to look like the ones they draw by hand, and five-point
asterisks are difficult by hand. (I find this a fairly weak argument,
I admit, because many mathematical symbols are not so easy to draw by
hand, for example fractur letters).

Post by Achim Blumensath

Post by Lev Bishop
Please consider to use the same model as is done with Lucida: keep the
original Slimbach text asterisk but for the mathematical asterisks use
a six-pointed form.

That leads to inconsistency.

Well, one way is inconsistent between the text and the mathematics,
but the other way is (as far as I can tell) inconsistent with general
mathematical typography.

At least four fonts that I could find in a few minutes of searching
the web (lucida, charter/ch math, hv math, pa math) preferred the
former kind of inconsistency, and none that I could find preferred the
latter.

Lev

[Achim Blumensath]

Hello,

Post by Lev Bishop

Post by Achim Blumensath

Post by Lev Bishop
I can understand the choice of 5-pointed asterisk for marking
footnotes, but I really think it seems wrong for the mathematical
asterisks to be anything other than six-pointed.

[...]

Post by Lev Bishop
I suppose the reason is that the five-pointed version seems very
unfamiliar. I don't recall ever seeing a five-pointed mathematical
asterisk in any books or papers I have read.

That's just because most mathematical texts are set in a baroque or
romantic typeface like Times or Modern, which have asterisks with six
tear-drop shaped arms.

Post by Lev Bishop
He says that mathematical asterisk should always be six-pointed with
one vertical stroke and two cross strokes. His explanation for this is
because mathematicians want the asterisk to look like the ones they
draw by hand, and five-point asterisks are difficult by hand.

I think the reason is rather the one you stated above: most people are
more used to six-pointed asterisks.

I do not accept this argument. In the same way you could argue that all
mathematical texts should be set in Computer Modern: that's what most
mathematicians are used to.

It would be something different if the shape were that unfamiliar to
readers that they could hardly read the symbol. But that's not the case
here: most people would instantly agree that it's an asterisk, although
one that looks a bit funny.

Post by Lev Bishop

Post by Achim Blumensath
That leads to inconsistency.

Well, one way is inconsistent between the text and the mathematics,
but the other way is (as far as I can tell) inconsistent with general
mathematical typography.

It's not inconsistent, just different.

Achim

[Lev Bishop]

Post by Achim Blumensath

Post by Lev Bishop
I don't recall ever seeing a five-pointed mathematical
asterisk in any books or papers I have read.

That's just because most mathematical texts are set in a baroque or
romantic typeface like Times or Modern, which have asterisks with six
tear-drop shaped arms.

I'm sure you are right that most mathematical texts are set in baroque
or modern typefaces. But it's more than that: even those texts that
are set in old-style typefaces still use 6-armed asterisks. For
example, the journals Nature and Nature Physics are currently set in
Minion, but they use 6-armed math asterisks (I can send you a reprint
if you'd like to see). Paul Hudak's book "The Haskell school of
expression" is set in Lucida Bright, but uses 6-armed math asterisks.
I'm sure there are other examples. These are just what I found now.

One final argument for using 6-pointed $*$: it makes the five-pointed
$\star$ much more easily distinguished from $*$.

Lev