This is unisym.html
Naturally I was somewhat irritated to find that Mozilla and Netscape _ hereinafter
Netzilla _
have disabled the symbol font. For them _ but not for
Opera _ the fixup is easy
enough on the user side:
- first find the file, fontencoding
- open it with notepad or another vanilla editor
- change
# Symbol font
encoding.symbol.ttf = Adobe-Symbol-Encoding
- to -
# Symbol font
# encoding.symbol.ttf = Adobe-Symbol-Encoding
encoding.symbol.ttf = windows-1252
I definitely recommend keeping their way around for the purpose of comparing the new
way with the old way. In fact, anybody who has the symbolps font can replace the last step with
- change
# Symbol font
encoding.symbol.ttf = Adobe-Symbol-Encoding
- to -
# Symbol font
encoding.symbol.ttf = Adobe-Symbol-Encoding
encoding.symbolps.ttf = windows-1252
That way, direct comparisons of characters from
http://www.mozilla.org/projects/mathml/fonts/encoding/symbol.html
and characters from symbolps can easily be made.
One of the major arguments in favor of disabling the symbol fonts has been
that symbol and symbolps have major alignment defects, and another may be that
some of their glyphs are unsatisfactory. (The l.c. epsilon, for example, is what
LaTeX calls \varepsilon.) Fair enough, but perhaps somewhat overstated -- here is
an
example that contains comparisons
of extended Times New Roman and symbolps characters.
O.K.:
The alignment of new bracket-parts is marginally better.
Some of the new glyphs may be thought to be more pleasing.
What about those \varepsilons?
What about A = πr2
and çA÷ = pr2 ?
(very browser-dependent!)
The codes for the two inline equations are
A = πr<u style=top:-3pt;font-size:10pt>2</u>
and
<s>ç</s>A<s>÷</s> = <s>p</s>r<u style=top:-3pt;font-size:10pt>2</u>
As far as I know, only Netzilla, modified in the second way described above, can render both equations,
and Opera can't render either one.
Enhancements I would like to see Netzilla and Opera adopt are:
- The client's choice between Adobe-encoding and symbol-fonts should
be included in edit preferences ... .
- Browsers should recognize some kind of directive that authors can use to indicate which style
of encoding their documents employ.
If the new style catches on -- it is marginally better, after all -- other purveyors
of browsers might then be urged to adopt similar enhancements.
Meanwhile, I'll continue to play around with both ways. And one more thing: the equations (with
symbol or symbolps enabled, if possible) definitely look better on newer browsers than they do
on NN4. Here is one that only works on
Netzilla.