[seqfan] Re: MathJax (Charles Greathouse)

[Brendan McKay]

I'm a great fan of MathJax and highly recommend that OEIS considers it.
Here are some random remarks prompted by the questions and comments.

1. One of the strengths of MathJax is that it renders the mathematics in
type of the same size as the surrounding text.  If you see oversized or
undersized mathematics on web pages, you can be pretty sure it isn't
MathJax.  To see it in action, try these pages:
http://www.combinatorics.org/ojs/index.php/eljc/article/view/v21i2p17
http://www.combinatorics.org/ojs/index.php/eljc/article/view/v20i4p28
http://mathoverflow.net/questions/172559/on-the-superior-of-generalized-ornstein-uhlenbeck-process
The first time you visit a MathJax page it might take a few seconds extra,
but after that the scripts will be in your browser cache and you won't
notice any delay.

2. Installing MathJax is trivial: the web master adds about 8 lines to 
the invisible
header that is included already with each web page.  It will take 15 
minutes.
Nothing needs to be done separately for each page and readers don't need to
do anything at all.

3. Once those magic 8 lines are there, their effect on the appearance of 
OEIS
pages will be none at all.  Everything will look exactly the same as it 
always
did until someone deliberately marks a formula as mathematics by putting
a dollar sign on each side of it (or two dollar signs on each side to 
display
an equation on a line by itself).  MathJax does not interpret TeX commands
that are not in mathematical formulas (so the overall layout of the page
is not touched).

4. Searching for formulas is a big problem for mathematics on the web,
but since the format of plain-text formulas is not consistent either, using
MathJax would only change the situation from "can't do it" to "can't do 
it".
Simple things won't be different ("$x+y/2$" versus "x+y/2") and complex
things like "$f(x) = \frac{\sin(x)}{x+1}$" versus "f(x)=sin(x)/(x+1)" can't
be searched for now with any useful success rate anyway; the first one
will look much better though.

5. There are a small number of things that don't work, but generally
speaking it is pretty robust and any formula in normal TeX syntax will be
displayed correctly.

Cheers, Brendan.