Parser dialects and unifying code flavors
iva2k
original ASCIIMath can parse a subset of LaTeX and David's version can
parse TeX, which are close but not identical. Each has a different set
of useful features besides the markup dialect. Looking at the
history, different versions keep diverging further away from one
another. Question 1 - is there a comparison table between branches &
versions?
Question 2 - I wonder if there was a discussion on various dialects
and potential of bringing them together? I vaguely remember some
discussions I saw some time ago but cannot find any pointers to any
conclusions.
If there's a way to combine everyone's efforts, there can be benefit
to all. For one, that will result in having a common code - browser
glue layer - maintained in one place (sourceforge?) so it can have the
best coverage of all clients (browser hooks, browser variants support,
image fallback). Different language dialects can be maintained as add-
ons that are independent of browser glue layer and potentially they
could be combined as needed depending on intended use.
For myself I highlight certain features that I would like to combine:
Core:
- widest browser support
- image fallback
- graphs created from ascii math
Language Add-ons:
- ASCII math
- LaTeX
- TeX
Am I dreaming?
Marc Grober
Hash: SHA1
MA has done the most on this, with his site at http://www.asciimathml.com/
I had discussed the idea of breaking things up again into small
libraries called by a pre-processor of sorts with David and Mauno a
year or so ago, but no one has jumped on this, and with the recent
developments with MathJax and jsxgraph there is "competition" for
users leaving the initial core functionality (a simpler way to write
math) as key, and even that may be soon addressed via jax thanks to
David.
I think it is still a great set of tools, and MA has done some great
work extending it.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG/MacGPG2 v2.0.16 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iQEcBAEBAgAGBQJOX6b3AAoJEMCi9wxDaNqERgQH/1eA9iuSiRziUmEvR56Vf2J4
1+IEkQY5bo5Dy8dbPErdXaJwmDGuIfskla9XhllVI5xWMnP+mbhPg61j5syjD+2z
NEex8hbRAj26tFuwWOQJ1QmN7ZD1oLkVsCX2txx2+qe5YC+R+2pQGIetKzuEZYZv
LtjDB2/eBaKQ/Ds0qHhA675XXPvu830OLu0/eW6axNXUvI0cn+m69OgxmZcpZ/Mw
W2RTvXN+8AAadXDzB+Nv0dlbOPBLAbq6XvExmufxjG2Ixyd+3OpFEYlsViXxePY0
ZMqQhh+VOeEcCATwpRNbvd6HAGArzBEafn6aUGC5PBUBugmLSA5y7VQfEqZp1fU=
=3j4L
-----END PGP SIGNATURE-----
David Lippman
what I do know.
The TinyMCE plugins and image-based fallback scripts I try to maintain
(and use in the TinyMCE plugins) are based on AsciiMathML version
1.4.7, which could parse the AsciiMath language and supported a small
subset of TeX/LaTeX symbols.
At some point, Peter released AsciiMathML 2.0, which merged the
original AsciiMathML, the LatexMathML derivative work, and the
AsciiSVG graph drawing scripts. I personally had no interest in the
Latex part, and didn't see the value in merging AsciiMathML with
AsciiSVG, which is why I continue to use the original un-merged
scripts. It is worth noting that the AsciiMath language itself is
identical between both versions - the only difference is the
LatexMathML extensions and possibly the symbol set.
Q2: Like Marc mentioned, I think that MathJax is the correct venue
for an add-on oriented approach to math that supports different
dialects and different display modes. As he mentioned, I started
working on building an input jax for the AsciiMath language (which
already works, just isn't cleaned up yet). I doubt I will actually
use it much, though, since AsciiMathML continues to work fine for my
applications.
Personally, I am only interested in working on maintaining AsciiMathML
as a tool for displaying the AsciiMath language. I will also continue
to extend and maintain AsciiSVG for creating graphs, but for my
purposes, I still prefer to keep them separate, rather than trying to
merge them like Peter did.
-David