[tug-summer-of-code] A couple of project proposals

[Arthur Reutenauer]

	Hello Scott,

>                                                                     I
> guess I had assumed that XeTeX was implemented as a set of TeX change
> files

  I understand.  That's really not the case :-)

  In fact, LuaTeX also represents a big change to TeX, so much so that
for the moment it doesn't make sense to try and merge it with XeTeX;
which of course doesn't mean it won't happen at some point in the
future.

>                            integrating XeTeX's font support at the
> macro-package (viz., fontspec) level, presumably using Lua to
> interface with external font libraries.

  In fact, that's not exactly what we meant: OpenType support can really
be implemented in Lua.  That's what ConTeXt does (in the font-ot?.lua
files of its base code; you can browse it online on the ConTeXt garden:
http://source.contextgarden.net/tex/context/base/).  Using external font
libraries would be what I called the "XeTeX way", and would of course be
also possible.  I have to look into that.

>                                          I'm sure that being able to
> write things like \setromanfont{MyObscureSystemFont} in a LaTeX
> document would be a great boon to LaTeX users.

  There's no doubt.  I have been contemplating the idea since years, but
nothing happened for a long time, and in the end this lack of interest
for LuaTeX in the LaTeX community made me abandon LaTeX for ConTeXt.
But recently things have started moving; see, for example the
discussions from February in the archives of the LuaTeX user list,
http://tug.org/pipermail/luatex/).

  About your second idea now:

> In addition to JavaScript and Java, Flash/ActionScript is another
> likely candidate.
  
  Yes, Flash also came to my mind, but I was hesitant to mention it,
because I didn't know if there were good open source tools available for
producing SWF files, and I wouldn't like to propose a project where the
requirements include "must have Flash Professional or Flex".  Would you
happen to know that?  I saw a book on the subject the other day, but
didn't match what I read with what was currently available.

> Also, it's probably not unreasonable to use a server-side C/C++
> program that does the actual character recognition.  It's probably a
> lot easier to find neural-network or OCR code written in C/C++ than in
> other languages.

  Absolutely.

  Would you like to write a proposal for this second project?  We will also
need a mentor in the end, but for the moment we could start by
mentioning the idea and listing the different possibilites, as Karl
said.

	Arthur