>>>>> "Jean-Marc" == Jean-Marc Lienher <oksid _at_ bluewin.ch> writes:
Jean-Marc> Hello, XD640 is a new GPL desktop environment for XFree
Jean-Marc> 3.3.X and later.
Interesting. Why spawn a new project?
Jean-Marc> XD640's goal is to provide a set of small graphical
Jean-Marc> applications. Our target machine is a i486 66MHz with
Jean-Marc> 16MB of main memory.
What do you want? I used to work on a system with such hardware. I
was happy and productive with Emacs (my UI to file management, RCS,
gcc, gdb, LaTeX, mail, news) + a few xterm + FVWM.
Jean-Marc> When possible we will be compatible with recent M$
Jean-Marc> Office products. (Since they use XML file format, it
Jean-Marc> will be easier).
XML isn't the magic panacea. If the don't *FIX* the document
structure by giving a definitive DTD or schema, you'd still have to
change a target that they can move at their own will and pace. Even
with a definitive DTD or schema, if the contents are not structured in
a "reasonable" way, you can't do much. e.g.
<?xml version="1.0">
<msword:document xmlns:msword="http://blah.blah.blah/...";>
<msword:binary-data encoding="base64">
a0sd0e0f9DSFs0dvs0df90s8adsv9adsSDFJSDVsoidjsOJNEW
...
</msword:binary-data>
</msword:document>
What can you do? Is XML making it easier? It depends! And it
depends on so many things. Blind worshipping of XML must be avoided.
Jean-Marc> Since XD640 uses UTF-8 strings it is fully
Jean-Marc> internationalized. (Right-to_left strings are only
Jean-Marc> supported in XHTML text widgets but not in other
Jean-Marc> widgets).
Why not move a level of abstract higher and say "Unicode" instead of
UTF-8? UTF-8 should be used only as an external format (e.g. file
transfer). Internally, the system ought to use unicode characters
(wide characters). UTF-8, again, is not the panacea. While it solves
many problems and may provide a better transition to real 16-bit
characters (hey, Unicode has moved faster: it is now 32-bit!), it is
not problemless. E.g., when a Chinese text is encoded in UTF-8, it
needs 3 bytes per characters. Encoding it in UTF-16 would take only 2
bytes per characters -- the same as the national encodings for
Chinese.
Jean-Marc> - file manager (FLFM http://www.oksid.ch/flfm/ ) -
Jean-Marc> XHTML web browser - mail client - news client - plain
Jean-Marc> text editor - word processor (Flwriter
Jean-Marc> http://www.oksid.ch/flwriter/ ) - spread sheet editor -
Jean-Marc> picture editor - ppp dialer
Maybe, you can start with writing a compiler that would read various
Emacs packages (w3, Gnus, rmail, ...) and run them. :D
--
Lee Sau Dan $B'u&u40(B(Big5) ~{@nJX6X~}(HZ)
E-mail: danlee _at_ informatik.uni-freiburg.de
Home page: http://www.informatik.uni-freiburg.de/~danlee
##########################################################################
# Send submissions for comp.os.linux.announce to: cola _at_ stump.algebra.com #
# PLEASE remember a short description of the software and the LOCATION. #
# This group is archived at http://stump.algebra.com/~cola/ #
##########################################################################
