Thanks for all the responses.
Philipp Lohmann wrote:
> However wouldn't it be better if MacOSX supported the
Devanagari
> and Tibetan scripts? This way the whole platform would
benefit
> instead of only X11 applications.
Well, Mac OS X *does* support Devanagari and Tibetan; it
just doesn't
do it the same way Windows does. I gather Windows users must
wait for
Microsoft to "support" their script in the OS, but
in the Mac
environment all necessary resources can be built into the
font; thus
font developers have always been able to add
languages/scripts
without asking permission from Apple, and there were
Devanagari and
Tibetan solutions (among others) in the 80s -- before
Windows
existed, I believe.
However, among a number of unfortunate decisions made by the
series
of not-very-bright managers at Apple during the early 90s
was failure
to pursue the Mac's early multilingual advantage, leaving
the non-
Western world to standardize on Windows, where extra-ASCII
characters
and non-Latin scripts were handled differently -- resulting,
among
other things, in the common problem of characters like
dashes and
curly quotes on legacy-encoded (non-Unicode) Web pages not
rendering
properly.
In recent years, Unicode has solved most of the problems in
cross-
platform multilingual computing, but for some scripts it's
not
enough. As Tom Gewecke's "Multilingual Mac" page
explains:
"Note, however, that viewing complex scripts which
require
reordering, contextual shaping, or stacking of characters
(such as
Arabic, Devanagari, Tibetan, Classic Mongolian, and Thai)
requires a
combination of font and rendering engine technology. On the
Mac this
is accomplished via an AAT (Apple Advanced Typography) font
and
ATSUI, while Windows uses an OpenType font plus Uniscribe.
The result
is that when you select a Windows font in OS X, complex
scripts are
unlikely to display correctly, and an Apple font should be
used if
available."
<http://homepage.mac.com/thgewecke/mlingos9.html#font>
Thus, while work can be done in any of these scripts in
either
Windows or Mac OS, the results can't be exchanged reliably,
as the
complex scripts will tend to decompose into their elements
(for
instance, the short-i character in Devanagari, which should
appear to
the *left* of the consonant it follows in speech, will move
to the
right, the order in which it was entered) when viewed on the
other
platform. (Since Linux uses the same system/fonts as
Windows, the
same is true for exchanging between Mac and Linux.)
I'm not clear on all the differences between AAT/ATSUI and
Uniscribe/
OpenType, but from what I understand the Apple system seems
to be
superior, since it empowers users/developers rather than
retaining
control in the hands of the OS publisher (i.e. Microsoft).
(Also,
according to the major provider of such minority script
systems for
the Mac, XenoType Technology <http://xenotypetech.com&g
t;, the Apple
system is superior to Windows in other respects as well,
though I
don't know the details.) However, once again, as in the case
of file
name extensions (where Apple has abandoned its superior
metadata
system in order to "conform" with the primitive
system used in
Windows), the Windows system is what nearly everyone uses,
so I
suspect Apple may eventually capitulate and switch to the
Adobe/
Microsoft/OpenType system as the primary complex-font
technology on
the Mac. (Just my speculation, no inside knowledge.)
In fact, Apple has already taken steps to embrace some
OpenType
technology, to allow Mac users access to features like swash
letters
in Windows OpenType fonts -- which unfortunately had the
effect of
breaking the only successful effort at a truly
cross-platform Tibetan
font, created by the developer of the only Unicode Tibetan
font/entry
system that presently works in the Mac environment:
<http://xe
notypetech.com/osxTibetan.html>
For a while this Kit contained both AAT and Windows OpenType
resources (so far as I know, this was the only developer who
figured
out how to do this), but the latter have had to be removed
since OS X
10.4. The developer wrote to me: "If a font contains
Windows-styled
OpenType tables (necessary for foreign fonts to work
properly under
Windows) AND native Mac tables, the Mac tables are disabled
and
ignored with preference given to the OpenType tables."
So for now, anyone who wishes to exchange files in
Devanagari,
Tibetan or similar languages/scripts (a small number, true,
but such
users do exist) must seek other methods -- and
X11/OpenOffice does
seem to be at least an interim solution. Thus my query.
Maho NAKATA wrote:
> Okay, I'll provide X11 version as far as possible.
Well, thanks. I'm not asking anyone to go to any special
trouble; I
was just wondering what the plan was. I don't really know if
anyone
*is* actually using X11/OpenOffice as a solution for
cross-platform
Devanagari or Tibetan on the Mac. For me all this stuff is
just a
hobby, combining my interests in Buddhism, Asian cultures
and
languages, linguistics and graphic arts. Tom Gewecke, the
Mac world's
resident multilingual computing expert, suggests X11/OO, so
I tried
it out, and found that indeed it does work -- it even
appears to work
(to my somewhat surprise) with Apple's supplied input method
(keyboard script) for Devanagari, along with the OpenType
Sanskrit
2003 font, to produce the correct complex glyphs.
If I then open such an X11/OpenOffice document in
NeoOffice/Aqua, the
complex glyphs are deconstructed -- as they would be, I
assume, in an
Aqua OO. Meanwhile, I've also discovered that NeoOffice does
not
support "glyph variants" -- which means that some
alternative
Devanagari characters, which can be used in Apple apps like
TextEdit
or Pages, cannot be used (not even entered from the
Character
Palette) in NeoOffice. (Some Latin fonts, such as Apple's
Zapfino,
also provide glyph variants -- alternative versions of Latin
letters
-- which can't be used in NeoOffice.) I don't know if this
results
from Neo's reliance on Java, or is an OO limitation. When a
usable
Aqua OO appears, I'll try it out. Eventually, of course, I
hope the
Aqua OO (and other open-source apps like Inkscape and
Scribus) will
provide the same typographical capabilities as other OS X
apps.
If anyone (especially OO/Aqua developers) is interested in
the Mac's
multilingual capabilities, Tom Gewecke's site
<h
ttp://homepage.mac.com/thgewecke/mlingos9.html>
is a goldmine of information. Also his blog
<http://m10
lmac.blogspot.com/index.html>
has in-depth information about specific problem
languages/scripts,
including Indic and Tibetan. (Tom is a retired foreign
service
officer who also spends a lot of time on the Apple
discussion boards
answering users' questions.)
Jayson Kempinger wrote:
> Although it's not as easy of a solution as the OOo
provided binary,
> fink (http://finkproject.org/)
provides builds of openoffice in
> several languages.
Thanks for the tip; I've read about Fink, but never done
anything
with it. Being a Mac user (since 1988), I've had little
occasion to
get into command-line or similar under-the-hood operations.
I'll look
into it sometime.
Andrew Main
------------------------------------------------------------
---------
To unsubscribe, e-mail: mac-unsubscribe porting.openoffice.org
For additional commands, e-mail: mac-helpporting.openoffice.org
|
