Arabic-L:GEN:iCab

[Dilworth B. Parkinson]

----------------------------------------------------------------------
Arabic-L: Mon 29 Oct 2001
Moderator: Dilworth Parkinson <dilworth_parkinson at byu.edu>
[To post messages to the list, send them to arabic-l at byu.edu]
[To unsubscribe, send message to listserv at byu.edu with first line reading:
           unsubscribe arabic-l                                      ]

-------------------------Directory-------------------------------------

1) Subject: iCab
2) Subject: iCab
3) Subject:OSX and Arabic

-------------------------Messages--------------------------------------
1)
Date:  29 Oct 2001
From: Frantz <erkidego at yahoo.com>
Subject: iCab

I had not problem at all with iCab.  It works better than any other browser
to read Arabic.  You have to be sure that in the text coding you choose the
right Arabic system that was used to create the page you are viewing.
There are three options in the menu.

I have been able to copy and paste from the html document into NisusWriter
without any problem at all.

Frantz

--------------------------------------------------------------------------
2)
Date:  29 Oct 2001
From: Albrecht Hofheinz <Albrecht_Hofheinz at web.de>
Subject: iCab

Thanks YemenL for pointing out the need to *register iCab for Arabic*
using Apple's Language Register program. This does indeed do the
trick. I haven't tried it under OS X yet, but it works fine in 9.2.
Really should have thought of this myself.

For the benefit of the list, here's what Knut Vikør
<fmikv at alf.uib.no> of the University of Bergen, Norway, kindly wrote
in response to the original query:

QUOTE

I was forwarded this to see if there was a cure. In fact, I am not
yet up to OS X, so I may not be the best to answer. But briefly, what
I know:
- Under Mac WorldScript (systems up to 9), all punctuation is
duplicated, there is an Arabic period and a Latin period, an Arabic
comma and a Latin comma etc. Only the Mac does this. In all other
systems (ISO 8859, Unicode, Windows), a distinction is made between
those punctuation that has a different shape in Arabic (is reversed:
comma, question mark, semicolon) and those that are identical from
left and right (period, colon, exclamation mark). The former have a
separate Arabic identity, the latter do not, they have only one
period and require some other code to tell the display whether to use
an Arabic or English font.

The Mac's way of doing things simplifies matters, but that does not
mean much netwise as long as no other platform does this. So, all
Arabic texts on the net follow the "single period" rule; i.e. only
the Latin period exists. Thus when iCab sees a period in an Arabic
text, it - according to the Mac's system - sees it as an English
period, and acts accordingly.

When English and Arabic texts appear on the same line, the
application must decide which language dominates; i.e. whether the
Arabic and English text blocks are to be ordered from left to right
on the line or right to left. Good Arabic programs like Nisus allow
the user to decide this; but iCab is not made with Arabic in mind. It
is hard wired to Latin left-right domination, and thus will always
organize the text blocks left to right; even if the "English block"
is just a single period.

Thus, this is not a bug, but just following from the logic of Mac/net
discrepancy. Once Unicode is implemented, this will go away, since
Mac of course then will follow the same Unicode standards in this
respect. -- I have alerted iCab's developers to this situation, but I
suspect they will not give priority to what may be a major rewrite on
this OS 9-only problem, but rather go for Carbonization.

As for OS X; Arabic is as you know not implemented yet, but I am told
by knowledgeable people (not having upgraded myself), that if you
install a full Unicode font that contains Arabic characters, while
you cannot yourself yet write Arabic, you should be able to read
Unicode Arabic pages on the Web using the OmniWeb browser for OS X. I
copy from his message:

>As for right to left languages, OmniWeb can display web pages in
>UTF-8 correctly if the text direction is set to R to L and direction
>marks are inserted properly.
>
Yusuke Kinoshita

I cannot thus confirm this, but it may indicate that things may be
looking up. (I have also heard that Netscape 6.sth has some support
for Arabic, but as far as I understand, it is not properly working
yet. However, it should perhaps be monitored in upgrades).

Knut S. Vikør

END OF QUOTE

--------------------------------------------------------------------------
3)
Date:  29 Oct 2001
From: dparvaz at unm.edu
Subject: OSX and Arabic

I went ahead and tried OmniWeb's "sneakypeek" release of their
browser (bad Javascript and DOM implementations have kept me from
them in the past), but all I get in Arabic pages (say, aljazeera.net)
is an Arabic keycaps-style glyph which screams "not yet implemented."
When I install an Arabic font (like Geeza), I get Arabic characters,
but they're all stand-alone (i.e., no "joining"), and everything is
still left-to-right.

The same problem exists in OS X's stock mail application, which would
indicate that both programs are relying on nonexistent OS-level
support.

*Sigh*,

Dan.

--------------------------------------------------------------------------
End of Arabic-L:  29 Oct 2001