🚀 go-pugleaf

I have experimented with installing multiple script systems on the same
machine at the same time. In my case it was a graft of the Kanji
J1-6.0.2 and the Arabic AB1-6.0.3 onto an Australasian Z1-6.0.3. I
created my hybrid system with ResEdit, copying the appropriate itl0,
itl1, itl2, itl4, itlb, KCHR and SICN resources (I can't remember if
this is an exhaustive list), and the appropriate fonts, into my System
file, and then I copied the appropriate INIT and cdev and other supporting
files into my system folder.

One other thing I had to do was change the ID of the FOND resource for
Chicago from 0 to 16383. The first word of the FOND needs to be changed to
the same number as well.

I didn't need to write any FKEYs or any other code to switch between
script systems. As soon as I restarted my machine with two or more scripts
installed, the appropriate SICN (small icon) would appear in the menu bar
to indicate the current keyboard script. I could rotate scripts by
clicking on this icon, or by pressing command-space.

Also, script-manager-friendly programs are supposed to check for a change
of script when you change fonts in a document, and update the keyboard
script accordingly.

All this *sort* of worked. I had particular trouble with the Kanji
software; every once in a while the Kanji input system would get into
Kana-only input mode and refuse to come out. Of course, the fact that I
had no documentation for any of this didn't help.

I seemed to find an incompatibility between MacroMaker and the Kanji
script system. Specifically, I was getting intermittent crashes about
once every couple of days. Removing MacroMaker seemed to make these
crashes less frequent. But I never managed to pin down the exact
cause. There could very well have been a problem with the mismatch
in versions between the Kanji and the other two systems.

The main hassles were with all this American-made software, that
behaved very strangely with non-Roman writing systems. WriteNow 2.0
(and some other applications) put large gaps after every sequence of
Kanji text, thinking that the double-byte characters were taking up
about twice the width that they actually were. MacWrite 5.0 (this was
before MacWrite II) seemed to be the best-behaved, though of course
its word wrap wasn't very effective with no spaces between Kanji words!
Also, it was quite a trick to avoid getting the insertion point between
the two halves of a double-byte character.

Funnily enough, mixing right-to-left (Arabic) and left-to-right
text seemed to work quite well. A more subtle problem with double-byte
(Kanji) characters is that it is possible for searches to match *parts* of
characters.

I wonder how much performance it would cost to make a word processor
Script-Manager-compatible as a matter of course. Any developers of
text-intensive software care to comment?

Lawrence D'Oliveiro
They-pay-me-to-mess-about Consultant
Computer Services Dept, University of Waikato
Hamilton, New Zealand

DISCLAIMER: all the above information was gleaned from non-intensive
experimentation over a period of a few weeks. I don't know any Japanese
or Arabic, though I did get some help from a lecturer in our Japanese
department (I don't know of anybody on campus who speaks Arabic). In
other words, try this at home at your own risk!