[sword-devel] Unicode Bible program

Joel Mawhorter sword-devel@crosswire.org
Wed, 1 Mar 2000 18:22:22 -0800


On Fri, 25 Feb 2000, Paul Gear wrote:
> Joel Mawhorter wrote:
> > 
> > ...
> > >From what I got out of cvs, it looks like Troy was porting the actual Sword
> > library to Java not creating bindings (Is that right Troy?). I think the best
> > option would be to use the stuff that Troy has ported to Java and create a
> > completely Java version of Sword (library and frontend). This option has a few
> > advantages:
> > ...
> > 2. The current Sword development could continue without getting side-tracked by
> > the changes that would be required for Unicode support.
> > ...
> 
> Let me say it another way: Unicode is not a side-track; it is a key
> development area.  It will happen in the core Sword libraries as soon as
> there is enough demand on the user side and time on the programmer side.
> 
> IMO, it's better not to duplicate effort, but don't let me tell you what
> to do.  I would just suggest that if at all possible that you
> collaborate with others with the same goals.

I agree with you here that it is better not to duplicate efforts. I think
though that it might be better to put the effort into a Java version of Sword
than to modify the C++ version. What I'm going to do is fool around with Java
and see what is required to implement the Unicode features. Even if I end up
working on the C++ Sword libs, I think this is the fastest way to figure out
what the issues involved are. I've already learned a lot just from seeing what
can be done in Java. By the way, the more I see the more I like it. Java has
almost everything needed to support Unicode already built in. 

> I have been convicted by the Lord recently that most of my reasons for
> doing things with Bible software come from my pride and stubborness, not
> his leading, and that sometimes my technical perfectionism needs to take
> a back seat to his kingdom's goals, and compromises need to be made.

This is a valuable warning and one that I'll make sure I keep in mind. My
personality is such that I have to worry about stubbornness and pride a lot more
than other things. I appreciate you sharing your personal experience in this
area.

> > I've been doing some playing around with Java to see if this is feasible and
> > easy. So far it looks like it would be very easy to support this kind of program
> > in Java. The only possible issue is speed. I'm going to play with it a bit more
> > to see if Java is fast enough. The Swing compontents I've been playing with
> > seem fairly responsive on my fairly slow Pentium 166. With well designed text
> > searching/indexing algorithms, I don't think that speed would be a problem.
> > I'll let you know what I discover when I've had a bit more time to experiment.
> 
> It was certainly a snail on my 586/133.  :-)  Another option between the
> two you mentioned would be to compile Java to machine code on specific
> platforms that supported it, or perhaps use a mixed interpreted/compiled
> setup.  I've read that this is possible now on gcj, although i don't
> know how well it works.

You should try some of the new Swing stuff. I downloaded the Java 1.2 SDK for
linux and some of the demos included with that were quite nice and ran at a
reasonable speed.

Joel

> Paul
> ---------
> "He must become greater; i must become less." - John 3:30
> http://www.bigfoot.com/~paulgear