RE: Exception in SWT + serial port application

[Øyvind Harboe <oyvind dot harboe at zylin dot com>]

> Øyvind> Q: If I'm on the right track, is there a solution to 
> this issue? 
> Øyvind> My GCJ superpowers are not quite yet up to the task of 
> Øyvind> implementing _Jv_ThreadInterrupt.
> 
> I'm afraid you'll have to find someone else to implement the 
> missing functionality.  I don't know enough about Win32 to do it.
> 
> There are a few things like this missing from the Win32 port; 
> that port is still a bit immature.

After some testing I've concluded that the only user-noticeable 
problem with _Jv_ThreadInterrupt for my application is that it 
puts up a message box. If it printed out a nasty warning to 
System.err instead of putting up a message box, I would be done.


This is not a train wreck for me, I have a couple of solutions
available:

- patch the GCJ binary 
- compile a special version of GCJ which does nothing in 
_Jv_ThreadInterrupt. This requires me to drink that mail-order
radioactive potion to get mutant GCJ superpowers :-)
- Somehow modify(patch?) the www.serialio.com library to not invoke
Thread.interrupt(). The source is available in a more expensive
version of their product. 
- There are some reverse engineering tools for my www.serialio.com
component, but the one that I tried from sourceforge produced
readable, but incorrect code. A different(or newer version) reverse
engineering tool, might be up to the task.
- I've had no luck getting javax.comm from Sun to work so far,
but that might be an avenue to pursue.


Most likely not invoking Thread.interrupt() has some side neagtive
effect, such as leaking resources. This is utterly unimportant 
for my application at this point as long as the user does not 
notice. My application is typically started, run for a minute or two, 
then quit, hence it is not the sort of application where these 
things are terribly important.


Øyvind