| |
In lugnet.robotics, Corey Snipes writes:
>
> Since LeJOS doesn't currently implement garbage collection, it seems
> there are 2 very different directions one could go from here. One
> would be to implement garbage collection in some fashion, and the
> other would be to implement a c-esque "free()" so that it could be done
> manually.
>
> I realize the second suggestions probably runs counter to many folks'
> java sensibilities. However, as I understand it the biggest obstacle
> to implementation of Java in real-time (ie, robotics) systems is
> garbage collection and the unpredictability that it adds to a running
> system. Since it's not presently a "problem" in LeJOS, might a free()
> in the .rcx package be useful?
I would not call it running counter to many folks' Java sensibilities,
this suggestion simply violates one of the main Java principles.
I sincerely hope that nobody even remotely considers implementing this.
The only correct way of implementing Java is either with GC or without
freeing memory at all.
That last option is actually not so bad (JavaCard has used it for years)
an traditionally embedded programs simply allocate all memory they need upfront
and then work with that. As long as you know the memory you allocate is never
reclaimed, it is easy to work around it.
--Søren
|
|
Message is in Reply To:
4 Messages in This Thread:
  
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
Active threads in Robotics
|
|
|
|
|
|
