Subject:
|
Re: Memory - Where did it go?
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Sat, 8 Jul 2006 11:30:11 GMT
|
Viewed:
|
4623 times
|
| |
|
|
In lugnet.robotics, Tom Stiles <hamrad45@hotmail.com> wrote:
> After reading the help file I was informed that out of the
> 256K of memory (which is small to begin with), only about 56K
> is available for user files including sound files.
Once you delete a lot of the "Try me" and on-brick programming stuff, this
jumps up to (IMS) 120k+ or so. Still not ideal (yeah, I'd like a Mb of memory as
well :-), but much better. As to what to delete, I've taken everything off but
the system sound files (and I left those on because I like them, not because
they're required; I've deleted those at times as well).
> not knowing which files I can delete with no problems.
I'm not sure there *are* any files that "can't be deleted". I'll have to
check, but I've deleted everything I think at one time or another.
> 2) Can someone show me one of their more complicated programs and
> how much memory it took?
Yes, but I'm not sure how informative this will be. Somebody asked this on
theNXTstep blog, and so I looked at some sizes. The largest program I've written
(nesting7.rbt) has 39 blocks, 6 loops, & 1 switch, and is about 2.2 Mb on the
comp[uter, and compiles down to 23.5 kb in the NXT. If the NXT has 120 k of free
memory, it would seem you could hold a program of about 240 blocks... but I'm
sure some of those are bigger than others*, so... I suspect (that would make
this an unwarrented, unsupported *guess*) a practical upper limit is around
150-180 blocks, if there's little else on the NXT itself. YMMV.
> to see how effiecent the program is when complied...
I think John Hanson has done a good bit of digging on this. The answer
appears to be "not as good as I'd like", with hand-optimised programs coming in
at (IMS) 2-3 times smaller (still rather large in some cases, but at least
there's some good room for improvement).
*execution time varies a good deal between different blocks, for instance, with
a Math (addition) block taking just 0.14 ms, (and variable read blocks being
even faster) to Draw image blocks taking 7 ms or up; sensor reads vary with the
sensor, but take around 4.4 ms in my testing.
--
Brian Davis
|
|
Message has 1 Reply:
 | | Re: Memory - Where did it go?
|
| (...) By "Block" you mean a simple operation like and add or a multiply or turning on a motor or reading a sensor? These 'blocks' ought to be turned into little interpreted byte-code segments - it's hard to imagine how a block could possibly take up (...) (18 years ago, 8-Jul-06, to lugnet.robotics)
|
Message is in Reply To:
| | Memory - Where did it go?
|
| I got my Mindstorms NXT this past weekend and built the demo unit and then the Claw with only minor programming errors. Tonight I started using some of the various programming functions. My first program was to use the sonar and detect the presence (...) (18 years ago, 6-Jul-06, to lugnet.robotics)
|
9 Messages in This Thread:
  
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
