| |
In lugnet.robotics.rcx.pbforth, Steven J. Morris writes:
> Ralph Hempel wrote:
> I believe that. 32k of RAM is a luxury for forth. That's another thing
> people don't generally understand about forth, they way it effectively
> compresses runtime software. I've heard of ram reductions of a
> factor of 5-10 for forth when compared to equivalent C code.
Pfff. As if that could possibly be important. If it isn't an object-oriented
client-server architecture in Java++, it can't be serious :-)
> Speaking of which how does one tell how much RAM is available?
> Is there more than one memory pool, i.e. one for ALLOT and another
> for dictionaries?
If I read the source correctly, memory is allocated sequentially, and in my
image I've got SOURCE DROP 258 - HERE -
U.
14502 ok bytes free (258 is the PAD size, which lies below the TIB).
> I'm beginning to think that some of my communication problems might
> be hardware. In an earlier post I reported long delays. I have been
> keeping track and it doesn't make much sense. I had a 4 second delay
> after the HEX command. By delay I mean that I type HEX <CR>
> and repeat <CR> roughly every second until I see "ok". I assume that
> there is absolutely no excuse for a delay i processing HEX which
> probably just sets a variable. On the otherhand maybe the delay
> is in Linux somewhere. Doesn't seem likely with all the RAM I've
> got. No swapping anyway.
Weird. How are you connecting? I'm using Minicom, port settings 2400 8O1,
neither hardware nor software handshaking.
Ernst
|
|
Message is in Reply To:
 | | How much RAM is left?
|
| (...) I believe that. 32k of RAM is a luxury for forth. That's another thing people don't generally understand about forth, they way it effectively compresses runtime software. I've heard of ram reductions of a factor of 5-10 for forth when compared (...) (25 years ago, 5-May-00, to lugnet.robotics.rcx.pbforth)
|
7 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
