2 Jan
2007
2 Jan
'07
5:54 a.m.
After a bit of a ponder this should be about as fast as you can get
without putting multiple statements on one line ..
1 GOTO 5
2 WAIT A,D,D
3 POKE B, PEEK(C)
4 GOTO 2
5 A=61456
6 B=61441
7 C=61457
8 D=1
9 GOTO 2
.. if multiple statements per line are allowed then this should be
faster ..
1 A=61456:B=61441:C=61457:D=1
2 WAIT A,D,D:POKE B,PEEK(C):GOTO 2
Lee.
2 Jan
2 Jan
8 a.m.
From: "Lee Davison" <lee at
geekdot.com>
After a bit of a ponder this should be about as fast as you can get
without putting multiple statements on one line ..
1 GOTO 5
2 WAIT A,D,D
3 POKE B, PEEK(C)
4 GOTO 2
5 A=61456
6 B=61441
7 C=61457
8 D=1
9 GOTO 2
.. if multiple statements per line are allowed then this should be
faster ..
1 A=61456:B=61441:C=61457:D=1
2 WAIT A,D,D:POKE B,PEEK(C):GOTO 2
Lee.
Hi
As Lee shows, not having to interpret the numbers each time
should speed things up. Reading decimal numbers requires
division. That is just about the slowest thing the computer can
do. If the values are stored in variables, they are already converted
to a computer friendly format.
Another thought is that your PC may not be keeping up with
9600. You might slow it to 1200 and see what happens.
Dwight
_________________________________________________________________
The MSN Entertainment Guide to Golden Globes is here. Get all the scoop.
http://tv.msn.com/tv/globes2007/
11:22 a.m.
>1 A=61456:B=61441:C=61457:D=1
>2 WAIT A,D,D:POKE B,PEEK(C):GOTO 2
On Tue, 2 Jan 2007, dwight elvey wrote:
Hi
As Lee shows, not having to interpret the numbers each time
should speed things up. Reading decimal numbers requires
division. That is just about the slowest thing the computer can
do. If the values are stored in variables, they are already converted
to a computer friendly format.
I'm coming in late on this discusssion, and I don't even know what machine
we're talking about.
1) Many versions of basic use single precision floating point as the
default variable type! Explicitly declaring A,B,C,D to be integer might
make a significant performance change.
2) Any particular reason to not do it in machine language?
11:29 a.m.
At 01:22 PM 1/2/2007, Fred Cisin wrote:
I questioned the assumption that the interpreter would parse
the integer value from source code each time. Wouldn't most BASICs
tokenize to a 16-bit integer in memory, once at loading?
- John
>1
A=61456:B=61441:C=61457:D=1
>2 WAIT A,D,D:POKE B,PEEK(C):GOTO 2
On Tue, 2 Jan 2007, dwight elvey wrote:
As Lee shows, not having to interpret the numbers
each time
should speed things up. Reading decimal numbers requires
division. That is just about the slowest thing the computer can
do. If the values are stored in variables, they are already converted
to a computer friendly format.
I'm coming in late on this discusssion, and I don't even know what machine
we're talking about.
12:15 p.m.
>> >1 A=61456:B=61441:C=61457:D=1
>> >2 WAIT A,D,D:POKE B,PEEK(C):GOTO 2
>I'm coming in late on this discusssion, and I don't even know what machine
>we're talking about.
On Tue, 2 Jan 2007, John Foust wrote:
I questioned the assumption that the interpreter would
parse
the integer value from source code each time. Wouldn't most BASICs
tokenize to a 16-bit integer in memory, once at loading?
BASICA, and other 5150 BASICs store those numbers in RAM as 32 bit
floating point! Thus, 10 X = [some number that does not come out evenly
in IEEE floating point], will actually come back differently when LISTed!
It is true that the code being executed is tokenized, NOT run from what we
think of as source code. But, at least in THOSE BASICs, all numbers are
floating point, unless explicitly declared otherwise, There is a
noticable effect on performance by declaring the variables to be integer.
2) WHAT machine, and what version of BASIC are we talking about here?
--
Grumpy Ol' Fred cisin at xenosoft.com
2:50 p.m.
New subject: ECMA 34 was : Recovering Cassette Tape Records
While speaking of cassette tapes :
I have two Philips cassette decks available. The rubber band has
disintegrated. I seem to remember that the ECMA standard was called ECMA 34.
The logic consists of Philips 6 volt TTL. It is a bit vague, but I am quite
sure that the bit rate is about 300 bps. The tapes can contain 250 K on each
side. There is a cut-out in the back of the cassettes, indicating side A or
B
They are available for the mailing cost (from Denmark)
Nico
6:16 p.m.
>2) WHAT machine, and what version of BASIC are we
talking about here?
On Tue, 2 Jan 2007, Grant Stockly wrote:
Altair 680. Version 1.1 Rev 3.2 of what seems to be a
modified Altair 8800
8k non extended basic.
Thank you
Then a lot of our assumptions based on newer BASICs may be invalid.
Does it have a DEFINT command?
12:23 p.m.
I questioned the assumption that the interpreter would
parse
the integer value from source code each time. Wouldn't most BASICs
tokenize to a 16-bit integer in memory, once at loading?
I don't know about this BASIC, but most Microsoft-derived BASICs do not
(I know for sure that Applesoft and Commodore BASICs don't).
In fact, pre-assigning constants to variables in Commodore BASIC is a very
frequent method of speeding up BASIC execution.
--
--------------------------------- personal: http://www.armory.com/~spectre/ ---
Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- California, land of cereal: some fruits, nuts, and a lot of flakes. --------
6564
days inactive
6565
days old
8 comments
7 participants
participants (7)
-
cisin＠xenosoft.com -
dkelvey＠hotmail.com -
grant＠stockly.com -
jfoust＠threedee.com -
lee＠geekdot.com -
nico＠farumdata.dk -
spectre＠floodgap.com