Hi Mark (aka Mark Probert [mailto:probertm@nortelnetworks.com]):
You enlightened last Wednesday, December 11, 2002 2:46 AM:
My own experience (based on developing real-time trading
systems for some merchant banks), is that the best solution
is to use scaled integers, to 100th of a cent. So, $154.34
→ 1543400. You then get very fast operations (they are all
integer operations) and, in general, well with ± 4 billion.
Ruby, with BigInt makes it trivial.
Hmmm… very clever/insightful, sir Mark. I’m putting your tip on my code
bank 
C:\family\ruby>type a1.rb
def t
(2.4)/(0.2)
end
x = t
p x
p x.truncate
C:\family\ruby>ruby a1.rb
12
11
C:\family\ruby>type a1.rb
def t
(2.4100)/(0.2100)
end
x = t
p x
p x.truncate
C:\family\ruby>ruby a1.rb
12.0
12
I do not get truncation problems 
Regards,
-mark.
Kind regards,
-botp
Hi, botp.
Hmmm… very clever/insightful, sir Mark. I’m putting your tip on my code
bank
Scaled integers can be very handy in lots of situations.
I guess that we have just got so used to having floating
point available that we forget about what we lose.
For example, using scaled integers you can represent 1/3,
something that you can’t do using (most) floating point
representations.
Interestingly, there are a bunch of rational approximations
for irrationals that allow you to do more “serious” math
using scaled integers:
const ratio error
pi 355/113 8.5e-8
sqrt(2) 19601/13860 1.5e-9
sqrt(3) 18817/10864 1.1e-9
e 28667/10564 5.5e-9
c 24559/8192 1.6e-9
I haven’t done any checks to see if this will make any difference
to Ruby’s performance.
Regards,
-mark.
···
At 10:25 AM 12/11/2002 +0900, you wrote:
Ah, but c is exactly 300,000 kilometers per second*, if you use metric,
its not a problem!
···
Mark Probert (probertm@nortelnetworks.com) wrote:
Interestingly, there are a bunch of rational approximations
for irrationals that allow you to do more “serious” math
using scaled integers:
const ratio error
pi 355/113 8.5e-8
sqrt(2) 19601/13860 1.5e-9
sqrt(3) 18817/10864 1.1e-9
e 28667/10564 5.5e-9
c 24559/8192 1.6e-9
–
Eric Hodel - drbrain@segment7.net - http://segment7.net
All messages signed with fingerprint:
FEC2 57F1 D465 EB15 5D6E 7C11 332A 551C 796C 9F04
Emmm, no
Speed of light in vacuum c = 1/sqrt(epsilon_0 * mu_0) =~ 2.997925 x 10^8 m.s^-1 (*)
It’s a very good approximation, though.
(*) S. Ramo, J.R. Whinnery, and T.V. Duzer, Fields and waves in
communication electronics, 3rd ed, 1994
Value recommended by National Academy of Sciences—National Research
Council Committee on Fundamental Constants in 1963.
You might find a better value, though but I’m staying w/ 3e8m.s^-1 for my
calcs.
···
On Thu, Dec 12, 2002 at 04:03:25AM +0900, Eric Hodel wrote:
Mark Probert (probertm@nortelnetworks.com) wrote:
Interestingly, there are a bunch of rational approximations
for irrationals that allow you to do more “serious” math
using scaled integers:
const ratio error
pi 355/113 8.5e-8
sqrt(2) 19601/13860 1.5e-9
sqrt(3) 18817/10864 1.1e-9
e 28667/10564 5.5e-9
c 24559/8192 1.6e-9
Ah, but c is exactly 300,000 kilometers per second*, if you use metric,
its not a problem!
–
_ _
__ __ | | ___ _ __ ___ __ _ _ __
'_ \ / | __/ __| '_ _ \ / ` | ’ \
) | (| | |__ \ | | | | | (| | | | |
.__/ _,|_|/| || ||_,|| |_|
Running Debian GNU/Linux Sid (unstable)
batsman dot geo at yahoo dot com
We apologize for the inconvenience, but we’d still like yout to test out
this kernel.
– Linus Torvalds, announcing another kernel patch
Hello Eric,
Wednesday, December 11, 2002, 10:03:25 PM, you wrote:
Ah, but c is exactly 300,000 kilometers per second*, if you use metric,
its not a problem!
are you sure that light speed was be known 200 years ago?
···
–
Best regards,
Bulat mailto:bulatz@integ.ru
Actually, the meter has been redefined to be 1/299,792,458th of the
distance that light travels in a vacuun in 1 second. Thus, the speed of
light is exactly 299,792,458 m/s.
Thins brings up an interesting point. Originally, the second was defined
as the period of a pendulum with length 1m. If that’s still the
definition, then this would be circular.
I wonder if the second has also been redefined.
Daniel Carrera
Graduate Teaching Assistant. Math Dept.
University of Maryland. (301) 405-5137
Interestingly, there are a bunch of rational approximations
for irrationals that allow you to do more “serious” math
using scaled integers:
const ratio error
pi 355/113 8.5e-8
sqrt(2) 19601/13860 1.5e-9
sqrt(3) 18817/10864 1.1e-9
e 28667/10564 5.5e-9
c 24559/8192 1.6e-9
Ah, but c is exactly 300,000 kilometers per second*, if you use metric,
its not a problem!
Emmm, no
Speed of light in vacuum c = 1/sqrt(epsilon_0 * mu_0) =~ 2.997925 x
10^8 m.s^-1 (*)
···
On Thu, 12 Dec 2002, Mauricio Fernández wrote:
On Thu, Dec 12, 2002 at 04:03:25AM +0900, Eric Hodel wrote:
Mark Probert (probertm@nortelnetworks.com) wrote:
It’s a very good approximation, though.
(*) S. Ramo, J.R. Whinnery, and T.V. Duzer, Fields and waves in
communication electronics, 3rd ed, 1994
Value recommended by National Academy of Sciences—National Research
Council Committee on Fundamental Constants in 1963.
You might find a better value, though but I’m staying w/ 3e8m.s^-1 for my
calcs.
–
_ _
__ __ | | ___ _ __ ___ __ _ _ __
'_ \ / | __/ __| '_ _ \ / ` | ’
) | (| | |__ \ | | | | | (| | | | |
.__/ _,|_|/| || ||_,|| |_|
Running Debian GNU/Linux Sid (unstable)
batsman dot geo at yahoo dot com
We apologize for the inconvenience, but we’d still like yout to test out
this kernel.
– Linus Torvalds, announcing another kernel patch
Found that somewhere:
second (s): In the International System of Units (SI), the time interval
equal to 9,192,631,770 periods of the radiation corresponding to the
transition between the two hyperfine levels of the ground state of the
cesium-133 atom. (188)
But perhaps this is old now, in fact I used to believe that it was
defined in terms of the time an electromagnetic wave takes to travel
299,792,458m in a vacuum
BTW, the metric system RULEZ! }:->
http://www.matsushita.co.jp/corp/timecapsule/record/4_10.html
The proof, that’s probably what matz uses
···
On Thu, Dec 12, 2002 at 09:54:08AM +0900, Daniel Carrera wrote:
Actually, the meter has been redefined to be 1/299,792,458th of the
distance that light travels in a vacuun in 1 second. Thus, the speed of
light is exactly 299,792,458 m/s.
Thins brings up an interesting point. Originally, the second was defined
as the period of a pendulum with length 1m. If that’s still the
definition, then this would be circular.
I wonder if the second has also been redefined.
–
_ _
__ __ | | ___ _ __ ___ __ _ _ __
'_ \ / | __/ __| '_ _ \ / ` | ’ \
) | (| | |__ \ | | | | | (| | | | |
.__/ _,|_|/| || ||_,|| |_|
Running Debian GNU/Linux Sid (unstable)
batsman dot geo at yahoo dot com
abuse me. I’m so lame I sent a bug report to debian-devel-changes
– Seen on #Debian
I’m attempting to spawn a thread when a user clicks a
button. The thread makes a system call. If I don’t
do a thread.join immediately after, the app crashes
(unknown software exception). Does FXRuby not like it
when a child thread makes system calls? A thread in a
’standard’ Ruby app (without FXRuby) can make these
calls just fine.
Any ideas?
···
#---------------------------------------
require 'fox’
include Fox
class MyMainWindow < FXMainWindow
def initialize(app)
super(app, “Thread Test”)
#layout screen
button = FXButton.new(self, "Push me.")
button.connect(SEL_COMMAND, method(:onButton))
end
def create
super
show(PLACEMENT_SCREEN)
end
def onButton(sender, sel, item)
thread = Thread.new do
`cmd /C dir`
end
#thread.join
end
end
theApp = FXApp.new
theApp.init(ARGV)
myWindow = MyMainWindow.new(theApp)
theApp.create
theApp.run
But perhaps this is old now, in fact I used to believe that it was
defined in terms of the time an electromagnetic wave takes to travel
299,792,458m in a vacuum
Ha!
This would make some interesting definitions:
Speed of light (c): 299,792,458 meters per second.
Meter (m): 1/299,792,458th of the distance that light travels in 1
second.
Second (s): The time that light takes to travel 299,792,458 meters.
Daniel Carrera
Graduate Teaching Assistant. Math Dept.
University of Maryland. (301) 405-5137
···
BTW, the metric system RULEZ! }:->
http://www.matsushita.co.jp/corp/timecapsule/record/4_10.html
The proof, that’s probably what matz uses
–
_ _
__ __ | | ___ _ __ ___ __ _ _ __
'_ \ / | __/ __| '_ _ \ / ` | ’
) | (| | |__ \ | | | | | (| | | | |
.__/ _,|_|/| || ||_,|| |_|
Running Debian GNU/Linux Sid (unstable)
batsman dot geo at yahoo dot com
abuse me. I’m so lame I sent a bug report to debian-devel-changes
– Seen on #Debian
Well, we all know that one of them should be wrong, right? In this case
it is the last one (for example, see
http://physics.nist.gov/cuu/Units/second.html).
Regards,
Bill
···
Daniel Carrera dcarrera@math.umd.edu wrote:
Ha!
This would make some interesting definitions:
Speed of light (c): 299,792,458 meters per second.
Meter (m): 1/299,792,458th of the distance that light travels in 1
second.
Second (s): The time that light takes to travel 299,792,458 meters.
Jason Persampieri wrote:
I’m attempting to spawn a thread when a user clicks a
button. The thread makes a system call. If I don’t
do a thread.join immediately after, the app crashes
(unknown software exception). Does FXRuby not like it
when a child thread makes system calls? A thread in a
‘standard’ Ruby app (without FXRuby) can make these
calls just fine.
Any ideas?
Which version of Ruby? If it’s Ruby 1.6.7, please try it under Ruby
1.7.3 to see if it works there.