| Author |
Message |
|
Haubensac
#1 / 11
|
 unsigned long ? Hi, under VB there is the long type on 4 bytes : -2 147 483 648 to 2 147 483 647
but is there a way to treat this a a real DWORD, i.e. an unsigned long
from 0 to 4 290 000 000 ? By treat I mean calculation, printing, etc...
a "single" does not seem to do that because it is in scientific notation...
thanks
david.
|
| Fri, 09 Aug 2002 03:00:00 GMT |
|
 |
|
Derek Ros
#2 / 11
|
unsigned long ?
Quote: > Hi, > under VB there is the long type on 4 bytes : -2 147 483 648 to 2 147 483
647
> but is there a way to treat this a a real DWORD, i.e. an unsigned long
> from 0 to 4 290 000 000 ? By treat I mean calculation, printing, etc...
> a "single" does not seem to do that because it is in scientific
notation...
A DOUBLE should give you the range that you need. Cheers
Derek
|
| Fri, 09 Aug 2002 03:00:00 GMT |
|
|
|
Frits van Bomme
#3 / 11
|
unsigned long ?
Quote:
> > under VB there is the long type on 4 bytes : -2 147 483 648 to 2 147 483
> 647
> > but is there a way to treat this a a real DWORD, i.e. an unsigned long
> > from 0 to 4 290 000 000 ? By treat I mean calculation, printing, etc...
> > a "single" does not seem to do that because it is in scientific
> notation...
> A DOUBLE should give you the range that you need.
Still in scientific notation... The Currency data type (a scaled 8-byte integer, if I recall correctly) has large enough a range and up to 4 decimals (actual ones, as in the number stored in memory is devided by 10000) Frits van Bommel
|
| Fri, 09 Aug 2002 03:00:00 GMT |
|
|
|
Judson McClendo
#4 / 11
|
unsigned long ? Quote:
>Still in scientific notation... The Currency data type (a scaled 8-byte
>integer, if I recall correctly) has large enough a range and up to 4
>decimals (actual ones, as in the number stored in memory is devided by
>10000)
The Decimal data type (subtype of Variant, use CDec), gives more than 28 digits, with number of decimal places variable from 0 to 28. --
Sun Valley Systems http://www.sunvaley.com "For God so loved the world that He gave His only begotten Son, that whoever believes in Him should not perish but have everlasting life."
|
| Fri, 09 Aug 2002 03:00:00 GMT |
|
|
|
Michael Carto
#5 / 11
|
unsigned long ?
Quote:
> > > under VB there is the long type on 4 bytes : -2 147 483 648 to 2 147 483
> > 647
> > > but is there a way to treat this a a real DWORD, i.e. an unsigned long
> > > from 0 to 4 290 000 000 ? By treat I mean calculation, printing, etc...
> > > a "single" does not seem to do that because it is in scientific
> > notation...
> > A DOUBLE should give you the range that you need.
> Still in scientific notation...
The scientific notation is just a representation issue. You might be able to use the Format$() function to overcome that. --
MikeC
Please reply to the group.
|
| Fri, 09 Aug 2002 03:00:00 GMT |
|
|
|
Frits van Bomme
#6 / 11
|
unsigned long ?
Quote:
> >Still in scientific notation... The Currency data type (a scaled 8-byte
> >integer, if I recall correctly) has large enough a range and up to 4
> >decimals (actual ones, as in the number stored in memory is devided by
> >10000)
> The Decimal data type (subtype of Variant, use CDec), gives more than
> 28 digits, with number of decimal places variable from 0 to 28.
I've not really got a lot of VB experience, I just recalled that from PDS... Frits van Bommel
|
| Sun, 11 Aug 2002 03:00:00 GMT |
|
|
|
Frits van Bomme
#7 / 11
|
unsigned long ?
Quote:
berichtnieuws
> > > A DOUBLE should give you the range that you need.
> > Still in scientific notation...
> The scientific notation is just a representation issue. You might be able
to
> use the Format$() function to overcome that.
The main problem I have with it is that it's inaccurate, since there's a lot of numbers that can't be expressed as well in binary scientific notation as they can be in any decimal form. For example, .2 can't be precisely represented in binary scientific notation, while it can be in any decimal form. Frits van Bommel
|
| Sun, 11 Aug 2002 03:00:00 GMT |
|
|
|
Michael Carto
#8 / 11
|
unsigned long ?
Quote:
> > > > A DOUBLE should give you the range that you need.
> > > Still in scientific notation...
> > The scientific notation is just a representation issue. You might be able
> > to use the Format$() function to overcome that.
> The main problem I have with it is that it's inaccurate, since there's a lot
> of numbers that can't be expressed as well in binary scientific notation as
> they can be in any decimal form.
> For example, .2 can't be precisely represented in binary scientific
> notation, while it can be in any decimal form.
Why are we discussing 0.2 here? Didn't you start this thread by asking for unsigned long? Were you planning to shove 0.2 into an unsigned long? Single and Double data types have no trouble representing whole numbers.
--
MikeC
Please reply to the group.
|
| Sun, 11 Aug 2002 03:00:00 GMT |
|
|
|
Frits van Bomme
#9 / 11
|
unsigned long ?
Quote: > Why are we discussing 0.2 here? Didn't you start this thread by asking for
> unsigned long? Were you planning to shove 0.2 into an unsigned long?
No I didn't. Somebody who called himself "Haubensack" did. Hasn't posted anything in this thread ever since. Right now I'm discussing the disadvantages of Singles and Doubles. Quote: > Single and Double data types have no trouble representing whole numbers.
Except they're not totally accurate when it comes to large numbers. Frits van Bommel
|
| Sun, 11 Aug 2002 03:00:00 GMT |
|
|
|
Derek Ros
#10 / 11
|
unsigned long ?
Quote:
> > Why are we discussing 0.2 here? Didn't you start this thread by asking
for
> > unsigned long? Were you planning to shove 0.2 into an unsigned long?
> No I didn't. Somebody who called himself "Haubensack" did. Hasn't posted
> anything in this thread ever since.
> Right now I'm discussing the disadvantages of Singles and Doubles.
> > Single and Double data types have no trouble representing whole numbers.
> Except they're not totally accurate when it comes to large numbers.
That's true of course, but a DOUBLE is perfectly capable of representing an unsigned LONG with complete accuracy. In fact even 2^48 can be represented exactly by a DOUBLE. Also DOUBLEs have the advantage that they exist in GWBASIC, QBASIC, QuickBASIC PDS, VBDOS, VBWin and powerbasic, so they can be used even when you are stuck with a system which doesn't have the useful CURRENCY type. Cheers
Derek
|
| Sun, 11 Aug 2002 03:00:00 GMT |
|
|
|
Michael Mattia
#11 / 11
|
unsigned long ? If you are calling a routine which wants an "unsigned long", just pass it a "signed long" making the sign adjustments in the call-ING routine. Whatever the routine is, it doesn't care HOW you set bit 31 (the sign bit), it just cares that it IS set, and that there are 32 bits worth of "owned" memory at the pointer location (if passing by reference) or on the stack (when passing by value). Let's say you want to send the value &hFE123456, which, as a positive
number, is beyond the range of a signed long. So, send it, with CALL MyProc
( CVL(CHR$(&h56) & CHR$(&h34) & CHR$(&h12) & CHR$(&hFE)) (I reversed the
bytes to get the data in Intel format).
Reading it back is left as an exercise for the reader. Arithmetic on the
result is the responsibility of the user.
--
Michael Mattias
Racine WI USA
Quote: [stuff about passing an unsigned long from VB to something)
|
| Mon, 12 Aug 2002 03:00:00 GMT |
|
| |
