If -1 is TRUE, what value is FALSE ? 3

[JohnBates]

Hi,

This VB app reads a variable from an .ini file.

The .ini file contains DDEenabled=-1 which mean that DDE IS enabled.

To make the value FALSE, would I have to change it to 1?

Thanks, John

 

[Sypher2]

Often times simply any number other than -1 will be a false value. If the app isn't critical, you might just play with the number and see what happens.

 

[thornmastr]

False is simply what is NOT true. The following will work nicely.

Iff DDEenabled is defined as boolean then DDEenabled = FALSE.

DDEenabled = 0.

The reason that TRUE is -1 harks back to the days when assembler was King and -1 simply loaded a register with FFFFx which insured that any OR operation would, if I remember correctly, force a "branch conditions reset". (BCR).

Robert Berman
Data Base consultant
Vulcan Software Services
thornmastr@earthlink.net

 

[strongm]

>Often times simply any number other than -1 will be a false value
>False is simply what is NOT true

Actually, we're all arse-about-tip here. FALSE is 0. Anything that isn't FALSE (i.e anything that isn't 0) is TRUE

 

[chiph]

I agree with strongm -

False = 0
True = Not False

You can run into trouble when talking to other platforms, as their runtime may define "not 0" to be "1", while VB seems to define is as "-1"

Chip H.


If you want to get the best response to a question, please check out FAQ222-2244 first

 

[xbafge]

I beleive that in Visual BASIC, the truth value true is specified by the Boolean literal 'True' ( case insensitive ); the trught value false is specified by the Boolean lietral 'False'. Not <Boolean-variable> results in the opposite truth value. If such is the case the underlying ( numeric ) representation of True and False is irrelevant except whilr crossing type boundaries.

 

[strongm]

You can believe what you like, but you are sadly wrong.

 

[IFRs]

If you use "True" and "False" in VB you can't go wrong. To make DDEenabled=-1 false, say DDEenabled = False. If you say "print DDEenabled" it will say "false".

 

[jel]

xbafge has a point. To me, -1 is a numeric, not a boolean value. In VB-code you could use -1 as 'true', and 1 just as well, but eventually you'll get confusion as
-1 does not equal 1,
while true equals true.

 

[johnwm]

True Booleans can only have 2 values, which is the equivalent of a single bit. VB native numeric type is a Long. Storing signed Integers as "1's complement form" results in -1 being stored as &HFFFF. In other words, all 32 bits are set to 1. Zero is stored as &H0, so all bits are set to zero.

The Not function (at processor level) results in an inversion of a bit pattern, so Not 0 will return -1. Once False is assigned a zero value, True becomes -1.

There is a processor level flag which is set when the main processor register contains zero, and a fast test is available to check the zero flag. Therefore although False will be stored as zero, and tested for as zero, True will be set as &HFFFF, but tested as any non-zero value

________________________________________________________________
If you want to get the best response to a question, please check out FAQ222-2244 first

'If we're supposed to work in Hex, why have we only got A fingers?'

 

[strongm]

Can't go wrong?

debug.? true = not(not 1)
debug.? not(not 1) = true

 

[johnwm]

strongm,
Agree entirely - Precisely as I said !

________________________________________________________________
If you want to get the best response to a question, please check out FAQ222-2244 first

'If we're supposed to work in Hex, why have we only got A fingers?'

 

[strongm]

Yep. Not you I'm disagreeing with... ;-)

 

[johnwm]

Quite so!


________________________________________________________________
If you want to get the best response to a question, please check out FAQ222-2244 first

'If we're supposed to work in Hex, why have we only got A fingers?'

 

[xbafge]

johnwm has put forth a powerful and lucid reason for the adoption of 0 to represent the Boolean value 'false' by 0 and 'true' by any non-zero number, especially -1 by the *implementor* of a high-level language. But what is the point of abstraction if you revert to basics when the abstarct mechanisms are serving the very purpose they were developed for?

E.g.: in C/C++ many people think the value of a null pointer is 0 but in fact it may be any invalid address. In the construct int *ip = 0, 0 is not the number 0 but the null pointer literal. An implementor of a 4-bit C compiler is free to implement the null pointer literal 0 as the number 2,000,000,000 ( an invalid address for 4-bit machines ).