Just to start a discussion

IMO bit and bool are not synonymous. 'bit' refers to storage size, whereas 'bool' refers to semantics. 'bit' is more restrictive, and can be used to implement a boolean variable.

In relationship with the thread on structs vs. classes: I could imagine that a class has 'bool' variables, which the compiler is free to move around and where the programmer doesn't really care too much if it takes 1 bit or 8 bit (for more efficient access perhaps). 'bit' on the contrary is very specific about this, I would be surprised if my compiler used 8 bits to store my bit...

Jeroen van Bemmel wrote:
> IMO bit and bool are not synonymous. 'bit' refers to storage size, whereas
> 'bool' refers to semantics. 'bit' is more restrictive, and can be used to
> implement a boolean variable.

D's bit has semantics identical to proper bool; it wasn't a misnomer when it was introduced, but it is now.  The name should be changed, Walter knows the problem, let's never mention it again.

In article <b13pgl$28lp$1@digitaldaemon.com>, Burton Radons says...
>
>Jeroen van Bemmel wrote:
>> IMO bit and bool are not synonymous. 'bit' refers to storage size, whereas 'bool' refers to semantics. 'bit' is more restrictive, and can be used to implement a boolean variable.
>
>D's bit has semantics identical to proper bool; it wasn't a misnomer when it was introduced, but it is now.  The name should be changed, Walter knows the problem, let's never mention it again.
>

Will the "bit" name ever go away and be replaced by "bool" or "Boolean" or will it be like pre-ISO C99 where everyone invents a new name for Booleans?

Once a lot of code gets written one way then the language is almost forced to carry that style forward,  example: K&R C idioms like defaulting to int return type had to be allowed in ISO C90, finally dumped in C99.

Why? I like the name. It's short. Cubistic. Minimalistic. Doesn't pretent to be more important than it is. And versatile. I don't see why some people can't just get used to calling bool a bit? Why has bit to go away? What would you prefer: to have a bitarray (what it always is), or a boolarray (sometimes a lie)? A bitarray might be a flag array, but it generally needn't. And how about bitstreams vs. boolstreams?


Mark T wrote:
> Will the "bit" name ever go away and be replaced by "bool" or "Boolean" or will
> it be like pre-ISO C99 where everyone invents a new name for Booleans? 
> 
> Once a lot of code gets written one way then the language is almost forced to
> carry that style forward,  example: K&R C idioms like defaulting to int return
> type had to be allowed in ISO C90, finally dumped in C99.
> 

"Ilya Minkov" <midiclub@8ung.at> wrote in message news:b25rbm$2a1v$1@digitaldaemon.com...
> Why? I like the name. It's short. Cubistic. Minimalistic. Doesn't pretent to be more important than it is. And versatile. I don't see why some people can't just get used to calling bool a bit? Why has bit to go away? What would you prefer: to have a bitarray (what it always is), or a boolarray (sometimes a lie)? A bitarray might be a flag array, but it generally needn't. And how about bitstreams vs. boolstreams?
>
bit is storage, bool is semantics;
a bool array _might_ be implemented as a bit array, but could be char or int
per bool
a bit array / stream is bits.

bits currently can not be passed by reference, you require a pointer to the
byte/int + shift
however a bool could be, because it is valid to store a bool in an int to
pass to/from func and bit within a local/member.

>
> Mark T wrote:
> > Will the "bit" name ever go away and be replaced by "bool" or "Boolean"
or will
> > it be like pre-ISO C99 where everyone invents a new name for Booleans?
> >
> > Once a lot of code gets written one way then the language is almost
forced to
> > carry that style forward,  example: K&R C idioms like defaulting to int
return
> > type had to be allowed in ISO C90, finally dumped in C99.
> >
>