Re: bug in bitfield code

From: Jaco van Iterson (J.C.vanIterson@ET.TUDelft.NL)
Date: 03/08/96


On Thu, 7 Mar 1996, Brian Pape wrote:

> > > I found a bug in the bitfield code.
> > > should be changed to:
> > >                                     *b &= ~(*c); }
> > 
> > Now I've heard a lot about bitfields recently but I'm mnot exactlyu sure
> > what iut is they do. Can someone let me know what advantage bitfields have
> > over bitvectors?
> 
> They're the same thing (at least as applied to circle code), it just lets
> you store up to (for instance) 32 on/off or true/false (whatever you want
> to call it) flags in a 32 bit integer.  Technically, you could consider a
> 'bit field' to be of a storage field of 1 or more bits, and a bitvector to
> be a value holding 'n' bitfields of the same size, but since they aren't
> used that way in the circlemud code, you can consider bitfield and
> bitvector to be identical.  (as in a value holding up to 'n' single-bit 
> flags).
> 
> Brian Pape
> bpape@ezl.com
> 
> 

You can of course store a bitvector in a bitfield but that wouldn't give you
much advantages above using a normal integer to do that (maybe you can use
a tiny bit less memory that way).

Better use of bitfields would be to make them all the size of 1 bit and put
them together in a structure. That way you can have a unlimited number of bits.
Every bit(field) is then a variable which can be 0 or 1 tightly packed into
a structure.


Jaco



This archive was generated by hypermail 2b30 : 12/07/00 PST