-=> On 27 May 96  01:02 you wrote to me <=-

Hi Paul,

MS> I rarely use bitfields, since they are not portable & I'm quite
MS> comfortable with the bitwise arithmetic operators.

PE> "portable" in what sense.  If you are using them as internal
PE> variables, they should be fine to use.  If you are writing them
PE> out to a file, then only chars will do.  BFN.  Paul.

Yes, I meant they aren't portable for reading/writing files, which
is one place where they could be quite useful.

PE> P.S.  Plauger has already been contacted, and admitted it was
PE> wrong.

Ok, so now he knows better.  :)

PE> I also contacted him about a bug in the ISO C standard, which
PE> even needed an explanation of a possible scenario.  (their
PE> example where they use mktime()).

Details?

PE> /* According to P.J.  Plauger, in his book "The Standard C
PE> Library", ISBN 0-13-131509-9 on page 217, it certainly does, to
PE> quote:  "When you subtract two pointers in a C expression, the
PE> result has type ptrdiff_t.  It is an integer type that can
PE> represent negative values.  Almost certainly it is either int or
PE> long.  It is always the signed type that has the same number of
PE> \ bits as the unsigned type chosen for size_t, described below.
PE> (I said above that the use of these definitions is essentially
PE> unrelated.  These two definitions are themselves highly
PE> related.)" */

PE> /* According to Paul Edwards, with reference to ISO/IEC
PE> 9899:1990, it doesn't do anything of the sort.  */

I admit that that passage had me worried when I first read it.

Michael Stapleton of Graphic Bits.

 * AmyBW v2.10 *
... This tagline is encrypted
--- AmyBW v2.10