Path: utzoo!attcan!uunet!ncrlnk!ncr-sd!hp-sdd!hplabs!decwrl!purdue!mailrus!tut.cis.ohio-state.edu!bloom-beacon!CS.BROWN.EDU!jb
From: jb@CS.BROWN.EDU
Newsgroups: comp.windows.x
Subject: Re: bcopy
Message-ID: <8812091358.AA06741@tracy.cs.brown.edu>
Date: 9 Dec 88 08:58:45 GMT
Sender: daemon@bloom-beacon.MIT.EDU
Organization: The Internet
Lines: 18

>			char a[20];
> /* first sample: */
>			bcopy (&a[0], &a[1], 19);
> /* second sample: */
>			bcopy (&a[1], &a[0], 19);

Assuming the array initially contains "abcdefghijklmnopqrst", the
first should yield "aabcdefghijklmnopqrs" while the second will
result in "bcdefghijklmnopqrstt".  This is what Sun means when they
say that "overlapping strings are handled correctly."  The second
one should behave that way in all cases.  The first might end up
with the array containing 20 'a's if overlapping copies are done
in a simple character at a time loop.

To make this a bit more X related, I had to fix bcopy on the Encore
Multimax to work around this problem.

				Jim