Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!unix.cis.pitt.edu!pitt!willett!ForthNet
From: ForthNet@willett.pgh.pa.us (ForthNet articles from GEnie)
Newsgroups: comp.lang.forth
Subject: X3J14 Holding Pattern Here
Message-ID: <2783.UUL1.3#5129@willett.pgh.pa.us>
Date: 17 May 91 11:36:55 GMT
Organization: (n.) to be organized.  But that's not important right now.
Lines: 42

Category 10,  Topic 12
Message 73        Thu May 16, 1991
R.BERKEY [Robert]            at 02:57 PDT
 
 
 re: Jack J. Woehr, 91-05-07
 Message-ID: <1991May7.174009.24072@mnemosyne.cs.du.edu>

I have also posited that objections with a current standard are one of the
motivations for participating in a subsequent standard's process.  I have also
stated that a corollary problem for those now involved is to damp the waves of
change between standards.

Good decisions, that take into consideration both the needs of implementations
and the needs of programs, are not necessarily easy to obtain.  Ever since I
began to study the division available in Forth-79 and Forth-83 I've wondered
why the

 ( d n -- n n )

signed division word was not included.  Only now that such are proposed for
inclusion (and I was one of those who voted for their inclusion), have I come
to understand why.  They are not primitives that can be used in building
higher precision operations. Additionally, their internal inconsistencies
create difficulties for programmers (notably in the coding of rounded-to-
nearest algorithms) even when used in their single precision form.  As such,
their placement in the required word set is dubious.

Intel with their 8088/8086 and Motorola with their 68000 have each
demonstrated the complexities.  Intel's difficulties in implementing -32768 as
a dividend have been reported in Forth dimensions as well as are recorded in
the Intel 80286 Programmer's Reference Manual. Motorola's signed opcode is
sufficiently inefficient that an ordinary implementation of floored division
based on unsigned division will divide positive numbers such as 8 by 3 more
efficiently than will the Motorola signed opcode (1 to 2% difference).

Robert
-----
This message came from GEnie via willett.  You *cannot* reply to the author
using e-mail.  Please post a follow-up article, or use any instructions
the author may have included (USMail addresses, telephone #, etc.).
Report problems to: dwp@willett.pgh.pa.us _or_ uunet!willett!dwp