Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 beta 3/9/83; site hplabs.UUCP
Path: utzoo!decvax!harpo!seismo!hao!hplabs!jf
From: jf@hplabs.UUCP (Joe Falcone)
Newsgroups: net.lang.st80
Subject: Re: Smalltalk-80 Licenses
Message-ID: <1513@hplabs.UUCP>
Date: Wed, 8-Jun-83 15:01:09 EDT
Article-I.D.: hplabs.1513
Posted: Wed Jun  8 15:01:09 1983
Date-Received: Mon, 13-Jun-83 09:46:23 EDT
References: <4543@cornell.UUCP>, <4562@cornell.UUCP>
Organization: Hewlett Packard Labs, Palo Alto CA
Lines: 87

I have noticed that most (if not all) of the discussion of Smalltalk in this
newsgroup has been by people who have (at best) played with Smalltalk.

Some of the mis-information has been quite amusing at time, including
that from "pavel" who apparently has some association with PARC.

I was one of the members of the HP Smalltalk project (part of the Xerox
test distribution program).  Yes, campers, HP, DEC, Apple, and Tektronix
all have licenses for Smalltalk.  Did we pay for them?  Yes and No.
We didn't pay money, but we did pay engineering time to help debug the 
distribution.  (I doubt whether any of these companies would have actually
PAID real money for licenses - it was far easier to justify having
engineers spend a year fiddling with it).

I have "Smalltalk: The Language and Its Implementation" on my bookshelf.
Nice reading.

We have gone through the galley phase (including some
wild things that some Xerox text editor did to some
of the papers) on "Smalltalk 80: Bits of History, Words of Advice".
It is likely that this book will be paperback for those who are budget
conscious.  It is a collection of articles from the 4 test sites,
plus University of Washington (courtesy Al Borning), UCB, and Xerox.
Glenn Krasner managed the feat of coordinating a book whose authors
were sprinkled across the country.  He deserves much praise.

I have a late, but now old draft of "Smalltalk 80: The Interactive
Programming Environment" by Adele Goldberg.  I really don't know
what has happened to this book.  My draft is dated August 1982 and
looks for all the world to be camera-ready.  I'll be very surprised
if this book is delayed much longer.

My impression is that the fourth book, "Smalltalk 80: Creating a
User Interface and Graphical Applications" has always been on the
bottom of everyone's list at PARC.  Last I heard, which was a few 
months ago, it had not been started, but that may have changed.
However, given the fact that the first book was started back
in early 1980, it may be another year before we see this book.

Incidentally, for those of you contemplating publishing a collection
of technical works, Glenn's original schedule for "Bits of History" was:

June 15, 1982   First drafts

July 15, 1982   Peer review complete (authors review one anothers papers)

August 15, 1982 Drafts consolidated at PARC

September 15, 1982  Book is sent to publishers.  Pre-publication copies
                    ready in a month, real copies hit shelves in six.

We probably slipped 3 to 6 months - they are predicting summer or
fall for the "Bits of History" book.

As for the implementations out there, here are my impressions.
I have had the good fortune to see all of the implementations,
except for the SUN version.  As part of the "Bits of History" book,
there will be a table of Benchmarks for each implementation.

The fastest one is on the Xerox Dorado.  What else is new?

The big surprise was for second place.
The second fastest is neither the Dolphin nor the VAX implementations,
but a surprisingly clever job done by Tektronix on the 68000.
Now, as previously mentioned on the net, Peter Deutsch put Smalltalk
on a SUN-type workstation.  However, he has released no performance figures.
The Tektronix implementation is quite useable, the UCB, DEC, and HP
implementations are useable if you kick everyone else off the VAX
(Smalltalk is for personal computers, but I don't know whether you can call
a single-user VAX a personal computer).
The Apple 68000 implementation is slower than the VAX versions, but
is eminently more useable since only $10K worth of hardware is necessary
and that hardware is commercially available as the Lisa, including mouse.

For those of you who are interested in the practical chore of
implementing a sophisticated system like Smalltalk, I would highly
recommend "Bits of History" (or whatever they decide to call it).
All four companies contributed interesting articles, each discussing
different solutions to the same Smalltalk performance problems.
Of course, take my recommendation with a grain of salt since I am
one of the contributors and hence have a financial interest.

Hope I have cleared the air for a while,

Joe Falcone
hplabs!jf