Newsgroups: comp.lang.smalltalk
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel-eecis!netnews.com!howland.erols.net!news.mathworks.com!blanket.mitre.org!world!carlg
From: carlg@world.std.com (Carl E Gundel)
Subject: Re: NEXT Smalltalk
Message-ID: <E6uwxs.KvC@world.std.com>
Organization: The World, Public Access Internet, Brookline, MA
X-Newsreader: TIN [version 1.2 PL2]
References: <5f5l61$i00$3@blackice.winternet.com> <331A754F.279E@concentric.net> <E6L8Au.DsB@world.std.com> <33239DCE.47E9@bobo.rd.wdi.disney.com>
Date: Tue, 11 Mar 1997 02:15:28 GMT
Lines: 27

Dan Ingalls (dani@bobo.rd.wdi.disney.com) wrote:
: Carl E Gundel wrote:
: > Are you sure?  Everything I've ever read at least implied that the
: > software VM came first, and that optimized hardware like the Dorado came
: > later.  Can you point me to some article that says the hardware came first?

: No, but I was there and I can tell you.
: I implemented the first Smalltalk system (ST-72) in Nova code and we 
: ran it on DG Novas before the first Alto was built.  AFTER the alto 
: was built, we continued to run for a while in Nova code, with our only 
: microcode assist being a music kernel (done primarily by Steve 
: Saunders) and an animation kernel (primraily by Steve Purcell).
...
: Before I wander too far afield, the bottom line is that the 
: microcodable hardware (Alto) came first, and spawned a number of 
: interesting kernels for music, graphics and language emulation, 
: Smalltalk being one.  The writable microstore worked so well that 

Thanks for the wonderful reply!  I'm glad we have the benefit of so many 
of the original Smalltalkers in this group.  :)

Carl
-- 
------------------------------------------------------------------
 Carl Gundel  carlg@world.std.com  Shoptalk Systems  508-872-5315
 author of Liberty BASIC, a 1996 PC Magazine Awards Finalist!
 http://world.std.com/~carlg/basic.html
