Forgot your password?
Cyhist Jan 23 1997 C
Document Actions
Date: Thu, 23 Jan 1997 01:54:44 -0800
Reply-To: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@SJUVM.STJOHNS.EDU>
Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@SJUVM.STJOHNS.EDU>
From: Efrem Lipkin <efrem@acm.org>
Subject: Re: CM> Re: Early Microsoft history
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
>Re emulation as a development technique: I agree with Robert Bickford that this was used long before Gates and Allen. A couple of cases I'm personally familiar with.
>
>1. At the Center for Research in Management Science, U.C. Berkeley, in 1972, we built a dual-processor timeshared system using Digital Scientific Meta4 microprogrammable processors (one running a custom-designed APL "virtual machine", and the other running a generalized systems programming language. We debugged our microcode on an emulator running on an SDS 940 under the Berkeley 940 timesharing system. The microcode was installed in the processor by using a knife to peel foil squares off a PC board; mistakes were hard to correct, so we worked hard to eliminate all the bugs on the emulator. (By the way, the principal investigators of this project, Chuck Grant and Mark Greenberg, went on to start "Kentucky Fried Computers", which became Northstar, a purveyor of CP/M computers featuring one of the first 5.25" floppy systems. Mark recently passed away.)
In the usual way history bounces off itself. Chuck & Mark were working on Community Memory's/Resource One's SDS 940 for at least part of that project.
In 1970 the people who created the SDS 940 timesharing system from the SDS 930 batch machine now under the name Berkeley Computing Corp I believe were emulating
their latest creation on a 940. In 1972 I was debugging a Data General Nova program
on a PDP-10 using a well-established emulator called MIMIC, which would also emulate
a PDP-11 and I believe several other minicomputers. Far superior to working on the
native box. All these extra bits in the PDP-10 word let you detect execution at,
reference of, or writing to any set of locations in your program. It also allowed
one to work via a modem from home!
Of course by then IBM had been emulating a 1401 on a 360 for years. Technically,
they were emulating both the 1401 and the 360 on a whole variety of strange microcoded
machines.
Efrem
================================================================== Efrem Lipkin 1811 Ward Street
CoDesign Berkeley, CA 94706
efrem@acm.org (510) 845-3170
______________________________________________________________________
Reply-To: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@SJUVM.STJOHNS.EDU>
Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@SJUVM.STJOHNS.EDU>
From: Efrem Lipkin <efrem@acm.org>
Subject: Re: CM> Re: Early Microsoft history
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
>Re emulation as a development technique: I agree with Robert Bickford that this was used long before Gates and Allen. A couple of cases I'm personally familiar with.
>
>1. At the Center for Research in Management Science, U.C. Berkeley, in 1972, we built a dual-processor timeshared system using Digital Scientific Meta4 microprogrammable processors (one running a custom-designed APL "virtual machine", and the other running a generalized systems programming language. We debugged our microcode on an emulator running on an SDS 940 under the Berkeley 940 timesharing system. The microcode was installed in the processor by using a knife to peel foil squares off a PC board; mistakes were hard to correct, so we worked hard to eliminate all the bugs on the emulator. (By the way, the principal investigators of this project, Chuck Grant and Mark Greenberg, went on to start "Kentucky Fried Computers", which became Northstar, a purveyor of CP/M computers featuring one of the first 5.25" floppy systems. Mark recently passed away.)
In the usual way history bounces off itself. Chuck & Mark were working on Community Memory's/Resource One's SDS 940 for at least part of that project.
In 1970 the people who created the SDS 940 timesharing system from the SDS 930 batch machine now under the name Berkeley Computing Corp I believe were emulating
their latest creation on a 940. In 1972 I was debugging a Data General Nova program
on a PDP-10 using a well-established emulator called MIMIC, which would also emulate
a PDP-11 and I believe several other minicomputers. Far superior to working on the
native box. All these extra bits in the PDP-10 word let you detect execution at,
reference of, or writing to any set of locations in your program. It also allowed
one to work via a modem from home!
Of course by then IBM had been emulating a 1401 on a 360 for years. Technically,
they were emulating both the 1401 and the 360 on a whole variety of strange microcoded
machines.
Efrem
================================================================== Efrem Lipkin 1811 Ward Street
CoDesign Berkeley, CA 94706
efrem@acm.org (510) 845-3170
______________________________________________________________________
