RE: System 3 Code

Mel,

Very nice recap.  One point, though, on your comment that relational
databases hadn't been invented yet.  The relational database model was first
described in an IBM research publication around the time the System/3 was
announced (1969/1970 timeframe, by a Dr. Edgar F. Codd).  RSI (Relational
Software, Inc.) released the first commercial relational database in 1979
(called Oracle V.2). IBM came out with its first commercial product based on
the relational model in 1983 with the Query Management Facility.

I believe everything you state is historically correct, but just wanted to
put the rdbms portion into the same historical perspective.

Steve Brazzell

p.s. ...kind of ironic, isn't it, that an IBMer came up with the relational
database concept but let another company (Oracle) take first advantage of
it.


-----Original Message-----
From: owner-rpg400-l@midrange.com [mailto:owner-rpg400-l@midrange.com]On
Behalf Of Mel Rothman
Sent: Wednesday, September 27, 2000 4:28 AM
To: RPG400-L@midrange.com
Subject: Re: System 3 Code


Most System/3 computers had disk drives.  Only the very early System/3 Model
10s
lacked disk I/O.

The first two models were System/3 Model 10 (first a card-only version with
a MFCU
(multi-function card unit); later one with disk I/O and MFCU)) and System/3
Model 6
(no punched cards, keyboard/printer console, disk).  System/3 was announced
in 1969.

Later models were quite powerful for their day.  My last project as an IBM
Systems
Engineer was designing and programming an online (3270 CRT terminals) system
for a
credit union on a System/3 Model 15D.  The system went live on January 3,
1978.

That particular System/3 had 384K bytes of memory, 3 partitions, and
about 140 MB of disk storage (two removable 3340 disk drives - the platters,
heads,
and arms were sealed inside the removable cartridge).  By October 1976, the
credit union
was running 25 3270 terminals on the System/3.  By that time, most System/3s
were cardless
with unit record I/O handled by direct-attached 3741 diskette data stations.

The programming language was RPG II, which supported sequential, and direct
(by RRN or KEY), disk I/O.  Exception output was also supported.  There were
no externally described files, no IF, DO, SELECT, FOR, CASxx, etc., op
codes.
There was no database (flat files only), no journaling, and no commitment
control.

RPG II was the only language at first; it was later joined by COBOL and
FORTRAN.

Relational databases hadn't been invented.  Spooling was supported only for
system printers (I don't remember whether more than one per system was
supported)
and the system spool file could fill up!  Maximum program size was 64K.

The Communications Control Program (CCP) supported online programming.
CCP programs could not exceed 32K.  The RPG II compiler supported external
buffers
and external memory resident overlays.  Screen I/O was supported by DFF
(Display File Facility), which I think was part of CCP.  There was no SEU,
but a field developed program for online editing of source members was
widely used.  The disk was organized in two major parts: a library section
(contained
procedures (OCL - operation control language), source, object (executable
programs)
and relocatable object (generally system-provided to be be link edited into
programs);
and, a file section for user data, including a VTOC (volume table of
contents).

System/3 was the ancestor of System/32, System/34, System/38, System/36,
and AS/400, in that order.  System/32, System/34 and System/36 all were
System/3 derivatives.  System/38 and AS/400, of course, were and are
completely different and far more advanced than the others.

That's the way it was (as best I can remember) and is (as I best can
understand).  Any errors are certainly my own!

I would have to agree that "System/3 code" is very old indeed!  It certainly
wasn't as bad as walking 5 miles to and from school in the snow, uphill both
ways,
or programming on card-only machines, but compared to today's programming
environment, it was pretty primitive.

Programmers whose experience only goes back as far as AS/400 (announced
1988)
or even System/38 (announced 1978, shipped 1980), often don't know how good
things are.


Mel Rothman


booth@MartinVT.com wrote:
>
> System 3 machines were mostly card machines.  They used 96  column punched
> cards to store data.  RPG responded to those needs, not to magnetic media.
>  You designed with record formats, matching records, level breaks, and
> left-side indicators.  there were no parms, no calls, no *INZSR, no
> externally described files, and no logical files.  you sorted the records
> to establish order for reports.   Source code was numbered in the first 5
> columns so you could put your card stack back in order when you dropped
> it.
>
>
> _______________________
> Booth Martin
> Booth@MartinVT.com
> http://www.MartinVT.com
> _______________________
>
> Lisa.Abney@universalflavors.com
> Sent by: owner-rpg400-l@midrange.com
> 09/26/2000 08:42 PM
> Please respond to RPG400-L
>
>
>         To:     RPG400-L@midrange.com
>         cc:
>         Subject:        System 3 Code
>
> Just a curiousity question ...
>
> A contractor recently described some very old code as "System 3" code. Any
> idea
> what this is?  RPG II is as far back as I go!  I do know it's running on
> an
> AS400 in 36 mode.
>
> +---
+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---

+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---