IBM computer system too complex to learn?

[Ben Pfaff]

I was reading _The Soul of a New Machine_ last night. One
passage referred to an IBM computer system project that was
canceled because, according to the book, it was to be so complex
that it would take more than one lifetime of study for an
engineer to understand it. What computer system was this, and
was the claim about its complexity mere hyperbole? (Why would
anyone ever contemplate a design that complex to begin with?)

[John Levine]

Sounds like the legendary Future Systems project. Bits of
it were recycled into S/38, but I've never seen a good
overall description of it.

Perhaps Lynn Wheeler has better sources.



--
Regards,
John Levine, jo...@iecc.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. http://jl.ly

[hanc...@bbs.cpcn.com]

On Nov 2, 2:37 pm, John Levine <jo...@iecc.com> wrote:

> Sounds like the legendary Future Systems project.  Bits of
> it were recycled into S/38, but I've never seen a good
> overall description of it.

The previously mentioned "IBM's 360 and Early 370 Systems" by Pugh et
al spends several pages (pg 538-) covering Future System--what it was
intended to do and why it was cancelled. In essence it called for a
"single level store" in which users would not have to concern
themsevles with the physical location of information in the system.
Information would be addressed in the same manner regardless if it was
stored in main memory, disk, or tape. It would have an all new
architecture.

The problem was providing it within an acceptable price and
performance, even with expanding technology of the 1970s.

Some hardware developments for FS were incorporated in the new S/370
line. Some FS concepts were used in the S/3x line. Most systems
designs, microcode, and softwrae were largely discarded, making FS the
biggest failure in IBM history.

[Charlie Gibbs]

In article <87liej7...@blp.benpfaff.org>, b...@cs.stanford.edu

(Ben Pfaff) writes:

> (Why would anyone ever contemplate a design that complex to begin
> with?)

Because too many people are so impressed with complexity that they
forget that sophistication is a totally different thing. The concept
of "elegant design" is totally beyond them - which makes them eager
victims for whatever piece of technological crap is foisted off on
them this week.

--
/~\ cgi...@kltpzyxm.invalid (Charlie Gibbs)
\ / I'm really at ac.dekanfrus if you read it the right way.
X Top-posted messages will probably be ignored. See RFC1855.
/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign!

[hanc...@bbs.cpcn.com]

On Nov 2, 7:14 pm, "Charlie Gibbs" <cgi...@kltpzyxm.invalid> wrote:
> In article <87liej7ulu....@blp.benpfaff.org>, b...@cs.stanford.edu

>
> (Ben Pfaff) writes:
> > (Why would anyone ever contemplate a design that complex to begin
> > with?)
>
> Because too many people are so impressed with complexity that they
> forget that sophistication is a totally different thing.  The concept
> of "elegant design" is totally beyond them - which makes them eager
> victims for whatever piece of technological crap is foisted off on
> them this week.

Compare the architecture of the S/360 vs. the 1401. Is S/360 too
complex?

[Charlie Gibbs]

In article
<8565bfb0-11ea-482b...@k21g2000vbj.googlegroups.com>,

Good one. The 360 architecture was pretty clean in many ways.
The enhancements for the 370 started to look kludgy, though.

As for OS/360, that's a totally different can of worms...

[Shmuel Metz]

In <87liej7...@blp.benpfaff.org>, on 11/02/2012

I suspect that he was referring to Future Systems, but the issue was
performance rather than complexity.

I nearly split a gut laughing, because the Hebrew word for "nothing"
is "efes", quite close to the English spelling out of FS.

--
Shmuel (Seymour J.) Metz, SysProg and JOAT <http://patriot.net/~shmuel>

Unsolicited bulk E-mail subject to legal action. I reserve the
right to publicly post or ridicule any abusive E-mail. Reply to
domain Patriot dot net user shmuel+news to contact me. Do not
reply to spam...@library.lspace.org

[hanc...@bbs.cpcn.com]

On Nov 3, 3:46 pm, "Charlie Gibbs" <cgi...@kltpzyxm.invalid> wrote:

> As for OS/360, that's a totally different can of worms...

The very early versions of OS/360 were horrible, as the histories and
authors will say. It was also ahead of its time, trying to do too
much too soon.

But later data centers had machines with large memories and many
peripherals. Such machines were used to handle multi-programming,
Keeping multiple application programs from encroaching upon each
other, working efficiently with memory, and handling tape mounts,
printer forms, disk mounts, and other special devices all took a
sophisticated operating system.

From reading the histories, I get the impression some customers were
still very anxious to get their System/360 even before the operating
system was even ready, and certainly well before the bugs were worked
out. I believe some early customers got a special quick & dirty
version just to get them going.


In my humble opinion, many of System/360's hardware and software
development problems were the result of Tom Watson Jr's arbitrarily
(his words) split of the business into two divisions (GPD and DPD)
based on machine size. There was an enormous rivalry and lack of
cooperation between Poughkeepsie and Endicott which I think could've
been avoided or at least toned down by better management than Watson
Jr offered.

Also, S/360 was a worldwide effort. The logistical capability to keep
"everyone on the same page" was rather limited in the early 1960s.

[Alexander Schreiber]

Shmuel Metz <spam...@library.lspace.org.invalid> wrote:
> In <87liej7...@blp.benpfaff.org>, on 11/02/2012
> at 11:00 AM, Ben Pfaff <b...@cs.stanford.edu> said:
>
>>I was reading _The Soul of a New Machine_ last night. One passage
>>referred to an IBM computer system project that was canceled because,
>>according to the book, it was to be so complex that it would take
>>more than one lifetime of study for an engineer to understand it.
>>What computer system was this, and was the claim about its complexity
>>mere hyperbole? (Why would anyone ever contemplate a design that
>>complex to begin with?)
>
> I suspect that he was referring to Future Systems, but the issue was
> performance rather than complexity.
>
> I nearly split a gut laughing, because the Hebrew word for "nothing"
> is "efes", quite close to the English spelling out of FS.

Interesting. "Efes" is also the name of a Turkish Brewery Group, with
"Efes Pilsener" as their flagship product. The beer is quite drinkable.
Not that we did have that much choice in Istanbul ...

Kind regards,
Alex.
--
"Opportunity is missed by most people because it is dressed in overalls and
looks like work." -- Thomas A. Edison

[Shmuel Metz]

In <1193.725T2...@kltpzyxm.invalid>, on 11/03/2012

at 11:29 AM, "Charlie Gibbs" <cgi...@kltpzyxm.invalid> said:

>In article
><8565bfb0-11ea-482b...@k21g2000vbj.googlegroups.com>,
>hanc...@bbs.cpcn.com (hancock4) writes:

>> Compare the architecture of the S/360 vs. the 1401. Is S/360 too
>> complex?

The S/360 was less complex than other machines of its era.

[Quadibloc]

On Nov 4, 6:19 am, Shmuel (Seymour J.) Metz

<spamt...@library.lspace.org.invalid> wrote:

> The S/360 was less complex than other machines of its era.

True enough if you're thinking of user-mode programming in assembler.
The subroutine calling convention would have been more complicated
compared to simple machines like an SDS 940 or a Honeywell 316, but
that would be minor. And other machines could do either binary
arithmetic or decimal arithmetic - not both.

Storage protection was another feature other machines didn't have in
as sophisticated a form.

But when it came to I/O, the 360 with its channels was a whole other
world of complexity compared to almost any other computer of that era.

John Savard

[Stan Dandy Liver]

On Sat, 03 Nov 2012 23:20:26 -0000, Shmuel Metz
<spam...@library.lspace.org.invalid> wrote:

> In <87liej7...@blp.benpfaff.org>, on 11/02/2012
> at 11:00 AM, Ben Pfaff <b...@cs.stanford.edu> said:
>
>> I was reading _The Soul of a New Machine_ last night. One passage
>> referred to an IBM computer system project that was canceled because,
>> according to the book, it was to be so complex that it would take
>> more than one lifetime of study for an engineer to understand it.
>> What computer system was this, and was the claim about its complexity
>> mere hyperbole? (Why would anyone ever contemplate a design that
>> complex to begin with?)
>
> I suspect that he was referring to Future Systems, but the issue was
> performance rather than complexity.
>
> I nearly split a gut laughing, because the Hebrew word for "nothing"
> is "efes", quite close to the English spelling out of FS.

> It's also a beer brewed in Turkey (and the UK under license).
http://en.wikipedia.org/wiki/Efes_Beverage_Group



--
[dash dash space newline 4line sig]

Money/Life question

[Stan Dandy Liver]

On Sun, 04 Nov 2012 00:12:17 -0000, Alexander Schreiber
<a...@usenet.thangorodrim.de> wrote:

> Shmuel Metz <spam...@library.lspace.org.invalid> wrote:
[]

>> I nearly split a gut laughing, because the Hebrew word for "nothing"
>> is "efes", quite close to the English spelling out of FS.
>
> Interesting. "Efes" is also the name of a Turkish Brewery Group, with
> "Efes Pilsener" as their flagship product. The beer is quite drinkable.
> Not that we did have that much choice in Istanbul ...
>
> Kind regards,
> Alex.

Bah, beer rays!