The Unix Heritage Society mailing list
 help / color / mirror / Atom feed
From: "John P. Linderman" <jpl.jpl@gmail.com>
To: Arthur Krewat <krewat@kilonet.net>
Cc: The Unix Heritage Society <tuhs@minnie.tuhs.org>
Subject: Re: [TUHS] Bell Labs data center in 1969/70
Date: Mon, 18 Mar 2019 11:04:31 -0400	[thread overview]
Message-ID: <CAC0cEp-8c1Kp1SiBR8JXrwXvW-ae7jec588zGau+D=eWR1SnoQ@mail.gmail.com> (raw)
In-Reply-To: <e4662baa-e275-905a-bc48-b9e619f4f865@kilonet.net>

[-- Attachment #1: Type: text/plain, Size: 7543 bytes --]

I can add an intolerable amount of detail to this story. I joined Business
Information Systems at the Labs in 1973, after completing my PhD at MIT. I
didn't like doing research: I wanted to write programs, not papers. (If I
had known what Bell Labs Research was about, I might well have applied
there). A small supervisory group, maybe 4 or 5 people, headed by someone
who had participated in the California test, was looking into
re-implementing the project. I don't know why the original didn't fly...
perhaps the way our effort died offers some insight.

"Information (411)" was a free service growing at an unsustainable rate.
The name was itself part of the problem. People would dial 411 to find out
about the weather. Changing the name to "Directory Assistance" helped. Just
talking about charging for calls helped even more. It cost AT&T nothing to
talk about it. But it was also clear that free paper-based search was
unsustainable. Operators got a daily "Frequently Called Numbers List
(FCNL)", things like hospitals and major retailers and IRS (in season).
Then there were the versions of the customer directories. These were
grouped by location, for which reason calls always started with "What City,
Please". If you didn't know the city, you were largely out of luck.
Operators couldn't plow through a stack of directories. The daily addenda
always struck me as bizarre. Operators were rated on call completion time,
so there wasn't much chance they'd take the time to check the addenda first
if the number showed up in the main directory. They probably served a
useful purpose for new numbers, which would be a logical reason for calling
411 to begin with. Typical calls were broken into three parts, getting the
information from the caller, searching for the number, and reporting the
number to the customer ("Oh, just a minute, I need to get a pencil.") Each
took about 10 seconds. Every second we could shave off a call was projected
to save the Bell System about a million dollars a year (back when 1 million
dollars was real money).

We had access to the directory information from the California trial, so we
did some analysis to suggest that 3 characters of Surname, plus 3 more of
City or of Street name, could be combined to produce a manageable list of
candidate numbers, and would shave several seconds off the search time.
(The approach would have mitigated the need to know City, and would have
consulted up-to-date information, which would have been regenerated daily).
A second phase, automated audio response to the customer, would have
eliminated the final phase of the calls. (It would also have eliminated
what little satisfaction the operators got from the job, which was already
well hated). It was easy to convince our department head that the project
was worth pursuing, and our director was also an easy convert. When we went
to our executive director, the final level of approval we needed to go
ahead, we expected an easy sell. Surprise! He was in charge of keeping
track of customer equipment, a huge and complicated effort. He told us, "In
Alabama, where I come from, we only hunt one rabbit at a time." He also
informed us that "micro-fiche had stolen the market". Our executive
director executed the fastest 180 I had ever seen, declaring that maybe the
project wasn't worth pursuing. I instantly lost all respect for him.

Two other notes. The executive director realized that by eliminating paper
directories, micro-fiche would save "the cost of glue for the binders". (I
and he were unaware that they were loose leaf, thanks, Ralph). He told my
supervisor to conduct a study of the cost of glue savings. My supervisor
(wisely) assigned the study to someone else in the group, knowing that I
would have walked before I wasted time on that. The executive director was
eventually promoted to "Vice President of Electronic Systems" at Western
Electric, possibly in recognition of his keen insight into the superiority
of micro-fiche over computers. With management like my director and
executive director, it's not too surprising that the original California
project folded.

On Sun, Mar 17, 2019 at 3:47 PM Arthur Krewat <krewat@kilonet.net> wrote:

> This kind of telephone history always get my "phreak" up ;)
>
>
> On 3/17/2019 2:52 PM, Ralph Corderoy wrote:
> > Hi Steve,
> >
> >> For a long time, California was viewed as hostile to phone companies,
> >> or at least AT&T, and I remember clearly people saying that Bell Labs
> >> would never have a location in CA as a result.
> > Here's what Larry Luckham told me in a private email that he's since
> > said could be copied to the list.
> >
> > Larry wrote:
> >> Of the thousands of web pages that I have posted the one of the Bell
> >> Labs photos is the one that generates a dozen queries a year.  Had no
> >> idea that would be the case when I posted it.  The photos are also the
> >> most ripped off and reposted of anything I've ever done.  But, to your
> >> points.
> >>
> >> The facility I set up in Oakland was temporary and for a specific
> >> experiment that ran for roughly 4 years.  You may recall that
> >> beginning in the mid 60's the Bell System was experiencing a huge and
> >> unpredicted demand for 411, information operator services.  The lead
> >> time to provide the trunking and other facilities for 411 operations
> >> was something like 25 years.  The public facing response was the "$55
> >> million dollar phone call" ad campaign intended to point customers
> >> back to printed directories.  The inward facing response was to figure
> >> out a way to handle each request for service faster so that the
> >> existing trunking and other facilities could meet the growing demand.
> >>
> >> At that time information operators relied on printed directories much
> >> the same as the customer's printed directory, except that theirs were
> >> loose leaf, reprinted monthly, and supplemented with a yellow daily
> >> addendum.  They were also printed in a larger format to make reading
> >> easier.  A division of the Labs called Business Information Systems
> >> Corp.  out of the Raritan River Center was tasked with the project and
> >> given a very short timeline.  A computer database and electronic
> >> display terminals driven by a very powerful search engine was the
> >> result.  Special operator terminals were designed and built by Western
> >> Electric.  The search engine was contracted out to Computer Corp. of
> >> America (CCA) which had been founded by some guys from Minsky's AI lab
> >> at MIT.  Then the idea was to try it out in a live environment.
> >> The San Francisco Bay Area was selected as reasonably representative
> >> and that's where I came in.  I was already managing the data center at
> >> the local Bell company, Pacific Telephone and Telegraph,
> >> San Francisco, so I was recruited to make it happen.  I built the
> >> mainframe data center, PT&T provided space in an information operating
> >> room a few blocks away and CCA came onsite to do the programming.
> >>
> >> The testing ran roughly 4 years.  I had moved on before it ended, but
> >> it was successful and was implanted, at least to some degree, but this
> >> shop was dismantled and everyone went home.  Then technology did what
> >> it always does, it ran over everything and changed the world.
> >> Along came the PC, the Internet, smart phones, etc.
> >>
> >> It's been a very long time and I'm sure I've forgotten, or
> >> misremembered stuff, but that's kind of what I remember.
> >> Hope it sheds some light.
>
>

[-- Attachment #2: Type: text/html, Size: 8933 bytes --]

  reply	other threads:[~2019-03-18 15:05 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-03-13 13:25 Doug McIlroy
2019-03-14  8:10 ` Rob Pike
2019-03-15  4:03   ` Kevin Bowling
2019-03-16 21:30   ` Steve Johnson
2019-03-17 18:52     ` Ralph Corderoy
2019-03-17 19:39       ` Arthur Krewat
2019-03-18 15:04         ` John P. Linderman [this message]
  -- strict thread matches above, loose matches on Subject: below --
2019-03-12 16:02 Dan Cross
2019-03-12 17:14 ` Clem Cole
2019-03-12 17:17   ` Jon Steinhart
2019-03-12 17:29     ` Clem Cole
2019-03-12 17:31       ` Jon Steinhart
2019-03-12 17:42   ` Paul Winalski
2019-03-13  0:17   ` Greg 'groggy' Lehey
2019-03-13  1:37   ` Dave Horsfall
2019-03-13  8:41     ` Peter Jeremy
2019-03-14 22:12       ` Al Kossow

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAC0cEp-8c1Kp1SiBR8JXrwXvW-ae7jec588zGau+D=eWR1SnoQ@mail.gmail.com' \
    --to=jpl.jpl@gmail.com \
    --cc=krewat@kilonet.net \
    --cc=tuhs@minnie.tuhs.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).