Bell System Technical Journal, 1922-1983 | Hacker News Reader

Bell System Technical Journal, 1922-1983

bstj.bell-labs.com

112 pointsssclafani20 comments

This one from 1978 is the UNIX issue:

http://bstj.bell-labs.com/oldfiles/year.1978/BSTJ.1978.5706-...

I chose to read parts of it for a class a few years ago and found it surprisingly modern-sounding.

The epigraph on the foreword: "Intelligence...is the faculty of making artificial objects, especially tools to make tools. — Bergson"

From the preface: "Because computer science is still in an early stage of development, no well-formulated theoretical sructure exists around which problems can be defined and results organized. 'Elegance' is of prime importance, but is not easily defined or described. Reliability and maintainability are important, but they also are neither precisely defined nor easily measured."

On the inside back cover: "This issue of the Bell System Technical Journal was composed, including all tabular and displayed material and final page makeup, using the document preparation software described on pages 2115-2135. It was phototypeset using the troff program, which was written by the late Joseph F. Ossanna, Jr."

The list of contributors includes one woman, Helen D. Rovegno, but Google doesn't say anything about her.

My grandfather was a technical editor for the BSTJ and other Bell Labs publications for most of his life, but unfortunately he never really talked about it in detail to me.

The telecom nerd in me just EXPLODED.

It's great that they've finally put all of these online; no more inter-library loans or having to troll eBay to look up an article.

My favorite: the September 1964 special issue detailing the 1ESS electronic switch.

http://bstj.bell-labs.com/oldfiles/year.1964/BSTJ.1964.4305....

http://en.wikipedia.org/wiki/1ESS

"A Mathematical Theory of Communication" Shannon, C.E.

http://bstj.bell-labs.com/BSTJ/images/Vol27/bstj27-3-379.pdf

A classic work.

Be sure to get part 2 in the next issue at http://bstj.bell-labs.com/BSTJ/images/Vol27/bstj27-4-623.pdf

Plus: Shannon, Claude E., The Synthesis of Two Terminal Switching Circuits, Bell System Technical Journal, 28: 1. January 1949, http://bstj.bell-labs.com/BSTJ/images/Vol28/bstj28-1-59.pdf

Ah, fond memories. I recall a bus ride through snowed in Denver in 1984, to get a copy of the original blue box article referenced in phone phreak lore at the downtown library, just to find that these pages had long been removed from the journal copy.

Here the article I believe in question: "Pushbutton Calling with a Two-Group Voice-Frequency Code, Schenker, L.", http://bstj.bell-labs.com/oldfiles/year.1960/BSTJ.1960.3901.....

There is an older article covering dial tones (DTMF): "In-Band Single-Frequency Signaling, Weaver, A.; Newell, N.A.", found in http://bstj.bell-labs.com/oldfiles/year.1954/BSTJ.1954.3306....

Bell did science for science's sake, not just for the telecom industry, and that's what was really unique about them, I think. Specifically, I remember reading an article a while back that talked about Bell Labs being the leading employer of Physicists in private industry at one time. There's not a lot of companies like that anymore, except Google (vicariously) and places like IBM's Watson Research Center.

A lot of places use to invest heavily in research. The modern desktop came from the research lab of a printer company! These days the big researchers are probably IBM and Microsoft (SPJ, 'nuff said). Sun was also very big (they did all kinds of very interesting work, like the Self language for example) but it looks like that's over now.

Interesting blog post regarding this: http://blogs.sciencemag.org/sciencecareers/2009/08/we-miss-y...

Slight Tangent: Before I got my BS in Applied Physics, I did a lot of research in experimental astroparticle physics. I learned a lot of interdisciplinary stuff, and I thought those combinations would easily land me at least a ~40k job doing something (programming, electronics stuff, whatever). Turned out I was wrong, and I couldn't find a job anywhere in my hometown (even though I could run circles around any of the new CS grads in most languages) I learned a hard lesson that I thought would haunt me for the next decade: People don't want to hire scientists, they only want engineers. I got a shitty $15/hr part time job as a test tech, got sick of that and then realized I couldn't get into grad school for financial reasons. I had a bit of money so I applied only to physics jobs at high profile schools or labs across the country (partially motivated by this site, actually). Within a week I got an email back from Stanford (SLAC) to do an interview for a developer. It turned out that I had all the right kinds of experience, and I just started last Thursday at what can only be described as the perfect job for me right now.

I think most companies these days just overlook the value of research, basic science, and scientists in favor of marketing, products, and engineers. The risk that research doesn't lead to a marketable product is too high for almost every company, and that's a shame.

I think it is more the freerider problem; Bell didn't suffer it significantly because they held a huge monopoly, and had few other ways to expand, even if others would benefit freely from their research.

Science and research take time. And there is no incentive for companies to take a long term outlook on things. Everything is measured quarterly.

It occurs to me that if you have a big company, probably the best way to "spend money for the benefit of human kind" is supporting a research lab in this style. I think this beats even having a foundation like Bill Gates. (By the way, Gates does both things...)

Remember that Bell had the advantage of the mother of all monopolies. When Bell Lab's was at its height, nothing could touch Bell. They had a monopoly that made Microsoft in the late 90s/early 00s look flimsy. I mean, they pretty much ran out of room for growth (beyond scaling with the overall growth of America).

I have to agree this is a telecom geek's dream. My father worked on a bunch of this stuff in the late 70s and 80s. I just sent him the link to geek out on.

This is a great source to build on top of the assorted 2600 articles from over the years.

I'd really like these in HTML or EPUB.

Quick math says 500 pages per journal by 382 journals gives 382,000 pages. The going rate on Mechanical Turk seems to be $1 / page for human OCR. All the decent hosted OCRs sites I ran tests on are about .15c per page.

Yikes! Out of my budget. :-(

However, http://code.google.com/apis/documents/docs/3.0/developers_gu... did a half-decent job on some random selections. Any Googlers want to (ab)use company resources?

If you could get the copyright holder to release them into the public domain, then they would be a great candidate for Distributed Proofreaders. 382kpgs would still take a while (although maybe not too much if it attracted new volunteers), but it would be by far the cheapest solution.

Am I the only one having difficulty loading these pages? Each article (.pdf) seems to be about 1.5 - 2.5 MB. But it's taking 3-5 minutes for anything to appear.

The lists appear immediately.

Does anyone know when this archive was released? The copyright at the bottom of the index page says 2010 so can we assume this is a recent, awesome event?

BTW: I Googled for some PR for this but didn't find any (at least, on the first page of results).

is there any centralized source for large lists of interesting documents like this? i'd love to know if there's a project gutenberg for juicy whitepapers.

Reply on news.ycombinator.com