Examining a technology sample kit: IBM components from 1948 to 1986 | Hacker News Reader

Examining a technology sample kit: IBM components from 1948 to 1986

righto.com

75 pointszdw20 comments

This is from the era when IBM had their own electronics packaging, Western Electric had their own electronics packaging, and everyone else had commercial packaging. Western Electric used grey rectangular cans, filled and sealed with a potting compound. They also liked inductors of all kinds, including saturable reactors as power amplifiers. They wanted components that would not fail.

By the way, does anyone know why it seems to be so difficult to print readable texts onto the packaging these days? I know orientation information can often be deduced from the orientation in the tape/reel, but is it really that hard to produce some more contrast?

Because they just laser etch and don't use paints anymore, it's faster.

The thing that always amazes me is that companies paid millions for these machines and got value from them and I have a device more powerful than any of them - and I use it to look at cat pictures and basic math.

I’m 50, so lived through it. But still wonder how companies operated, how did anything get done? Without computers and automation. Similar how did people learn things or know what is happing in world before the internet? I’m seconds away from any known fact.

Lots and lots of manual processes, and lots and lots of "man"power (in quotes because often it was literally acres and acres of women doing computations by hand - a "computer" was a person, often a woman, doing computations).

> But still wonder how companies operated, how did anything get done?

At the small (13 people) software house I joined in 1993, our information management was physical filing cabinets. Key project plans were printed and stored. A visit from the ISO auditor always involved diving into the filing cabinets. Source code wasn't printed (phew!) but the project records would explain where on the servers the software could be found.

Repo-based software configuration management as we would recognise it today was virtually non-existent. Projects tracked releases to customers using configuration notes that (in theory) allowed you to identify the resources required and how to recreate / build the delivered system. We had nothing like CI/CD - releases were few and intermittent.

We were disciplined and the directors were quality focussed, so the system actually worked reasonably well. The company had evolved some very good checklists for projects that were used at start-up to build the information framework for a project of a given type and size (we had three or four different generic types of project).

When I left ten years later the basic processes were still broadly identical, although there was less emphasis on the physical filing.

[Edit]

> Similar how did people learn things

For software, from product manuals and usenet. For domain issues, from experts and sometimes books. E.g. when we built in ~1995 a system to support analysis of satellite constellations, the team concerned actually bought and read books on orbital mechanics and spherical geometry. I was a space nerd and still remembering being pleased at explaining to the programmers that the modelled satellites were changing speed as they orbited because the programmers had correctly implemented Kepler's laws.

Armies of people and layers of management. One thing though, there was a natural bottleneck that was in tune with human needs. Today computerized sides of the system (say automatic complaint system on internet) are throwing too much requests that end up in now crippled human services operating half on legacy paper paradigms and half with bad software.

ps: chances are some big fund blockchain project will evaporate all of this in a few moments. It's the kind of thing large enough and capable enough to wipe legacy.

This is an idea that spins in my head regularly since around the iphone 5 say.

At the same time I found a thread where a dude made some napkin calc saying that desktop around the core 2 era were in the ballpark of cray-1 or cray-2 crunch power (on peak IPC only).

The computing revolution feels extremely dull to me now. 2010s internet is not helping.

And real jobs either.. my last gig at a court room, the information system (let's call it that) was so ridiculously bad that to factually improve the justice system you'd only need a well programmed vba program running on a pentium-II.

Progress is only so for investors, it's highly unevenly distributed and where it could matter it's not happening.

Author here for your vintage IBM questions :-)

Do you know what SMS-card based computers are still operational? Or if anyone has done anything with the spare cards (besides scrapping them, anyway) or made replica systems out of similar modular cards?

The Computer History Museum in Mountain View has two IBM 1401 computers operational. The 1401 is based on SMS cards. The museum had an IBM 1620 working for a while, but it is not currently operational. There are also some operational IBM 1130 computers around (http://ibm1130.org/today/).

Awesome, thanks!! Hope I get a chance to visit one day. I love that people are still keeping these giant piles of cards going.

btw love your blog :-)

Thank you for publishing that nice bit of history. I really enjoyed reading.

Great article! I tried to help my local university's computer museum set up a similar display but there is no comparison.

About "IBM is older than you might expect. It was founded (under the name CTR) in 1911 and produced punched card equipment for data processing, among other things.", while this is 100% true it is a little misleading since what happened in 1911 was a merger of several older companies:

https://en.wikipedia.org/wiki/Computing-Tabulating-Recording...

It would be like saying Unisys was founded in 1986 without pointing out that Burroughs was from 1886 and Sperry from 1910.

Thanks. I don't want to get too much into the corporate history but I updated the article to mention that it was formed through merger.

It seems like there was a vacuum going with IBM between 1933 and 1946?

IBM introduced a few products during that time period, but there wasn't a whole lot of development going on during this period due to the depression and then World War II. During WWII, IBM produced a lot of bombsights and rifles, strangely enough.

One interesting thing about the Depression is that IBM weathered it better than most companies because it rented equipment rather than selling it. Companies weren't doing any new investment, but they mostly kept renting the IBM equipment they had. Plus, IBM was selling millions of punch cards per day.

They also supplied machines to Nazi Germany during that period, which helped run the holocaust.

Which they really don't like people talking about.

> By the 1930s, IBM was producing complex electromechanical accounting machines for data processing, controlled by plugboards and relays.

And, crucially, supplying them to Nazi Germany to help run the holocaust.

Reply on news.ycombinator.com