The Story of PDP-1 (1998) | Hacker News Reader

The Story of PDP-1 (1998)

gordonbell.azurewebsites.net

94 pointstosh16 comments

It's also a bit of a tragic story. The PDP-1 was designed by Ben Gurley, an engineer at DEC. He took just 3.5 months to build the machine. Tragically, Gurley was later murdered in 1963 while eating dinner with his family, when a co-worker fired a rifle through the window of his house. The co-worker was later admitted to a hospital of the criminally insane, where he died.

Wow. Here is a story about that story:

https://www.informationweek.com/desktop/updikes-passing-reca...

Maybe interesting: A log of implementing a video game (Computer Space) for the PDP-1 over a month, a project for RetroChallenge 2016/10:

https://www.masswerk.at/rc2016/10/

(The experience was great and the instruction set is a joy to use. The single time, I felt somewhat hindered, was due to the limit of just up to 3 characters for any names to be used for assembler variables, notably not the PDP-1's fault. BTW, the games runs at 60 frames a second.)

I was at the Computer History Museum in Mountain View yesterday and got to play Spacewar!, the game that inspired Computer Space (amongst other things), on their restored PDP-1. Stephen Russell and Peter Samson were on hand to chat about it.

Sounds like something they do at the museum regularly, worth a visit

Here's another interesting read in a similar vein... "The Soul of a New Machine", a Pulitzer Prize winning book about the creation of a new computer at Data General Corporation.

https://en.wikipedia.org/wiki/The_Soul_of_a_New_Machine

This is an incredible book, largely worth a read or two.

My first Data Processing job (pre-IT) was an operator on DEC PDP-11/45. Later we upgraded to a DEC PDP-11/70.

In Australia, we had 10% of the world's Pick systems. I have use Wang OIS, Data General RDOS, IBM System AS400 and a few others as well.

I threw away an original DEC Flea Circus (Field Service) leather toolkit as the Australian Computer Museum does not have any more room.

Core memory from this era is elegant and beautiful.

https://en.wikipedia.org/wiki/Magnetic-core_memory

We used to have a big frame of it hanging in our office. It was quaint, but I didn't get why, in 1995, why you would hang obsolete junk.

A year or two ago I started to feel a little nostalgic and looked for it on eBay. That stuff is probably worth more now than when it was new tech.

I once heard a story about how Brazil used to manufacture magnetic-core memory. When production stopped, lots of people lost their jobs and somehow this shaped Brazil's computer policy (they basically closed the market to foreign companies). But I haven't been able to find any proof of this.

The reserved market policy was created in 1977 specifically to keep IBM from launching their minicomputer, which would have preemptively killed off the effort by the government owned company Cobra (which would only launch their 530 mini in 1980)

The relation with jobs was that the policy had full support from academics who were upset that foreign companies would hire their engineering students only for sales, with all design work done in their home countries. I am not aware of any big stink about factory line workers losing their jobs.

Those were he days: “the computer is available in word lengths of 18, 24, 30, or 36 bits”. You can do multiplication using an instruction or via subroutine (why?)

Woah, I never read that before. The PDP-2 never happened and the PDP-3 was a 36 bit PDP-1 design only built once by a customer of DEC. Thanks for bringing that to my attention.

"If you're not playing with 36 bits, you're not playing with a full DEC!" -Doug "DIGEX" Humphrey (Pro-PDPer and Anti-VAXer), teasing the 32 bit VAX weenies at DECUS.

Early PDP-1's had the MUS (MUltiply Step) operation, which had to be applied repeatedly in order to perform multiplication. Singe-step multipy (MUL) was first available as an extension, later it became the default.

More on the multiply and divide subroutines at https://www.masswerk.at/spacewar/inside/insidespacewar-pt6-g... – automating these subroutines in hardware meant adding a special clock, which would take over control in order to perform the according steps.

Reply on news.ycombinator.com