"Has it got the Cobalt?" (story)

[chopnhack]

If you haven't manually entered the bootloader to a PDP-11 then you are not a real programmer.

LOL, here's to you sir!


Owen: "What I really need is a droid who understands the binary language of moisture vaporators."
C3PO: "Vaporators! Sir, my first job was programming binary loadlifters very similar to your vaporators in most respects!"

 

[chopnhack]

And custom languages like PICAXE BASIC - yuck!

What's wrong with PICAXE Basic? It looked approachable from the examples I saw.

It's not the GOTO function is it?

I think it's important to get involved in programming - programming in general - when you're fairly young, but I don't think it makes much difference which language you start with. Eventually, it's probably a good idea to know one or two from each category.

It teaches a discipline of logic, I wish I had more exposure to it. I had limited hands on experience in elementary and again in high school. When young I also had some tinkering in basic on a commodore 64.

At 37 I don't think its too late

 

[BobK]

You might notice that the instruction word is broken up into groups of 3 with the high bit a singleton. There is a good reason for that. The instructions are written in octal (base 8, digits 0-7).

The instruction word was all based on 5 fields:

1-bit 0 for byte 1 for word operation
3 bits: opcode
3 bits: address mode for operand 1
3 bits: register for operand 1
3 bits: address mode for operand 2
3 bits: register for operand 2

So, after learning 7 opcodes (that last one was an escape for less often used instructions), and 8 address modes, one could encode an instruction into octal or decode it in one's head.

This was the most elegant machine ever invented.

Bob

 

[Supercap2F]

If you haven't manually entered the bootloader to a PDP-11 then you are not a real programmer.

See the 16 switches. That is one instruction word, you flip them up for 1 down for zero, then you toggle another switch to load that instruction into memory and advance to the next address. The bootloader was on the order of 20-30 words depending on what device you were loading from. You would typically load from DECTape (which small wide magnetic tape) or later on a disc.


Bob

View attachment 14526

Hey, that's pretty neat! Is that how they programed the first computer?
Dan

 

[KJ6EAD]

If you haven't manually entered the bootloader to a PDP-11 then you are not a real programmer.

See the 16 switches. That is one instruction word, you flip them up for 1 down for zero, then you toggle another switch to load that instruction into memory and advance to the next address. The bootloader was on the order of 20-30 words depending on what device you were loading from. You would typically load from DECTape (which small wide magnetic tape) or later on a disc.


Bob

View attachment 14526

I'm sad to say I got the Cobalt/COBOL alliteration without cheating.

Bob, the PDP-11 was fancy new stuff when I loaded the bootloader on the Data General Eclipse at my high school...then loaded the language translator via octal paper tape!

Wasn't Tracy Kidder's book, "The Soul of a New Machine" about the development of the PDP-11?

 

[(*steve*)]

This was the most elegant machine ever invented.

And spawned countless other processors (not to mention the C language). In fact C owes so much to PDP-11 assembler that it has quite justifiably called "smart assembler".

Unfortunately, like Betamax, the PDP-11 inspired instruction sets were lost beneath a wave of VHS style 4040 -> 8080 -> 8086 -> etc. lineage of processors with their nasty and segmented architecture that has only very recently extricated its claws from our code.

 

[gorgon]

When i started working as a young service engineer back in -75, the serious programmers in the company wrote all their cobol programs on punched card programming sheets by hand. Those were then transferred to punched cards by them, or a secretary, and read by the big IBM computers we had. Many of those COBOL programmers was 'old' already then
My first programming beside the BASIC done at school, was doing capacitor calculations in Fortran as an exercise. I had to run them at night because they took a lot of capacity of the CPU . The reason was that I mostly used CPU time, and little IO capacity in my programs, draining the cpu capacity from the normal business batch runs.
The JCL=Job Control Language was a pain, since you needed that to run your programs in the batch system.

Those were the days

Regarding the Bootstrap loading above. One of the jobs I had at that time was to supervise and old TTY server running a dustributed net of BASIC callup TTY terminals, running 110baud, and processing BASIC programs for remote customers. Every time there was a thunderstorm or powercut, we had to restart the old system. First we entered around 30 words of switch flipping commands, then we loaded the operating system from papertape, and then it was up and running after some commands, on an old magnetic drum storage and magnetic core memory. I remember I admired the console terminal, a mechanical wonder, no solid state there. I don't remember the brand of the machine, but it was old already then.

 

[BobK]

Wasn't Tracy Kidder's book, "The Soul of a New Machine" about the development of the PDP-11?

Actually, no, it was about the DG Eclipse!

Data General was founded after there were two designs for the next generation (after the PDP-8) at Digital. One, designed by Gordon Bell was the PDP-11, the other was degined by Edson DeCasto who left Digital to found DG and it became the NOVA.

Bob

 

[BobK]

At the risk of self-aggrandizement and name-dropping, I would like to say a few words about Gordon Bell. He was one of a kind. Designer of the PDP-11, he became the senior vice president and chief technologist at Digital (#2 man in the corporation).

I was hired by Digital to speed up the Cobalt, er, COBOL runtime. My second week of work, I was coming in late one day, and an associate grabbed me and said, "Hurry, Gordon Bell is looking for you". My response was "Who?". That was my first meeting with Gordon. From his high position, he made a point of communicating directly with the "worker bees", which made him an outstanding manager.

He was also famous for sending out emails widely with ideas and asking for comments. Some of them, however, were purposely stupid ideas. Anyone who responded in favor of those was scratched from his credibility list.

When I applied for a job at Microsoft, where Gordon was then working, I gave him as a reference. The HR person was very skeptical and asked if he would actually know me. I confidently said yes. They contacted him, and I got the job.

Bob

 

[KJ6EAD]

Thanks for sharing that excellent bit of personal work history. You were there participating in a pivotal era in computer development.