The Internet, 1973

theoldwizard1 wrote:
... Mail was a big problem because different manufactures handled mail on their machines differently, including different address format

Yup, did as lot of coding back then so different mail systems could talk to each other. We used Western Union Easylink as the intermediary for transport.

Dave

I started on a friden 5610/flexiwriter. We had acct cards and inventory cards and a small canned package. Took a lot of rewrites and tape patches, but it ran a long time. Mid 70's bought a SingerFriden System 10(maybe). WE ended up with 3 10 meg HDD. I think it was some sort of IBM assembler and was 7 but to the best of my memory. Had model 50/51 printer and model 80 crt's. Life was good.

older_fossil wrote:
I got involved in 1969...

A Bunch of youngsters on here. :)

I was hired by IBM in 1960, their IBM 704 "scientific" computer used vacuum tubes with no parity check on 32 bit "words". When something failed, the only symptom was wrong results. Computers for business use had parity checks as wrong results would not work for that market.

The only thing related to connecting computers was a Transceiver that was a glorified key punch (for punched cards) that could transmit a punched card from one Transceiver to another via a phone line connection.

In 1961 they started delivering their first transistorized computers with discreet transistors, 3 or 4 soldered to a 3x5" circuit card that plugged into "mainframes" that contained hundreds of said circuit cards. All of which took up a large room and had less computing power than any smart phone of today.

I got involved in 1969 when the four node protoype for the ArpaNet was being put together. I was an undergrad student at UCSB and was involved with connecting the UCSB IBM-360 mainframe to the local Arpanet IMP. I also designed an interface for Digital Equipment Corporation's UNIBUS interface that was used in PDP-11 and VAX minicomputers. That interface was used by many universities and research facilities to get onto the ArpaNet. It was amazing to see the diagrams of the ArpaNet quickly grow in complexity until it became an unmanageable task.

My first paying gig in the computer field was using a PDP8 with 4K words of core. Programmed in assembler with paper tape on a Teletype ASR33 - 110 baud.

Fortran and punch cards - luxury!

2oldman wrote:
Amazing. I don't know what any of those acronyms mean.

I did not look at all of them but I do know what many of them mean

PDP-## PDP is the name of a computer company so that's a computer

(As an example)

I first got "online" in 1977.. though it was not ARPA net it was a closed network.. NO dial up modems.. All hard wired dedicated lines ASR 31 Teletype terminals for most of the net though I had a nice IBM high speed I could use as well and later a Teletype Dataspeed 40 and then a Racal-Milgo, then PC's with Racal-Milgo cards, then finally PC's with software, All Dedicated lines though, not "internet".

That's how long I've been doing this stuff.

What is even more amazing is that basic design managed to scale from about 0 to ...oh... 1 or 2 billion clients in about 30 years. Pretty good design work, I would say.

PDP 8L - programmed in binary by switches on the front... Fortran IV - punch cards... Now I can't run my phone... ST

Now they are mostly built in China and sold here.....

Thanks for the memories, the bad memories:E