4 ms·

Don’t forget BBN. The TCP/IP code in BSD was originally from them. Let’s call it a prototype.

by e40 2y ago

p_l 2y ago
BBN got a lot of the initial work, as well as built the initial infrastructure including routers. BSD got a contract to port TCP/IP to Unix because DoD found themselves facing that the main system they worked with TCP/IP on was discontinued. Thus a grant to port the stack to Unix and make it available as wide as possible.
- e40 2y ago
  I was on mobile and didn't want to type this out: I was there when Bill Joy, Sam Leffler, and Bob Henry (not Robert R Henry) were crowed around a huge HP terminal in the common area of the 5th floor of Evans Hall. I used the public/common terminals for my work and was sitting there one day when the trio walked in a were excitedly talking about the TCP/IP code from BBN. I have no idea what they were doing, but I think they were trying to build it.
  - p_l 2y ago
    Generally a lot of networking code evolved on the big 36bit machines. There were implementations for unix that predated BSD sockets, but AFAIK Berkeley got a grant to make an implementation that would be freely accessible when TOPS-20 future was indicated to be in jeopardy - or it might have been just because before AT&T being able to demand licensing fees, everyone was putting out clones of Unix on various hardware.
    - e40 2y ago
      I'm pretty sure Berkeley (or UC?) had a System V license and was allowed to redistribute code based on it in BSD. That became a point of contention later, when the *BSDs started to become popular. This is all very old stuff so I might have the details wrong. It was definitely a fun time to be at UCB. I remember getting the early Sun workstations and porting stuff to them. It was a really good port of BSD, but that Motorola chip was a problem for some things. I ported the Franz Lisp compiler (liszt) to it.

4 ms·

Don’t forget BBN. The TCP/IP code in BSD was originally from them. Let’s call it a prototype.

by e40 2y ago

Don’t forget BBN. The TCP/IP code in BSD was originally from them. Let’s call it a prototype.

p_l 2y ago
BBN got a lot of the initial work, as well as built the initial infrastructure including routers. BSD got a contract to port TCP/IP to Unix because DoD found themselves facing that the main system they worked with TCP/IP on was discontinued. Thus a grant to port the stack to Unix and make it available as wide as possible.
- e40 2y ago
  I was on mobile and didn't want to type this out: I was there when Bill Joy, Sam Leffler, and Bob Henry (not Robert R Henry) were crowed around a huge HP terminal in the common area of the 5th floor of Evans Hall. I used the public/common terminals for my work and was sitting there one day when the trio walked in a were excitedly talking about the TCP/IP code from BBN. I have no idea what they were doing, but I think they were trying to build it.
  - p_l 2y ago
    Generally a lot of networking code evolved on the big 36bit machines. There were implementations for unix that predated BSD sockets, but AFAIK Berkeley got a grant to make an implementation that would be freely accessible when TOPS-20 future was indicated to be in jeopardy - or it might have been just because before AT&T being able to demand licensing fees, everyone was putting out clones of Unix on various hardware.
    - e40 2y ago
      I'm pretty sure Berkeley (or UC?) had a System V license and was allowed to redistribute code based on it in BSD. That became a point of contention later, when the *BSDs started to become popular. This is all very old stuff so I might have the details wrong. It was definitely a fun time to be at UCB. I remember getting the early Sun workstations and porting stuff to them. It was a really good port of BSD, but that Motorola chip was a problem for some things. I ported the Franz Lisp compiler (liszt) to it.