Arpa address 3483

cpremote was approximately 1970 between machines at the science center. approx. 1971 vnet was used between science center and endicott for distributed development project enhancing cp67 to provide virtual "370" virtual machines (there were some number of new instructions added to 370 architecture that had to be simulated and the hardware virtual memory tables had different definitions).

in mid-70s there was new joint product announcement for vnet and jes2-nje (vnet in addition to its own line drivers had a large collection of line drivers to talk nje).

jes2-nje had a severe problem because, jes2 managed network node definitions using a one-byte index of 255 entries (that was already being used to define jes2 psuedo devices, which typically might occupy 60 entires) ... and by the time of the joint vnet-jes2 announcement, the internal network was well pbutted the limit of total nodes that could be defined in a jes2 system. jes2 nodes on the internal network had to be carefully managed ... also a jes2 feature was that it would discard network traffic where either the origin node or the destination node wasn't in its table definition. a new activity was invented for managing jes2 network definitions ... managing the subset of internal network node definitions that were critical for the users of that system.

sometime after the internal network pbutted 1000 nodes, jes2 enhanced their networking support to allow for defining up to 999 nodes ... later, after the internal network had pbutted 2000 nodes, jes2 was enhanced to support definition of 1999 nodes.

as i've frequently posted in the past, the corporate communication group had started sna in the early 70s ... aka system network architecture ... which was basically a large sophisticated terminal controller infrastructure. it scaled up for large corporations and insbreastutions that had tens of thousands of dumb terminals (or things controlled like dumb terminals ... atm cash machines, cable tv settop boxes, etc, later insbreastutions like airline res. system that might have several hundred thousand terminals).

this wasn't "networking" ... it was communication control for large numbers of dumb terminals (despite its label sna). the extensive use of the term "networking" by the communication group ... forced other organizations to start calling their stuff "peer-to-peer networking" to differentiate it from the communication stuff that was called sna.

as in this recent posting, my wife somewhat ran afoul of the communication group in that time-frame ... when she co-authored "peer-to-peer networking" architecture (awp39) with bert moldow (as a real networking architecture).

in the 80s, the communication group buttisted in the rapid uptake of PCs in corporate environments with terminal emulation ... dumb corporate terminals, and their millions, could be upgraded to a PC (the price of the dumb corporate terminals and PCs were approx. the same) ... and for a single desktop footprint ... the person got a display-keyboard that provided both access to corporate dataprocessing and some local computing facilities.

in the late 80s, as client-server networking started to emerge ... the communication group attempted to counter with "SAA" ... which attempt to preserve as much as possible of the terminal emulation paradigm as possible.

in that time frame, my wife and I had come up with 3-tier networking architecture and were out pitching it to corporate IT executives and taking some amount of heat from the communication group (which wanted to continue the terminal emulation paradigm and stave off any progression to networking).

I've butterted that one the reasons that the number of nodes on the internet pbutted the number of nodes on the internal network sometime mid-85 ... was that while the communication group couldn't prevent the internal network ... they were able to do a lot to head-off being able to treat workstations and PCs connected to the internal mainframes as networking nodes (maintain their status as terminal emulation) ... at a time when internet was seeing a big explosion in workstations and PCs as networking nodes.

the following article lists cpremote as completed mid-68 plus 1.

it was then replaced with scnode ... followed by RSCS which was subsequently announced and released as a product. The article mentions that the internal network grew at a relatively constant rate of one node a week through 1977 and then in 1978 the growth doubled to two nodes a week.

another reference that discusses some of this ... cpremote, rscs, vnet, rscs, etc.

I don't know the number of DEC machines internally. I've been able to post the information that I have access to. In the past, I've posted the number of vax machines sold

and noted that in the late 70s and early 80s ... there were as many or more 43xx machines sold as there were vax machines sold. there was some sort of transition-threshold that happened in this period and you saw corporate orders for several hundred 43xx machines at a time.

I've repeatedly and frequently posted both data that I have as well as other references. To some extent the world knows about arpanet because a lot of people made it their business to publish information about it for one reason or another. During that period, there was little or no public documents about the internal network ... in part because of the communication group's effort to focus customers totally on their SNA activity and terminal communication paradigm. in fact, the communication group strongly objected to vnet being part of the joint jes2-vnet networking announcement in the mid-70s (and had prevented it from being announced earlier).

however, despite all the efforts by the communication group to put a totally different face on "communication", "networking" and products announced by the corporation for customer use ... that doesn't mean that the internal network didn't exist. I conjecture that there are a lot of things that go on within various insbreastutions and govs. ... that may not be public knowledge ... but just because it isn't public knowledge ... doesn't mean that it didn't happen.

--