ARPANET (I)The Internet. Infinite expanses. That was not always so.
At the very beginning, the "Internet" was still manageable and you could count the computers that were involved, even if not on a
Hand but in small numbers.
Before you had the possibility to be cyberspacig online, to shop online, to write mails
to spread his mental mischief on Facebook or in blog form ( ;) ) the Internet looked different and was called ARPANET.
The ARPANET was limited at its emergence purely to the USA and was designed as a research network.
Who wants to learn more about the historical development of the ARPANET the must get appropriate literature,
because on the historical aspect I will deal only conditionally.
Since the ARPANET was so clearly structured, a replica offers itself quasi.
But why?Mainly for the fun of learning and because I am fascinated and captivated by ARPANET for many years.
Since I do not have arbitrarily much time at the disposal the project will stretch with security uach in the length.
Who wants to see fast progress and the finished project within a week, I must disappoint
or advise him to do it yourself ;)
DESIGN OF ARPANETBut now enough of the words. May I introduce, the ARPANET (state: May 1973):
... At least times roughly sketched.
Now what do we see on the map?
First of all, we see all the universities that were involved in the ARPANET (Harvard, UCLA, Carnegie, etc.)
and which computers were used there or were connected to the ARPANET (PDP-1, PDP-11, PDP-15 etc.)
ModulesThe individual components/modules of the ARPANET are as follows:
IMPs (Interface Message Processors) are packet switching nodes and the direct predecessors of today's routers.
The very first Request for Comments (RFC1) is about this, among other things. The IMPs are based on mini-computers of the series 16
from Honeywell. The first nodes were Honeywell DDP-516, later the faster H316 were used.
The H316 IMPs were the successors to the originally placed DDP-516 and offered special high speed serial interfaces.
The H316 were among the first ever 16bit computers.
TIPs, or Terminal IMP written out, were the way for users of the ARPANETs to access the network directly via console.
The first successful connection came between Stanford University and UCLA.
One of the first commands tried was "login". However, when entering "login" the computer crashed after the second
letter. The problem was fixed a few hours later, however, and login was working.
So for the rough network structure it needs:
15x - TIPs
15x - IMPs
06x - 316IMP
Different mainframes were attached to the individual nodes of the universities.
02x - PDP15
15x - PDP10
04x - PDP11
02x - PDP1
01x - H645
02x - IBM 360/67
01x - IBM 360/75
01x - IBM 360/65
01x - IBM 360/44
01x - IBM 370/145
01x - IBM 360/91
01x - NOVA
01x - MAXC (PDP-10 Klon)
01x - TSP
01x - TX-2
01x - MICRO 810
01x - B6700
01x - IBM-1800
So our ARPANET includes a total of 36 nodes and 38 mainframes.
Since 38 mainframes and 36 nodes take up a lot of space, the project will be scaled down a bit.
ESP32 microcontroller boards will be used for mainframes and nodes.
They are sufficient in performance, more space efficient than one (or more) PDP-11 and last but not least affordable in quantity.
In Part II it continues with the hardware.