H316: Add Bob Armstrong's Readme documentation for the IMP tests
This commit is contained in:
parent
964b177616
commit
bbb545d4bd
1 changed files with 129 additions and 0 deletions
129
H316/tests/00readme.txt
Normal file
129
H316/tests/00readme.txt
Normal file
|
@ -0,0 +1,129 @@
|
|||
IMP DEMONSTRATION
|
||||
-----------------
|
||||
|
||||
This directory contains a version of the IMP firmware, ca 1973, and some simh
|
||||
command files to help you run it. These command files define a three IMP
|
||||
network consisting of IMP nodes #2, 3 and 4.
|
||||
|
||||
* IMP 2 has a single virtual modem line which connects to IMP 3.
|
||||
|
||||
* IMP 3 has two modem lines, one connects to IMP 2 and the other to IMP 4.
|
||||
|
||||
* IMP 4 has a single modem line, connected to IMP 3.
|
||||
|
||||
A picture is worth a thousand words -
|
||||
|
||||
IMP 2 <-------> IMP 3 <-------> IMP 4
|
||||
modem modem
|
||||
|
||||
The interesting thing to notice about this arrangement is that there is no
|
||||
path between IMP 2 and IMP 4. Any messages between those two machines must
|
||||
pass thru IMP 3. Sounds simple, but for that to work requires all IMPs to
|
||||
discover their adjacent nodes, find out the neighbors of their neighbors, build
|
||||
routing tables, and then figure out the path to a remote node.
|
||||
|
||||
|
||||
SETUP
|
||||
-----
|
||||
|
||||
Assuming you've already built the H316 simulator with IMP support, open three
|
||||
terminal windows and run on instance of H316 in each.
|
||||
|
||||
Window #1 -> H316 imp2.cmd
|
||||
|
||||
Window #2 -> H316 imp3.cmd
|
||||
|
||||
Window #3 -> H316 imp4.cmd
|
||||
|
||||
These command files will configure the hardware, load the IMP firmware,
|
||||
set the IMP node numbers, and establish the modem connections. These command
|
||||
files DO NOT actually start the simulation, however, so after you've loaded
|
||||
all three IMPs, revisit each of the three windows and type "GO".
|
||||
|
||||
Initially you should see a message like this in all three windows
|
||||
|
||||
DBG(737742)> WDT LIGHTS: changed to 177400
|
||||
|
||||
This message refers to the status of the lights on the IMP's display panel -
|
||||
177400 means all lights are on. After about 30 seconds (yes, 30 seconds!)
|
||||
you'll see this message on IMP 2 and 4
|
||||
|
||||
DBG(19558711)> WDT LIGHTS: changed to 077400
|
||||
|
||||
and these on IMP 3
|
||||
|
||||
DBG(15492720)> WDT LIGHTS: changed to 077400
|
||||
DBG(17096499)> WDT LIGHTS: changed to 037400
|
||||
|
||||
This means IMP 2 and 3 have turned off the light corresponding to modem line 1.
|
||||
The IMP designers used these lights to indicate problems or errors, so a light
|
||||
off means that all is well. IMP 2 and 3 have just told you that their modem
|
||||
line #1 is UP. Remember that IMP 3 has two modems, so we see two lights turn
|
||||
off there.
|
||||
|
||||
BTW, the time it takes for the IMP lines to come up is controlled by
|
||||
intentional delays in the IMP firmware and is driven by the RTC. simh simulates
|
||||
the RTC in real, wall clock, time, and the delay you see here is comparable to
|
||||
the delay you'd see on real IMP hardware. IT HAS NOTHING TO DO WITH THE SPEED
|
||||
OF THE PC RUNNING THE SIMULATION!
|
||||
|
||||
|
||||
NOW WHAT ?
|
||||
----------
|
||||
|
||||
Now that you've got your three IMP network running, what can you do with it?
|
||||
Well, the IMPs had a very simple DDT like command language. Typing "0/" will
|
||||
display the contents of location 0 -
|
||||
|
||||
0/ 0
|
||||
|
||||
The "C" command was used to "cross patch" the terminal on your IMP to another
|
||||
IMP. For example, typing "2C" on IMP 4
|
||||
|
||||
2C
|
||||
|
||||
will connect the console keyboard on IMP 4 to the console printer on IMP 2.
|
||||
Anything you type on IMP 4 will now be echoed on IMP 2 - try it... In the
|
||||
Olden Days ARPAnet sites would use this facility to communicate, operator to
|
||||
operator.
|
||||
|
||||
By adding 100 (octal) to the destination node, you can cross patch to DDT.
|
||||
For example, typing "102C" on IMP 4 will cross patch that TTY to DDT, not the
|
||||
operator, on IMP 2. This allowed guys at BBN and the NCC to debug IMP problems
|
||||
remotely.
|
||||
|
||||
|
||||
FILES INCLUDED
|
||||
--------------
|
||||
|
||||
The files included in this software kit are -
|
||||
|
||||
00readme.txt - this file
|
||||
impcode.cmd - simh script to load the IMP firmware
|
||||
c-listing-ps.txt - assembly listing of the IMP firmware
|
||||
impconfig.cmd - simh script for common IMP configuration
|
||||
imp2.cmd - IMP node #2 specific configuration
|
||||
imp3.cmd - IMP node #3 specific configuration
|
||||
imp4.cmd - IMP node #4 specific configuration
|
||||
imploop.cmd - IMP single line loopback test
|
||||
imploop4.cmd - IMP 4 line loopback test
|
||||
mdmtest1.cmd - simh modem simulation test, part 1
|
||||
mdmtest2.cmd - simh modem simulation test, part 2
|
||||
testrtc.cmd - simh RTC simulation test
|
||||
|
||||
|
||||
MORE READING
|
||||
------------
|
||||
|
||||
Dave Walden's website,
|
||||
|
||||
http://www.walden-family.com/bbn/
|
||||
|
||||
has a wealth of BBN papers, proposals, reports, journals, and other material.
|
||||
There is an entire section devoted to ARPAnet and IMP related content.
|
||||
|
||||
|
||||
Bob Armstrong
|
||||
bob@jfcl.com
|
||||
30-Nov-2013
|
||||
|
Loading…
Add table
Reference in a new issue