[ih] Why is there (still) spam...?

[Jack Haverty]

This list reminds me of the Internet meetings in the early 1980s. 
Whatever the topic was, we commonly wandered off into other topics.   
That even motivated the invention of a new protocol - the RatHole 
Protocol, or RHP.   Basically anyone in those early Internet meetings, 
at any time, no matter who was talking, could shout out "Rathole!".  
That would cause an immediate survey of the room to see if the majority 
agreed we had gotten off track, and needed to get back to the topic at 
hand.   There is probably some similar mechanism in Robert's Rules of 
Order, but we weren't familiar with such stuff back then.

I'm still interested in Why There Is Still Spam, but hold your Rathole! 
thoughts for a minute more, while I recall ITS a bit...

To answer DaveC's comment - there was a small army of operators and 
administrators caring for the MIT-DM PDP-10 in the 1970s.   I know 
because I was one of them.  We called them "System Managers" (SM) , and 
everyone on the research staff had an assigned day to be SM. The SM's 
job was to fix whatever was broken.  Our PDP-10 broke a lot, usually a 
memory problem, and unfortunately I was one of the two on our staff who 
dealt with hardware.

Historical Trivia question - when fixing a memory, I often had to create 
a Shmoo Diagram and use it to tweak the controls inside the memory 
cabinet.  Does anyone know what I was doing?  Hint - it's not mentioned 
in the Wikipedia writeup.

Re commands like GUN - Such commands didn't appear until after more 
memory, and paging hardware, allowed such a waste of space.  System 
hackers used the shortcut character sequences in DDT to do everything.

So, for example, to "attach" to a running job you would type 
<jobname><ESC>J   To then destroy that job, IIRC you would just type 
<ESC><CTRL-X>.  (the period is part of the command).  Or something like 
that - it's been a long time.....

You could attach to the running ITS system by typing SYS<ESC>J  Anyone 
who did that could change the ITS system itself, while it was running.   
I don't remember exactly, but by changing the contents of some memory 
location, you could cause the system to shut down gracefully.  Most 
often though, ITS just committed suicide, and the SM had to go fix it.

Of course Security was an issue, especially with hordes of very smart 
and curious undergraduates and visitors running around at all hours.  At 
one point, ARPA insisted that we add passwords to ITS. Mike Brescia (MB) 
wrote the code, and after many announcements it was made "live" one 
Friday afternoon.

On Monday morning, a friendly, but unsigned, MOTD (Message Of The Day) 
announced to everyone that a new utility had been built, to help people 
who couldn't remember their password.  Simply typing (something like) 
"passwd jfh" would print out jfh's password.

So we practiced Security By Obscurity.

For example, before you could change the running ITS system at all, you 
had to type a magic incantation - <ESC><ESC><CTRL-R> - but in case 
someone was watching you type, the system echoed $$^D instead.  The D 
key was right above the R key, to fool a lurker about what you actually 
typed.   If that lurker noticed and subsequently tried $$^D it wouldn't 
work and the system wouldn't respond to $$^R any more either.

By the way, I've always thought that such techniques inspired some of 
the aspects of Zork.  Same people were involved.

Another Security By Obscurity technique was the Maze Guncher (MG). 
MazeWars had become quite popular, but it could easily interfere with 
"real work".

Dave Lebling wrote MG, which occasionally would methodically "map" every 
running job on ITS into its own address space, and look for certain 
instruction sequences that were unique to the MazeWars program.  Then it 
would silently alter some instruction, guaranteeing that the program 
would shortly crash, with an indication that a memory error had occurred.

For a while, SYS:ATSIGN MG existed, primarily patrolling during "work 
hours".  IIRC, after gamers discovered MG, we put it into the printer 
spooler to give it some protection.  Security By Obscurity.

I remember "fixing" a lot of memory problems that were actually MG's 
work.   Those needed no Shmoo.

RatHole!!!!

---------------------

So, ... Why Is There Still Spam?   I've never concluded that it's just 
part of human nature, or a problem that cannot be solved. Vint's comment 
in the Podcast that Joly posted seems to me to be a way forward:

"Dr. Vinton Cerf: I think that a primary challenge, is accountability. 
Because of the
way that the Internet can be abused, it's essential that we find ways of 
holding
parties accountable for abusive behaviors, and that will interfere with 
anonymity in
some cases, and so under the right circumstances, you have to penetrate 
the veil of
anonymity in order to hold parties accountable."

Maybe someone will solve the Spam Problem.  I think the solution will 
span not only technology mechanisms but also legal and even 
international agreements.

Meanwhile, for History, what did we do, or not do, that allowed Spam to 
infect The Internet?

Jack Haverty


On 7/29/25 07:15, Lars Brinkhoff via Internet-history wrote:
> Steve Crocker wrote:
>> In ITS, I believe any user could kill any of the processes run by any
>> user, but I may not have all the details.  I don't recall there being
>> a command to bring down the operating system itself.
> I have been puzzling over this.  The JARGON file (as stored on ITS)
> says:
>
>    GUN [from the GUN command on ITS] v. To forcibly terminate a program
>       or job (computer, not career).  "Some idiot left a background
>       process running soaking up half the cycles, so I gunned it."
>
> But there is no straightforward GUN command on ITS.  There is no TS GUN
> program, and no GUN command in DDT.  There is a GUN command in LOCK,
> which is used to kill an entire user job tree.  LOCK also has a command
> to bring the system down: DOWN a.k.a KILL.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature.asc
Type: application/pgp-signature
Size: 665 bytes
Desc: OpenPGP digital signature
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20250729/7868c2cb/attachment.asc>