Hi Julian! The message I sent to everyone (and I'm quoting from very distant memory here, so it may not be word-for-word accurate) was: "This is a test of area coverage. If you get this, please send your nasty, indignant flames to jkh@violet.berkeley.edu". Yes, the text was obviously tongue-in-cheek, but that was the culture at the time and I expected to get such replies ONLY from people at U.C. Berkeley, almost all of whom in the various computer departments knew me personally, so I figured they'd just laugh and send me an email with "Got it, you *&^%$#@! HOW DID YOU DO THAT??" or they wouldn't get anything at all since I didn't expect it to actually WORK. Needless to say, I got a lot of nasty, indignant flames that day from people who did NOT know me, which is why I sent the personal replies. I did not, however, get a personal message from Dennis Perry (the inspector general of the ARPANET) even though he got scribbled on, something I am both grateful and a little sad about - it would have been a message to save! For awhile, I also held the dubious distinction of "the internet's first spammer" and got an honorable mention in the 1989 "International SPAM Olympics", but sadly no modern records of this exist.
I found the text in a mailing list archive. it was: "This is a test of area coverage. Please send me a nasty indignant flame if you get this broadcast. Jordan (jkh@violet.berkeley.edu)" so your memory was almost word-for-word accurate
Thanks for the story. From now on when I screw up with a bit of code I'm going to remember this and have a warm feeling inside that at least I didn't accidently message the entire internet. You've got to love commands / languages that have inverse danger to length relationships. SQL is the classic danger language, the shorter the command the more damage you can do.
@@aka5 Yeah, everyone on the LAN instantly got a "hello" popup (with corresponding *ding* and *beep* noises) from a computer in the reception area (I didn't do this from my computer, of course). As I recall the IT dept were drafted in to manage the "hello virus".
I used to wall pranks like "this system has been hacked". Very few knew the Unix systems they were working on. Most of the techs there just had to run a few scrips to test our cards & such. NOT that I was a guru or anything, especially at that time. But I always felt that C & Unix are KING.
We used "write" all the time in our CS lab to mess with our friends. That, and "eject". Both commands got banned by senior year because things got out of hand
My employer was mainly a VAX/VMS shop back in the day. I remember when the PHONE program made an appearance, and was a novelty for all of ten minutes. Then everybody made sure to add SET BROADCAST/NOPHONE in their login scripts ...
I worked for a company of 10,000 employees back in the 90's, a Windows NT shop. I discovered the windows "net send" command. The entire building ricochet to the sound of floors of workstations lighting up. They soon switched that off in the default policy.
When I was doing remote tech support for customers' servers, the fastest way to get me to *not* work on your issue was to bypass the ticket system and use wall(1) to communicate.
I heard a story from guy who used to teach IT to soldiers back in the windows XP days: A few of the more knowledgeable soldiers were using NET SEND from the command line to secretly chat in class. One of them forgot to include their buddy's computer name in a message and it was broadcast on all computers in the lab. The teacher quickly figured out what they had been doing and said "You just broadcast this to the entire base, maybe even the entire Army!" After class, the soldier was presented with multiple NCOs and officers of ever-increasing rank (up to Major, IIRC) telling him stories of how his message had popped up all over the base and country and affected ongoing operations. They then told him the truth and let him go with a warning to stop chatting in class.
Over 20 years ago I sent Happy Easter via SMS to everyone on the E+ mobile net because their Windows SMS app erroneously allowed for it. It was even in the local news for I think some 17.000 people received it. Then again you can easily write to that many peoply today on reddit or in some facebook groups.
Reminds me of the time my friends and I in school discovered that you can programatically send emails through google drives script feature and we put a "send to entire student body" in for loop that ran a couple thousand times LMAO. IT was very unhappy with us, but thankfully it was easy for them to clear up the issue so we didn't get into to much trouble.
Ah yeah, we used to 'talk' to each other a lot back at my Alma mater :) good times... buddy got wall'd once for this command 'dd if=/dev/zero of=/dev/null' :D
Funny story - a fellow student once send me a message on one of our HPC nodes.. I had a program running via MPI across the Cluster while his message slipped in between the stdout from mpi.. my ssh client intepreted the EOF of the message as EOF of the shell and terminated the whole connection 😂😂 as the program had no resume functionality I had to start over again 🙄
@@lawrencedoliveiro9104 I believe ssh clients use more that just EOF to end the connection. Meaning they use special character sequences, which ofc they can appear in the stream but ssh knows when to listen for them. Search for ssh control sequences. For example right after a newline typing `~.` will terminate the ssh session.
My story is me in high school in the late 90s on the school computers I discovered a command called netsend I think it was called followed by something like -all. I told someone and they did it with a rude message not realizing everyone knew exactly who it came from so a minute later someone came to the classroom and plucked them out lol
Yep I remember talk and it was one to one. I extended it to be multi-user (in 1985) when I was supposed to be learning something that was probably less interesting. I think my version of mtalk (as I named it) has been lost to history.
Oh, what fun. :) I haven't used the wall command in ages but I did use it to alert users of a system shutdown or reboot. I just checked my machine and rwall isn't installed but there is an rwall package available.
Some of my classmates even thought that was the main purpose of the command prompt. So when I used it for something harmless, they almost reported me. (this was a few years ago on a secondary school)
Unless the messaging is opt-in then even something like "write" sounds bad. You could just use the "who" command featured at the start of the video along with a script to mirror the behavior of "wall", run that in a second loop and you're now spamming everyone.
Do a video featuring another channel explaining why information can't travel faster than light theoretically, and what holes in our reality would appear if information was able to be instantaneous
r-commands were a Sun thing mostly (NIS or later Yellow Pages, a sort of primitive service/user discovery system). It spread a little, but it got bad reputation for incidents like the one explained in the video.
Yeah that annoyed me... My ears work great but sometimes I want to watch when I'm in a place where I can't turn the volume up loud enough... But it's really going to cheese off deaf people and seems at odds with the otherwise professional appearance of the channel.
Around that same time, my high school's computer class was a room with an IBM System/360 and 8 terminals (plus one PC) connected to it. There was a similar command to write to any other terminal, and the "bell" was a super-loud mechanical buzzer (more of a clatterer) in the keyboard. We would have fun annoying one another with this, but then someone discovered that if you were on the Console (terminal 2), you could send to all. And they'd all go off at roughly the same time…
@@shaytal100 Attempting to invoke it in bash (Ubuntu machine) says "Command 'rwall' not found, but can be installed with: sudo apt install rwall" I ain't gonna! but it's in the standard repo.
@@thegardenofeatin5965 That is strange. It is neither in the Arch linux repos, nor in the AUR. And usually things that aren't in there are also not in other repos.
What I'm wondering is, how does this work when someone is actively working in the Terminal? Does the message wait until they finish typing something, or are they immediately interuppted?
They aren't interrupted per say, they just get a line feed, a header message (time, sender etc), another line feed and the message that is sent. It doesn't alter their own input, just what text is displayed on screen. If you refresh your editor/screen the message goes away without altering what you've typed in any way. If you are typing a command, your command will still work as you hit enter even if you got a message right in the middle of it.
It appears wherever in the terminal the cursor is at that point. So if you were, let's say, editing a file, you would get some text added to your terminal that wasn't really part of the editor but was indistinguishable and you would try to delete the message but obviously fail because it wasn't really part of the text editor. You can test this by running "tty" and then in another terminal, running "echo hi > {output of tty}" (replacing the curly braces). Something similar can happen when you run a process concurrently without redirecting output and then it outputs something to STDOUT.
The rwall message wouldn't have been accepted by every computer on the internet. It's a bit of an exaggeration to say that everyone on the internet received it.
Agreed, I found this one a bit subpar for a Computerphile video. What was rwall using as network transport? Broadcast? Multicast? And if so, how did the routers from that era deal with multicast messages? By now multicast is all but abandoned when it comes to WAN routing, and it is my understanding that broadcast was never supposed to cross border routers.
@@MrRedstoner Oh, the hosts file, right. Sorry I missed that. Now it also makes sense why the process was taking some time to broadcast the message all over the net.
in the video they said they had to refresh the screen. I suppose that the message was only written to the screen but was not part of the editable text.
Hi Julian! The message I sent to everyone (and I'm quoting from very distant memory here, so it may not be word-for-word accurate) was: "This is a test of area coverage. If you get this, please send your nasty, indignant flames to jkh@violet.berkeley.edu".
Yes, the text was obviously tongue-in-cheek, but that was the culture at the time and I expected to get such replies ONLY from people at U.C. Berkeley, almost all of whom in the various computer departments knew me personally, so I figured they'd just laugh and send me an email with "Got it, you *&^%$#@! HOW DID YOU DO THAT??" or they wouldn't get anything at all since I didn't expect it to actually WORK. Needless to say, I got a lot of nasty, indignant flames that day from people who did NOT know me, which is why I sent the personal replies. I did not, however, get a personal message from Dennis Perry (the inspector general of the ARPANET) even though he got scribbled on, something I am both grateful and a little sad about - it would have been a message to save!
For awhile, I also held the dubious distinction of "the internet's first spammer" and got an honorable mention in the 1989 "International SPAM Olympics", but sadly no modern records of this exist.
I found the text in a mailing list archive. it was:
"This is a test of area coverage. Please send me a nasty indignant flame if you get this broadcast.
Jordan (jkh@violet.berkeley.edu)"
so your memory was almost word-for-word accurate
Thanks for filling in the blanks! It was a different time wasn't it!
Thanks for the comment Jordan :) -Sean
It would be great if this comment was pinned! It complements the original video really well.
Thanks for the story. From now on when I screw up with a bit of code I'm going to remember this and have a warm feeling inside that at least I didn't accidently message the entire internet. You've got to love commands / languages that have inverse danger to length relationships. SQL is the classic danger language, the shorter the command the more damage you can do.
People got ONE spam message and were massively annoyed by it 😂
Yes, we was not so bothered with mail, as we got a handful a day. if there was a lot.
Remember doing `net send * hello` on a Windows LAN about 20 years ago. Chaos ensued - I did not think in a million years it would work.
what did that do, send hello to everyone on the network?
@@aka5 Yeah, everyone on the LAN instantly got a "hello" popup (with corresponding *ding* and *beep* noises) from a computer in the reception area (I didn't do this from my computer, of course). As I recall the IT dept were drafted in to manage the "hello virus".
@@fuzzix I put mine in a bat file after I copied pasted a few hundred times.
@@piyh3962 You know you can use GOTO in batch files, right? :)
@@fuzzix apparently he doesn't (or didn't).
Funny how "remote", "write", and "all" just sound dangerous now, even on their own
Also to be carefull with: recursive
"Hello World!" would have been an appropriate message
And the award for best underrated comment goes to...
Love stories like this!
wonder if there are love stories with wall
I sure love listening to the pioneers talk about old times even though I grew up in the runescape days. Invaluable information.
I used to wall pranks like "this system has been hacked". Very few knew the Unix systems they were working on. Most of the techs there just had to run a few scrips to test our cards & such. NOT that I was a guru or anything, especially at that time. But I always felt that C & Unix are KING.
We used "write" all the time in our CS lab to mess with our friends. That, and "eject". Both commands got banned by senior year because things got out of hand
find / | write username
I Appreciate these Memories, Literally and LITERALLY.
"wall" is the ancestor of "@everyone" on Discord
At least there's no equivalent to rwall. (And no, I don't think the update notifications count; those are waaaay less annoying than rwall could be.)
No, it's like if "@everyone" would notify people outside of the server.
@@IIARROWS outside of the server would be rwall
@@NameGoesHere285 Exactly.
11:21 The Elders of the Internet 😅
Stephen Hawking himself disinfected the internet for you
“...and that’s what the flashing light is for.”
My employer was mainly a VAX/VMS shop back in the day. I remember when the PHONE program made an appearance, and was a novelty for all of ten minutes. Then everybody made sure to add SET BROADCAST/NOPHONE in their login scripts ...
PHONE was great back in the day. I loved it...
This makes me feel better about the time I sent a sarcastic response to a professor's email using BSD's 'mail'. Yep, the default was "Reply All" ...
I worked for a company of 10,000 employees back in the 90's, a Windows NT shop. I discovered the windows "net send" command. The entire building ricochet to the sound of floors of workstations lighting up. They soon switched that off in the default policy.
When I was doing remote tech support for customers' servers, the fastest way to get me to *not* work on your issue was to bypass the ticket system and use wall(1) to communicate.
I heard a story from guy who used to teach IT to soldiers back in the windows XP days:
A few of the more knowledgeable soldiers were using NET SEND from the command line to secretly chat in class. One of them forgot to include their buddy's computer name in a message and it was broadcast on all computers in the lab.
The teacher quickly figured out what they had been doing and said "You just broadcast this to the entire base, maybe even the entire Army!"
After class, the soldier was presented with multiple NCOs and officers of ever-increasing rank (up to Major, IIRC) telling him stories of how his message had popped up all over the base and country and affected ongoing operations. They then told him the truth and let him go with a warning to stop chatting in class.
Pucker factor > 9000
0:55 I remember Prof. Eric Foxley! I still have my printout of his C programming course from 1992.
Over 20 years ago I sent Happy Easter via SMS to everyone on the E+ mobile net because their Windows SMS app erroneously allowed for it. It was even in the local news for I think some 17.000 people received it. Then again you can easily write to that many peoply today on reddit or in some facebook groups.
Reminds me of the time my friends and I in school discovered that you can programatically send emails through google drives script feature and we put a "send to entire student body" in for loop that ran a couple thousand times LMAO. IT was very unhappy with us, but thankfully it was easy for them to clear up the issue so we didn't get into to much trouble.
Ah yeah, we used to 'talk' to each other a lot back at my Alma mater :) good times... buddy got wall'd once for this command 'dd if=/dev/zero of=/dev/null' :D
Is that the digital equivalent of leaving the water running?
Tell everyone on the 'net, "Shall we play a game?"
Great story Julian! Enjoyed it!!
Funny story - a fellow student once send me a message on one of our HPC nodes.. I had a program running via MPI across the Cluster while his message slipped in between the stdout from mpi.. my ssh client intepreted the EOF of the message as EOF of the shell and terminated the whole connection 😂😂
as the program had no resume functionality I had to start over again 🙄
SSH client bug? How would the client even know which part of the output was from wall versus your program?
@@lawrencedoliveiro9104 I believe ssh clients use more that just EOF to end the connection. Meaning they use special character sequences, which ofc they can appear in the stream but ssh knows when to listen for them. Search for ssh control sequences. For example right after a newline typing `~.` will terminate the ssh session.
@@AnastasisGrammenos Yes, but none of that could “accidentally” be generated by server-side output.
My story is me in high school in the late 90s on the school computers I discovered a command called netsend I think it was called followed by something like -all. I told someone and they did it with a rude message not realizing everyone knew exactly who it came from so a minute later someone came to the classroom and plucked them out lol
Yep I remember talk and it was one to one. I extended it to be multi-user (in 1985) when I was supposed to be learning something that was probably less interesting. I think my version of mtalk (as I named it) has been lost to history.
Oh, what fun. :) I haven't used the wall command in ages but I did use it to alert users of a system shutdown or reboot. I just checked my machine and rwall isn't installed but there is an rwall package available.
I used to use write, rtalk, etc. It was the first form of online instant messaging. It was how I first heard about Linux too.
The time before e-marketing... just imagine how many emails went around on (the days preceding) black Friday. Billions.
Learnt a lot about the internet as a newb from a benign 'talk' friend popping in from 'nowhere' in the 90s 😊
so inspiring, so beautiful.. and oh so informative
I remember issuing remote shutdown commands in high school.
anyone remember NET SEND on windows 98? probably was similar to this.
Some of my classmates even thought that was the main purpose of the command prompt. So when I used it for something harmless, they almost reported me. (this was a few years ago on a secondary school)
Had no clue about some of the commands in this video e.g. 'who' and 'w'. Great stuff!
Unless the messaging is opt-in then even something like "write" sounds bad. You could just use the "who" command featured at the start of the video along with a script to mirror the behavior of "wall", run that in a second loop and you're now spamming everyone.
The who results at 1:25 put up some names I recall from when I was at Rutland Hall studying maths and physics in the late 1970s.
I used talk and ytalk quite a lot back in the late 80s/early 90s.
Wonder if this command was on Zuckerberg's mind when he (or whoever) coded the wall
Do a video featuring another channel explaining why information can't travel faster than light theoretically, and what holes in our reality would appear if information was able to be instantaneous
Terminal Room. Scary!
Lovely. Just checked my current RaspberryPiOS and the who and w commands still work!
rwall not found though,
however, "wall" still exist, even on macOS :)
@@roostertechchan Yep, on the RPiOS, too...
r-commands were a Sun thing mostly (NIS or later Yellow Pages, a sort of primitive service/user discovery system). It spread a little, but it got bad reputation for incidents like the one explained in the video.
Computerphile really needs to start adding captions, don't y'all listen to tom scott haha
Yeah that annoyed me... My ears work great but sometimes I want to watch when I'm in a place where I can't turn the volume up loud enough... But it's really going to cheese off deaf people and seems at odds with the otherwise professional appearance of the channel.
Around that same time, my high school's computer class was a room with an IBM System/360 and 8 terminals (plus one PC) connected to it. There was a similar command to write to any other terminal, and the "bell" was a super-loud mechanical buzzer (more of a clatterer) in the keyboard. We would have fun annoying one another with this, but then someone discovered that if you were on the Console (terminal 2), you could send to all. And they'd all go off at roughly the same time…
zsh: command not found: rwall
:(
Does that even depend on the shell? I guess that just means rwall isn't installed on your system
It can almost certainly be installed
@@WilcoVerhoef No it does not. But I use zsh and because of this zsh answers :)
It was a joke. I don't think rwall exist anymore.
@@shaytal100 Attempting to invoke it in bash (Ubuntu machine) says "Command 'rwall' not found, but can be installed with: sudo apt install rwall"
I ain't gonna! but it's in the standard repo.
@@thegardenofeatin5965 That is strange. It is neither in the Arch linux repos, nor in the AUR. And usually things that aren't in there are also not in other repos.
When people don't read the docs: Build that wall! Build that wall!
Now for a breakdown of photon quantum computing
Was expecting something much worse. How nice it would be that the worst you'd get from the internet was a random annoying mesage.
Nothing changes. "This is yor folt." - or some more recent similar things....
rwall sounds like it could actually be useful for emergency broadcasts, like for natural disasters. Maybe limited by area, guessed by IP.
To the windows to the rwall....
awwwwwwww. I wanna beable to do this!!!
And the rest of us want you to NOT be able to do this.
how would write something that connected to everyone on the internet nowadays from the terminal i wonder
i remember talk, but not write and certainly not a wall to the entire internet
What I'm wondering is, how does this work when someone is actively working in the Terminal? Does the message wait until they finish typing something, or are they immediately interuppted?
They aren't interrupted per say, they just get a line feed, a header message (time, sender etc), another line feed and the message that is sent.
It doesn't alter their own input, just what text is displayed on screen.
If you refresh your editor/screen the message goes away without altering what you've typed in any way.
If you are typing a command, your command will still work as you hit enter even if you got a message right in the middle of it.
It appears wherever in the terminal the cursor is at that point. So if you were, let's say, editing a file, you would get some text added to your terminal that wasn't really part of the editor but was indistinguishable and you would try to delete the message but obviously fail because it wasn't really part of the text editor. You can test this by running "tty" and then in another terminal, running "echo hi > {output of tty}" (replacing the curly braces). Something similar can happen when you run a process concurrently without redirecting output and then it outputs something to STDOUT.
The rwall message wouldn't have been accepted by every computer on the internet. It's a bit of an exaggeration to say that everyone on the internet received it.
Agreed, I found this one a bit subpar for a Computerphile video. What was rwall using as network transport? Broadcast? Multicast? And if so, how did the routers from that era deal with multicast messages? By now multicast is all but abandoned when it comes to WAN routing, and it is my understanding that broadcast was never supposed to cross border routers.
He does get to that around 13 minutes in.
@@MrRedstoner Oh, the hosts file, right. Sorry I missed that. Now it also makes sense why the process was taking some time to broadcast the message all over the net.
@Dusty 99 As much as an early adopter as I was, I never had the chance to have my hosts file contain "the entire internet" 😉
9:54 First occurence of SPAM?
hey.... apt-get install rwalld :)
Nice socks
confused at the amount of comments. way too little
How does that did not interfere with a e.g. running vim?
in the video they said they had to refresh the screen. I suppose that the message was only written to the screen but was not part of the editable text.
rsync my fav command , comment your fav one 👇🏻
*Me reading the thumbnail title*
rwall er mi ot te e?
Oh yes no security and able to remotely write to someone's sound device ;)
Yh
ah yes the time when you could call up sun and demand them to patch away the client for the insecure server :'D
I'll send an SMS to the world /
I'll send an SMS to the world /
I hope everyone gets my /
I hope everyone gets my /
Message in an RWALL
Talk about write and wall, but not of talk??
They did.
Yeah he did some talk talk
Next video topic: Noice Reduction....
Wait did he said SSH is not secure? Why?
No, he said rsh was subsumed by ssh because it (rsh) wasn't secure.
Literally no one:
Samsung: 1
Third!
First!
Edit: Second. Fml