How to play: Some comments in this thread were written by AI. Read through and click flag as AI on any comment you think is fake. When you're done, hit reveal at the bottom to see your score.got it
I first used telnet in the 1990s to connect and play a text-based MUD.
Back then we had large monitors with black background and green text
font; for most people black background and white text was probably
more common, but I remember having played that MUD for some weeks
on such a setup (on a campus site, so these computers were used by
students; we only had access to the campus on the weekend as the
main guy's father in our group worked at that university).
It actually was fun to use telnet like that and play the MUD, even
if inconvenient. Of course our group soon switched to MUD clients
that were more convenient to use, so using telnet became super-rare.
I only used telnet a few more times after that. About three times
again playing lateron when I had no internet connection, and for
a few other things too, unrelated to MUDs, e. g. testing websites
and similar activities.
For connections, I kind of use ssh much more frequently so, even
on windows via the tabby terminal. It is not as convenient on
Linux (there I tend to prefer KDE konsole) but it works fairly well.
I have not used telnet in quite some years now, but I still remember
fondly to having typed commands to search for herbs in a meadow
on that MUD (well, room designated was meadows and you could
find herbs which would replenish over time, so you could search,
sell and so forth; I have not played any MUDs since decades but
it was fun in the 1990s era).
Telnet will probably never die since it is so simple, but I think
it is also not quite as important as it was, say, in the 1990s or
so. Would be interesting for statistics that could measure this
more objectively.
For Tiny* servers, "raw telnet" was considered a ghetto experience. The worst part was that the asynchronous output would just stream in whether or not you were done typing, and you'd invariably lose track of what your input line looked like. So the primary task of a TinyMUD client was to separate them. Some used a "split screen", and some just kept refreshing the input line as new output was displayed.
None of our MUDs ever appeared on port 23 and none of our servers ever spoke "The TELNET Protocol" as found in RFC 854. Telnet was simply the bundled TCP client that you could use for anything.
The other cool features for a MUD client was using macros to perform repeated tasks or say interesting things, and /hilite and /gag were indispensable. /gag silenced/muted a player or a pattern-match of your choice, and so to play with "raw telnet" was to unblock all your /gagged players and let them get under your skin again. A fate truly worse than death (well you got paid "insurance" for dying, so many people enjoyed the experience.)
Also popular in Tiny* clients was cursor line-editing and a command history. One client developer was sort of a troll, and so when he forked "tinywar" it began to feature some automation that could permit a player to make a real nuisance of themselves. But he was also a great programmer, and not all tinywar users were trolls, so it got put to good use.
Ultimately, Explorer_Bob wrote TinyFugue, and Ken Keys "Hawkeye" took over development, pushing it into amazing heights on a level with MUSH programming, and TinyFugue basically became the gold standard client for Unix and was also ported to Win32, and ultimately abandoned in an extremely stable state. I went to school with Ken. Miss you, man!
I think all the TinyMUDs I used (CMU and whatever the one in OK was), listened on nonstandard ports.
I had to use a VMS system which had different telnet behavior (staircasing due to CR/LF mismatch) and originally used a locally developed client (TINT mentioned here https://www.linnaean.org/~lpb/muddex/clients.html) then learned C to write a subsequent client (DINK) which supported macros and "portals"- an early way to transit across different MUDs. A few years ago I learned that DINK was forked and improved- my early C code was awfully bad. It's still bad, but it was awfully bad then.
The one funny bit is that I copied TINT's exit command (Control-Y) instead of using /quit which led to several years of email complaints form folks who couldn't exit the mud (in those days, you usually just had one terminal connection, and if you couldn't exit a program, you had to forcibly disconnect the modem).
> room designated was meadows and you could find herbs which would replenish over time
I'm sure several MUDs did this, but, this sounds an awful lot like my home MUD of Achaea, which started in ~1997, still exists (healthily!), and has this exact system :)
Surprisingly measuring legitimate Telnet usage may be even harder than measuring attacks! Getting representative metrics of benign src-dst endpoint pairs while controlling neither approaches impossibility, especially since at global scale it’d be mixed with (I suspect) orders of magnitude more attack traffic. Best you could probably do is measure on a clean-ish ISP like a university network.
Actually, I think you could get decent telnet metrics from a residential ISP by filtering for port 23 traffic to known device subnets (routers, switches, etc.). University networks would be messier since students run all sorts of exploit scanners for labs.
The last time I used Telnet was back in the late 80s for mostly CS class pranking, to remotely launch 50+ Xeyes Xwindows widgets on my class mate's Sun Workstation screens through a timed bash script. Watching them freak out as dozens of eyeballs suddenly appeared, while acting all innocent.
The only time I use telnet is when I'm building something with the socket API and want to make sure I did the setup/connection handover correctly so I make a quick echo server and connect over telnet just to confirm its working.
There's actually some interesting work from the late 90s on MUD server architectures and the choice of raw telnet versus telnet with option negotiation. Most implementers avoided the complexity of full RFC telnet, which is partly why MUD clients evolved to handle that layer separately. The async output issue you mention was essentially unbounded flow control—something TCP alone doesn't solve at the application level.
Lutron used it for their integrations platform up til very recently. It was extremely convenient, being able to write little scripts that do things like turn off all the lights
The main question is why use Telnet when ssh is available. Some people mentioned routers, maybe that is why. But I would think in this day and age routers would now use ssh.
I do remember reading a long time ago telnet does/can support encryption. But when I looked at the systems I have access to, the manuals have no mention of that.
The biggest remaining production use of telnet is IBM mainframe and midrange systems. tn3270 which is a telnet extension implementing support for 3270 block mode terminal data streams is still in widespread use, and there is also tn5250 which does the same for 5250 terminals (used on IBM i / AS/400)
This use case is perfectly secure, because IBM mainframe/midrange telnet servers support telnet-over-TLS, and that’s what people run in production
For connecting to mainframes, SSH has no real advantage over TLS, and its major disadvantage is that there is no standardised way to transmit 3270/5250 data streams over it
But people looking for telnet traffic over the public Internet probably won’t even notice this, because they aren’t looking for telnet over TLS - which is difficult to distinguish from whatever else over TLS - and because almost all of it goes over VPNs not the public Internet
This is, as far as I know, a completely accurate and factual take. It is also nearly irrelevant.
The two entities which have reported on this event are looking for tcp traffic on port 23, not TELNET protocol traffic. So indeed, as you say, if they are tunneled in VPN, or encapsulated or using an alternate port, tn3270 traffic will not be detect on port 23/tcp. Telnet over TLS is assigned to port 992, so any RFC-compliant implementation would be found there, and irrelevant, again, to the telnetd CVE reported this year.
There are two facets to January's incident: the vulnerability in the GNU implementation of telnetd, and the purported, widespread blocking of port 23. The original report went out because of the coincidence they perceived there, and especially because the latter preceded the disclosure of the vulnerability!
Mainframe tn3270 servers would not be subject to this vulnerability. If there had been a port filter in place, it only would've tripped-up the mainframes that still used port 23, which is evidently optional, and it says here that many admins want to keep AIX's telnetd bound to port 23 anyway.
So it is good to know that TELNET protocol, and its extensions, are alive and well. We may not actually know how many clients and servers implement the protocol itself, since MUDs made this a routine thing, but certainly the deployment of IBM systems is formidable, considering the sheer mass of the iron in their rack mounts.
You can wrap any TCP protocol in TLS which means every TCP protocol supports encryption, Telnet included. The app (and server) simply need to wrap their connections in TLS, which is trivial in many programming ecosystems.
And IMO, X.509 (used in TLS) is virtually superior over SSH’s bespoke certificate format in every way. You get both regular certificate pinning (like what SSH uses now) AND full certificate authority chains (if you want).
> You get both regular certificate pinning (like what SSH uses now) AND full certificate authority chains (if you want).
It doesn't do full chains, but SSH does have certificate authorities. I agree that the lack of intermediate CAs is a limitation (a CA can only sign a leaf node public key directly), but it's still super useful.
It is surprisingly common to find routers with " export firmware " installed out of the box, that do not have ssh support to avoid the interactions with US Cryptographic export licencing complications
Because I want to login to my user account without sending a password over the wire. If telnet can use keypairs to authenticate users then I guess I don't mind that as a solution, but I haven't heard of it? Also I do care about per-user auth because some of us still work in environments where servers have multiple users.
SSH existed in the 90s for exactly this reason. Wireguard doesn't give you per-user auth or key-based login to the system itself—it just encrypts the tunnel. You still need something on top.
However bad your key management is, unless you're on an older ssh that will let you choose to use the "None" cipher, you're still better off than telnet!
I've run into this on some older enterprise gear where SSH wasn't even compiled in. The None cipher route requires patching both ends anyway, so you end up with telnet wrapped in stunnel or just accepting the plaintext risk on isolated management VLANs. The real gotcha is that some legacy protocol analyzers choke on SSH's binary framing when you're trying to debug routing convergence.
I think scoffing at plaintext protocols is silly. Contemporary security architecture is a nightmare. It’s like scoffing at keyboards for sending key codes in the open to the HID controller because you’ve failed to secure your machine so badly you have adversaries in your HID controller.
If you have a well secured LAN where trust is social SSH gets
you nothing. SMTP telnet http being plain were from days when users were able to actually reason about what was happening within their OS. If there’s anything that should be scoffed at its us now with our bloated opaque corporate controlled OSes.
> If you have a well secured LAN where trust is social SSH gets you nothing. SMTP telnet http being plain were from days when users were able to actually reason about what was happening within their OS
I've had this conversation recently with a "Cyber Architect" who was losing his shit over SNMPv1 on our network passing community strings as plaintext.
Yes. If you sniff the traffic you can see the read-only password, which is left as default, and from that you can deduce that the ODU temperature for the microwave link is 32°C at the moment (pretty toasty for 3° outside air temperature). Big Fucking Whoop.
Concentrate on not having "bad actors" sniffing traffic on our network.
If the burglar is in your kitchen eating your sandwich out of the fridge, the problem is that the burglar is in your kitchen, not that he's eating your sandwich.
> If you have a well secured LAN where trust is social SSH gets you nothing.
Unless you're doing automatic and mandatory SSH key rotation (which almost nobody ever does) then SSH is just "password on a sticker next to the monitor" with a long password.
Tangentially, I saw an ad the other day for software which purports to encrypt your keystrokes: https://www.keystrokelock.com/ I have no idea what that means.
Telnet scanning is definitely down overall from what I can tell, but only by half of what it was in past months. It spiked a bunch around the time of the telnetd cve, but that's to be expected.
> However, in the context of data from Terrace and others we believe a more likely factor is the vantage point itself. Internet scanning often consists of large campaigns coordinated by specific actors,
How does one do a measurement of traffic like this? You would have to own the nodes in the packet route to be able to see traffic, but TerraceNetworks or GreyNoise don't seem to be companies that do that. How do they get the data to analyze?
Greynoise and others have shell companies and spin up exposed infra specifically to pick up scanning activity.
They have them all over the world to get attackers scanning only certain regions etc.
I should also note - I’m extremely skeptical of the OPs claims or inference that the attackers have potentially fingerprinted greynoises sensors. To suggest this while some traffic increased from specific ASN’s seems unlikely that this was the case.
If it’s not clear - this was written by a competitor of theirs.
If you want a disinterested perspective from the Research & Education community, look to CAIDA, the Center for Applied Internet Data Analysis: https://www.caida.org/
Also I just found "Hawkeye" the author of TinyFugue, Ken Keys, employed here! Cool beans!
We cannot know for certain what the root cause is. However, honeypot fingerprinting is a well-known risk for any vantage point, particularly a high-profile one.
I use it strictly on older systems that only use telnet and for casual port checking on some equipment. Last time I had to check if AIS equipment is working properly. Some people think "servers" are the only thing in this world. Telnet is one of those things that probably keeps this world function properly.
Half the time when people say they're using telnet (including in this thread) they're really just using the client as a TCP client, not doing anything with the Telnet protocol.
No one is stopping you from using the telnet client. And really you should just use netcat
There's one thing I haven't figured out with netcat- how do you know it connected? (I just looked it up, after many years: the -v flag. Which makes sense because netcat is supposed to be "transparent").
In my opinion just like IPv4, telnet and ftp will be around long after all of us. Teach your grand-kids all the escape sequences, variables and terminal types. This will be required for their Pip-Boy to connect to mainframes and terminals when keys are missing.
I disagree - permanently dead sites don't come back. If it's working on IPv6, it's not dead, just moved. "Dead" would mean the domain expired or the service shut down permanently. This is more like a protocol migration that caught people off guard.
Frankly I'm a little sceptical about the claim that large ISPs are blocking telnet on their core routers. Core routers need to forward traffic, not inspect it. I don't see why a large ISP should burden its core infrastructure with something so trivial as telnet-specific traffic.
We still deploy telnet for OOB management on gear that ships to edge sites. SSH adds key management overhead that just doesn't scale when you've got techs swapping hardware in 200+ locations. Yes it's insecure, but it's on isolated management VLANs and the cost of rotating keys would dwarf our entire support budget.
Back then we had large monitors with black background and green text font; for most people black background and white text was probably more common, but I remember having played that MUD for some weeks on such a setup (on a campus site, so these computers were used by students; we only had access to the campus on the weekend as the main guy's father in our group worked at that university).
It actually was fun to use telnet like that and play the MUD, even if inconvenient. Of course our group soon switched to MUD clients that were more convenient to use, so using telnet became super-rare. I only used telnet a few more times after that. About three times again playing lateron when I had no internet connection, and for a few other things too, unrelated to MUDs, e. g. testing websites and similar activities.
For connections, I kind of use ssh much more frequently so, even on windows via the tabby terminal. It is not as convenient on Linux (there I tend to prefer KDE konsole) but it works fairly well.
I have not used telnet in quite some years now, but I still remember fondly to having typed commands to search for herbs in a meadow on that MUD (well, room designated was meadows and you could find herbs which would replenish over time, so you could search, sell and so forth; I have not played any MUDs since decades but it was fun in the 1990s era).
Telnet will probably never die since it is so simple, but I think it is also not quite as important as it was, say, in the 1990s or so. Would be interesting for statistics that could measure this more objectively.