The "WAS" in that title just made me feel really old. I remember learning about this in college and it being a thing that I would need to work with and understand how it works why it's different than ethernet. And now it's just a big "WAS".
I remember learning about TokenRing in high school circa 2005 for the CompTIA+ certificate.
Never ever seen it in my life though. It was a big "WAS" even then.
The i386-i486 machines we had in middle school were on the Ethernet bus (I remember the BNC cable going along the classroom, with T-joints to the network cards).
IIRC Token Ring wasn't a thing in practice for 100Mbit speeds and beyond, and went the way of the dodo faster than DOS did.
It'll still come in useful, conceptually, if you work with 10base-T1S (multi-drop automotive Ethernet) -- a 2019 standard -- which uses a virtual token ring to implement collision avoidance.
I completed a short computer networking course just earlier this year, and token ring was mentioned in almost as much detail as ethernet. (Frankly, we weren't given a great deal of detail on either: it was very much an introductory course.)
My local hospital system (Premier Health) was all-in on IBM and token ring. I remember being chagrined when they built a new hospital in 1998 and wired everything for token ring. The writing was clearly on the wall for token ring by 1998 but their IT department bled IBM blue.
Fast-forward to 2013, when my wife was giving birth to our daughter in one of their hospitals, and I noticed an unused type-1 token ring wall jack in our room right beside some occupied 8P8C (aka RJ-45) jacks.
I remember seeing those same plugs in one of the old office where I work, always thought it was for some old IBM connection, now I know exactly what it is!
We used to love token ring style networks (specifically ARCNET) for playing Doom, Descent and Quake back in the day. ARCNET was more annoying to set up and had thicker cables, but it was rock-solid reliable afterwards. Totally bulletproof.
In comparison to that, traditional ethernet (think concentric cabling and terminators) glitched often and one flaky part of a cable could ruin your day.
ARCNET was much slower in theory, but robust, fair and stable and did not totally die under load from a single client.
People may forget that Ethernet originally had a shared transmission medium and collisions were a serious limitation. Switching was a later addition that mostly solved collisions and really paved the way for Ethernet as we know it today.
Interestingly, early Ethernet performance studies suggested that about 37% utilisation was the best possible on a heavily contended network due to all the collisions. So a 10mbps Ethernet cable run was arguably similar in performance to 4mbps TR? That was essentially IBM's marketing argument for 4mbps token ring. But TR was more predictable under heavy traffic contention. That’s what we were told anyway…
Trivia: because IEEE 802.3 working group was formally in charge for standards of "local area networks with CSMA/CD", 1000BASE-T has half-duplex mode defined (which nobody implemented) as otherwise it would fall outside of working group purview.
Hubs were still just barely still relevant when I first started learning networking, but it was late enough that I don't think I've ever encountered one in serious use. Especially after port mirroring became a common feature even on affordable switches. I think we had an old pile of industrial/scientific test equipment in grad school nobody wanted to touch that was using an old Cisco hub, but my memory is fuzzy.
I've definitely never seen any old Thicknet networks in use, though I have seen Token Ring in active use semi-recently. I never learned anything about Token Ring other than I remember old CompTIA mentioning it existed.
I do have a fondness for seeing this old tech that was way out of my reach or just barely too old to encounter back in the day being used, it's so neat to learn about. ClabRetro on YouTube is one of my few "drop everything and watch a new video" creators.
This reminded me of my first encounter with Thicknet (and thinnet) just a few years ago.
We'd never gone over it when I went to college in 2012 but come 2022, I found myself learning about it by pulling it out of the ceiling during a decluttering project.
Made some sense, though. Being a TV station, coaxial cable was probably very plentiful when it was installed. We saved a couple of the vampire clamps at the time just for laughs. I wonder if they're still in the office.
Back in 1999 I was still doing the crawling-under-desks thing. I was at a tobacco companies sales office... and they had a wallplate with ports labeled token ring, ethernet, and appletalk (looked rj11, so phonenet?). A year or two later I worked at their central location in Virginia in a warehouse decommisioning old equipement, got to see some FDDI and HIPPI hardware. Never saw it being used though.
And at my previous job, in a quite normal office, there was a wallplate labeled DEC with the weirdest jack I've ever seen. I couldn't visually identify it, and I thought I knew everything. 1.5" inches wide, with a plastic spring-loaded cover that came down from the top. Maybe only two conductors in it, not normal rj-style pins at all. I'd guess something Vax, but I thought those were always ethernet since the early 80s, and anything older still wouldn't likely have survived until 2025.
Hubs were pretty common when I went to college (2003-2007), but not for the school's infrastructure. They would lend them to students because the rooms only had a single network port. Not sure how long that use case lasted; I imagine by now renovations have happened and rooms have multiple network ports (and the dorms would have wireless in any case).
You were one generation before me, I started college in 2007. By then it was all those super cheap little 4 port Netgear switches most people reading this have probably owned at some point. I don't remember us having official dorm wifi but we were allowed to have our own wireless routers, probably an IT nightmare by today's standards. I recall we also all got publicly routable IP4s.
My first exposure to networking was a 10base2 setup I got from a closing business. Setting that up to run an ipx/spx network which I could run doom/descent/quake on was my very first intro to networking. Was so cool to see computers across the room communicate back then. Something very satisfying about the way those coax connectors twist in and click.
> My first exposure to networking was a 10base2 setup I got from a closing business.
Exact same! My neighbors (still friends to this day) managed to put their hands on lots of used 10base2 network cards and coax cables / Ts and terminators from a company that closed shop.
So we put those cards in our PCs and created a LAN and we had a coax cable going from our bedrooms (my brother and I) to the neighbors' house. On the third floor! We contemplated using utility cables in the middle of the street to hook the neighbor across the street but decided it wasn't wise (we were teenagers so we still contemplated the idea). We also fancied the idea of drilling a hole between the two houses: my neighbor's father was fine with it and provided us with a really long drill but my parents didn't want to hear about it.
And we didn't have the proper dedicated BNC connectors crimp tool though (and had zero idea where to buy one and were too broke anyway), so we became masters at building coax cables with BNC connectors by using regular pliers.
Them little orcs and humans from Warcraft 2 would then transit through the coax cable from house to house (so around 1995).
First college computer lab was coax 10base2. Got introduced to MUDs on that one. And yeah, coax was unwieldy to work with, but we weren't pulling cables out on accident (at least not without taking the whole case with it).
TokenRing was already an academic curiosity when I was getting my CS degree in the mid-90s. The CS department by that point had already switched to twisted-pair Ethernet. By the time of my first job in the late 90s, we used FDDI for our web farm's backend network, but we tore that out with the arrival of gigabit, so FDDI didn't last too long either.
Token Ring topology was included in Microsoft's Networking Essentials certification at least as late as 2000. I started my first tech job in 1998 at a large bank and they still had production token ring networks.
I worked with Prime minicomputers when I was a late teen. They introduced RingNet, a token ring product, in 1979.
Fun fact: each system had an IN and an OUT coax port, and the coax cable was about as thick as a thumb. If a system died, a small hardware box short circuited, tying the IN and OUT coaxes together. This kept the ring intact, but it would also make that segment of the network longer and there were strict distance limitations between nodes. If the distance got too big, the entire ring broke. Also, if any coax cable got damaged, that also lead to a ring break. Prime had a special diagnostic program they could use on multiple nodes to try to find ring breaks.
Token-passing networks were interesting for hard real time system analysis, as they featured deterministic access control instead of collision detection. They made possible extending Rate Monotonic Scheduling theory to the network layer as well. The priority work (priority ceiling protocol, multi level token priority schemes) survive today in token bucket rate limiting.
Interestingly, the CAN bus features deterministic access control through collision detection, using a bus topology rather than a ring topology. I've had bad experiences with CAN, but I believe they're due to hardware bugs and/or driver bugs rather than the underlying design.
Arbitration in CAN is still one of the coolest things I've seen at the hardware layer.
Being able to handle a collision without needing the higher priority message to skip a single beat seemed like magic to me when I first learned about it
the token bus combines the one-user-at-a time of a bus with the potential-token-loss of a ring.
IEEE 802.4 token bus standard was withdrawn
polled buses with a master are pretty common when you don't mind the performance hit and central-point-of-failure. i believe cable networks work this way.
Interesting, I'd never heard of token-bus networks, or IEEE 802.4, which was indeed a token-bus standard.
CAN isn't a polled bus, doesn't have a master, and doesn't impose a performance hit. However, on a bus network, every device on the bus is a single point of failure: if it starts jabbering, the network stops working.
I love watching this channel as a 20-something who never had a chance to see/touch the hardware he's working with. Great way to show the real-world applications of this tech rather than just reading historical summaries in textbooks.
Some of it is very telling about how close we were to having a completely different tech ecosystem. Have you seen his video on Sun thin clients? They had this tech working back when PCs were $2k a pop. They had this cheap thin client that had all the peripherals and authentication, plus a very rich java runtime and libraries, but due to mismanagement the tech never got off the ground.
IBM POWER is another odd one. They are still going strong because their shit Just Works so anyone doing Serious Business just keeps using them. They have hot pluggable CPU and ram ffs! Compared to cloud computing they are tiny, but they aren't going anywhere anytime soon.
> Coax connectors for Token Ring did exist, but were more common on Ethernet networks.
I don’t disbelieve the author here but anecdotally the only time I’ve worked with coax BNC plugs for networking is with TR. And it be never seen a TR network in the wild that wasn’t coax.
It’s interesting to read that other cable types existed for TR.
Conversely, the only time I've worked with coax BNC plugs for networking was with 10BASE2 ethernet (aka "thinnet" and also "that stupid broken thing").
That was just one long, shared transmission line that happened with RG-58 coax. Each end was terminated with a 50 Ohm resistor. There was no central brain-box or hub. All nodes were peers.
Nodes could be inserted on an existing run by bisecting the coax at any random point, terminating the bare ends with BNC male connectors, and plugging those into a BNC T connector which ultimately connected to a NIC. Doing this brought down the entire network segment in unison until it was reconnected.
In fact, any discontinuity in the transmission line brought down the entire network segment.
So when Bob in Accounting decides to move his PC to a different spot on his desk and breaks the transmission line? The network dips out completely, and without any good tools for a IT folks to pinpoint exactly where that break happened.
Fun times. (Except: It actually was kind of fun for homelab-type stuff when 10BASE-T twisted-pair networks were still kind of expensive. Adding more nodes just meant installing more connectors and maybe extending the line. Now we just add another magical non-blocking N-way $15 gigabit network switch wherever, or skip it all and use wifi.)
---
I've never worked with Token Ring at all. If I ever even saw it in the wild, I did not recognize it as such.
Yeah, it is. I even Googled it to make sure I got the hyphenization (or lack thereof) right, and then I still successfully screwed up the actually-distinguishing part.
I think we spent more time in the computer lab in middle school running NET SEND and then imperceptibly loosening the coax on the back of our machine to take the network down than doing the actual exercises we were supposed to do
Thanks for sharing. That sounds a lot like my experiences as well so now I’m wondering if I’m misremembering those networks as TR (it was a generation ago).
Coax Ethernet with BNC connectors was absolutely the norm for LANs from the late 1980s to mid 1990s. Then, slowly, UTP took over, as 10base-2 was replaced by 10base-T, then 100base-T, then switches instead of hubs, then Gigabit Ethernet: 1000base-T.
I worked with Token Ring in the '80s and never ever even heard of a coax version. It was D9 connectors on the card to fat IBM connectors on the MAU, as pictured here:
Token Ring was a networking alternative to Ethernet.
With Ethernet when a node wants to transmit it just does so, then backs off and retries if there was a collision (someone else trying to transmit at the same time).
Token Ring avoided collisions by passing a "permission to transmit" token from node to node in circular (ring) fashion.
> With Ethernet when a node wants to transmit it just does so
Depends on the Ethernet variant. Older versions have collision detection as you describe. More recent versions are point-to-point full-duplex and thus collision free, with the outlier of 10base-T1S, which uses a virtual token ring for collision avoidance.
A very nice trip down memory lane. The big banks bought token ring because nobody gets fired for buying IBM, and Ethernet looked so amateurish compared to the heft of TR cables and chunky spaceship MAUs. Hermaphrodite connectors, remember those? I knew a bank that ran TR at 4mbps, connecting its head office buildings nearby and was considering an upgrade to 16 - it seemed mysteriously fast in those days. But then Ethernet came along, and SCO Unix servers on 486, so it was relatively easy and affordable to get a department of 30 people up and running with PCs as terminals. Those were the days…
> and unlike Ethernet, all the cards on the ring had to talk at the same speed.
Essentially that was the same for Ethernet too, everything on the same leg needs to talk at the same speed, though IIRC a different speed device wouldn't take down everything, it would just cause a significant excess of collisions due to the out-of-sync cross-talk and not be able to talk to anything itself. It wasn't until switching hubs and full switches were around that you could mix speeds on the same leg (with a switching hub you'd actually have two legs, one at each speed, bridged, with a full switch you have a separate collision domain per port so can mix speeds quite freely).
Oh, flashback to learning queueing theory decades ago!
In principle, token ring was faster than Ethernet (which was CSMA/CD in the day): the mathematics of token ring were exact and analytic, and, when solved, gave higher speeds and better reliability than Ethernet, the mathematics of which were so complex (think three-body-problem) that analytic solutions were available only for bounding equations, which resolved to slower than token ring.
In practice, Ethernet was much faster.
A classic "the model is not the territory" problem.
There's a Minecraft YouTuber called TangoTek who's building a deck building game called Decked Out 3 in Minecraft. That is using a token ring system to control events and messages sent in the game to avoid conflicts [1].
I just read through his Wikipedia entry and didn't see that he'd gone ''crazy'' unless you mean his political views? He's dead now anyways. I linked to that comic because I had bosses like that way back when.
I think you mean "went" and I wish there was a very brief plain text summary rather than a long video that I don't have time to watch. Anyway, I enjoyed the token ring comic and don't have strong feelings about the cartoonist now that he's gone.
At a 2600 conference long ago, we built a cluster of public X11 dumb terminals, so hackers at the conference could browse the web and such. We hardened the server against local user attack, and the clients booted from a read only solid state media (aka, floppy disk) with no port access. But how do you harden a physical network against a thousand hackers? Solution: token ring. We made the network so ancient that nobody would have the hardware to hack it that weekend. Worked great.
I only ever worked on one token ring network in the 90s - the client used mostly IBM hardware, and I guess that's what the consultants who set them up installed. It worked fairly well, but the main thing I remember was the switches were really, really noisy.
Every other network I worked on was ethernet, with the exception of one arcnet network. Which also worked really well, except it was very, very slow.
When I started my first job in the late 90s, they had Linux floppy disk based thin clients running over coax token ring.
The cable was laying right under my desk. When I touched it with my feet, there would be a disconnect because the coax t-connector was not sitting firmly enough on the network card of my device.
All unsafe work would be lost.
Now we have time slicing for high reliability Ethernet (IEEE 802.1Qbv). I have never worked with it and don't properly understand it, but I think that instead of passing a token around, there are time windows for high priority devices to transmit.
Low priority devices must stop transmitting in the window, and switches will interrupt any transmissions not allowed.
When I was an intern at Merck in the late 1990s, the networking team was switching over from Token Ring to Ethernet.
It was so much work that they were openly recruiting anyone who was an employee or intern to come in on a weekend and help out. They eventually offered to pay people extra and a couple of my intern buddies took up the offer.
The first time I read about this was about 35 years ago. I read about it in a computer science textbook my father owned. I was just a child so I never thought much about it, except that the name was pretty cool.
That textbook was sort of system textbook which describes a lot of low level stuffs such as networking, hard disks and such.
FYI cable networks (upstream) are essentially token ring.
In fact, broadcast ethernet is essentially token ring, except instead of a token it just grabs the wire. Only one machine can grab the wire at a time. Switches hide that from most people.
No, DOCSIS is nothing like token ring. It doesn't use tokens, it uses channels for up/down and TDMA/CDMA.
No, 'broadcast ethernet' is not token ring. There is no token. It's not a ring. Switches do hide the details, but it's hiding collision domains, not token passing.
The last token ring standard appeared in 2001 and there has been a huge amount of network-technology development since then. Tokenring is tiny and incredibly niche. It may as well be dead.
Ethernet got its latest in 2022 and another is in the pipeline.
I find this guy kinda annoying. I mean, yeah...seems like he did some networking back in the day, but everything he posts is just enough off (or ia Google-correct, not Been-There-correct) that I can't quite take him seriously.
The main reason Ethernet won the war was because it was cheaper.
He should have stopped there. That was it. And because it was cheaper, less complicated to build the chips, 'good enough' networking, a bunch of folks jumped on the bandwagon. And here we are.
If you had a 16 megabit network and brought up a system at 4 megabits, it took down the whole network.
Well, yes, everyone who has ever mentioned TR tells some variation of this. And if you left the terminator off of the coax ethernet, you brought the whole network down. But TR got smart MAUs and ethernet got UTP hubs around the same time and problem solved.
And hubs were a lot cheaper than smart MAUs. Sooo...
Ethernet was junk if you went with the cheapest options all around, but you could make it acceptable if you invested some of the savings in improved technology.
Just like TR. If you were deploying dirt cheap 8228s after 1990 or so, your network still sucked. You didn't do that.
But the other reason Ethernet won was because of scale.
Yes, there was a thing called 100Mb Token Ring. Someone might have even shipped some number of them. I never saw one in the wild, and I don't know anyone who claimed they had who didn't work for IBM or Madge (which, btw, wasn't an 'obscure' network company at the time and made more than just TR cards).
That's because basically everyone who did Token Ring (IBM most of all) had decided that you wanted to scale TR you were going to ATM (for a bunch of mostly political and marketing reasons). Which, unlike TR, was in reality a very bad idea, but that's another story.
I've never even heard anyone even suggest 128Mb Token Ring, and I suspect someones LLM is hallucinating. There was a gigabit Token Ring standard (IEEE 802.5v), but I feel pretty confident saying noone ever shipped it. Like I said...ATM was gonna be the future.
Ethernet used a 15-pin connector that looks quite a bit like a PC joystick connector. Token Ring used a 9-pin connector that looks like a 9-pin video connector or an Atari style joystick plug.
Ok...I'll just be snarky and point out there are ethernet cards that use a 9-pin D-sub connector to reuse a TR wiring physical plant. I have several.
Coax connectors for Token Ring did exist
Citation needed. I've seen coax for 802.4 industrial networking stuff, which is not TR. I'm betting more hallucinations.
ARCNET was another competitor from the same timeframe. It, too, was a contender for a while because it was backed by Datapoint, a pioneer in office automation using small computers.
Arcnet was a contender because it was even cheaper than ethernet and it was (and still is) widely used in industrial environments. And some popular early PC file sharing systems (e.g. LANTastic) would use Arcnet.
> Datapoint, a pioneer in office automation using small computers
In 1970, Datapoint created a desktop computer, the Datapoint 2200, with an 8-bit TTL processor. Intel created a microprocessor version of this processor for Datapoint, called the Intel 8008. Datapoint decided to stick with TTL, so Intel ended up selling the 8008 as a general-purpose microprocessor. Intel improved it to make the 8080, then made a 16-bit version, the 8086, as a stopgap product until the iAPX 432 micro-mainframe was ready. The iAPX 432 was a disaster and that's how you ended up with x86.
Ken...I know you tore open an Intel 82586 ethernet chip and an IBM Tropic TR chip. It might be interesting to do a comparison of the complexity of those two, being that they're roughly contemporary.
Apollo Domain computers had a token ring which used coax. RCA phono plug connectors. We had a pair of those at the aerospace company for testing.
Worked OK, but too weird. We ended up with Sun workstations.
At various times in my early career, I was involved with strange networking hardware only of historical interest now. Apollo Domain (sort of like Unix, but not.) "Thick Ethernet" 10BASE5. Later "Thin Ethernet" (Yes, you can mix thick and thin Ethernet without an impedance mismatch, but you may need to use a time domain reflectometer to check. The adapters were built to tolerate TDM pulses.) Hyperchannel, 20Mb/s which required 6U of rack space for the interface. (We had a pair, surplus from another project, but didn't do much with them.) DECNET triaxial cable for connecting things called the DMC-11.
These were all coax, and they were all true "baseband", that is, the signal was meaningful down to DC. No differential signals, and no scrambler to make the number of 1 and 0 bits roughly equal.
You can't use decoupling capacitors or transformers. They were all susceptible to noise and ground loops. The coax shield should only be grounded at one point for the whole chain. Otherwise, noise in the ground comes through as noise in the cable.
All this was no problem in a place full of RF engineers, who knew all that, and considered hooking up a TDR or a scope to a data network normal operational setup. And in an RF R&D facility, where copper busbars an inch wide were run in parallel with electrical ducts to solidly ground out noise, ground was solid.
So coax worked fine at Xerox PARC, it worked fine at Stanford (although inter-building coax didn't work all that well), it worked fine in single-room installations, and it worked fine for organizations which understood the limitations of non-differential baseband over coax.
The coax technology thus needed too much basic electronics knowledge for general office use.
Modern Ethernet is pretty much idiot proof at the electrical level. It's differential signals over a pair, immune to common-mode noise and ground level differences. If you plug standard cables into standard interfaces, it will work.
I ran a lab of Apollo 3500/4500 era machine in college (thanks for poking that memory). None of them had RCA phono connectors for token ring, but they did have a really weird connector that looked like it had 2 coax(ish) connections, along with some pins, in a D shell. Think something that looks like a 13w3 video connector, but only 2 signals and fewer pins. Are you thinking of an earlier generation of Apollo machines?
We had some early models which had two ordinary coax lines going in. Unplug either and the ring went down. They probably converted to something with one connector, and maybe a mechanism to close the loop if unplugged. That's a basic problem with token ring - anything removed must be jumpered out.
We just had two loaners for a demo, so we might have been given an old model.
We were evaluating workstations, and had various devices for a few months at a time.
I saw quite a bit of rare early hardware over my career. Much of it was rare for a good reason.
I wish I had a smart mau but she is standing on the windowsill mau-ing at the part that doesn't open while the other side is wide open. In the rain. Which is getting inside while she maus to get inside.
He is writing about tech since 1991, but I read a few articles today and most of them seem to repeat the same thing 3x in the first paragraph which is a bit off putting. This article doesnt but I agree, its ‘just enough’? Anyway cool to have all this in one place!
The "WAS" in that title just made me feel really old. I remember learning about this in college and it being a thing that I would need to work with and understand how it works why it's different than ethernet. And now it's just a big "WAS".
I remember learning about TokenRing in high school circa 2005 for the CompTIA+ certificate.
Never ever seen it in my life though. It was a big "WAS" even then.
The i386-i486 machines we had in middle school were on the Ethernet bus (I remember the BNC cable going along the classroom, with T-joints to the network cards).
IIRC Token Ring wasn't a thing in practice for 100Mbit speeds and beyond, and went the way of the dodo faster than DOS did.
It'll still come in useful, conceptually, if you work with 10base-T1S (multi-drop automotive Ethernet) -- a 2019 standard -- which uses a virtual token ring to implement collision avoidance.
I completed a short computer networking course just earlier this year, and token ring was mentioned in almost as much detail as ethernet. (Frankly, we weren't given a great deal of detail on either: it was very much an introductory course.)
It sounds like you should complain about the crappy course that you paid for (with your tax dollars if nothing else).
When I first took the CCNA test it still has Frame Relay questions.
Ex Novel Network MCNE, came here to say the same thing.
Sometime I wonder if any of the old protocols have new merit because of some other advance.
My local hospital system (Premier Health) was all-in on IBM and token ring. I remember being chagrined when they built a new hospital in 1998 and wired everything for token ring. The writing was clearly on the wall for token ring by 1998 but their IT department bled IBM blue.
Fast-forward to 2013, when my wife was giving birth to our daughter in one of their hospitals, and I noticed an unused type-1 token ring wall jack in our room right beside some occupied 8P8C (aka RJ-45) jacks.
BTW there are adapters so you can run Ethernet over Token Ring cabling. We had a lot of them at IBM since all the offices were wired for Token Ring.
I have a couple of D-Link ethernet cards with 9-pin DB9 connectors built in for exactly this purpose.
And that was Ethernet at 4mbps over IBM “structured cabling”?
No, Token Ring cabling was so over-engineered that you could run 100 Mbps Ethernet over it.
Four milli bits per second sounds very slow to me.
I remember seeing those same plugs in one of the old office where I work, always thought it was for some old IBM connection, now I know exactly what it is!
We used to love token ring style networks (specifically ARCNET) for playing Doom, Descent and Quake back in the day. ARCNET was more annoying to set up and had thicker cables, but it was rock-solid reliable afterwards. Totally bulletproof.
In comparison to that, traditional ethernet (think concentric cabling and terminators) glitched often and one flaky part of a cable could ruin your day.
ARCNET was much slower in theory, but robust, fair and stable and did not totally die under load from a single client.
People may forget that Ethernet originally had a shared transmission medium and collisions were a serious limitation. Switching was a later addition that mostly solved collisions and really paved the way for Ethernet as we know it today.
Interestingly, early Ethernet performance studies suggested that about 37% utilisation was the best possible on a heavily contended network due to all the collisions. So a 10mbps Ethernet cable run was arguably similar in performance to 4mbps TR? That was essentially IBM's marketing argument for 4mbps token ring. But TR was more predictable under heavy traffic contention. That’s what we were told anyway…
Trivia: because IEEE 802.3 working group was formally in charge for standards of "local area networks with CSMA/CD", 1000BASE-T has half-duplex mode defined (which nobody implemented) as otherwise it would fall outside of working group purview.
I remember finding an old hub off in a far corner in 2008 working on my school’s network. It was a rare find then!
Hubs were still just barely still relevant when I first started learning networking, but it was late enough that I don't think I've ever encountered one in serious use. Especially after port mirroring became a common feature even on affordable switches. I think we had an old pile of industrial/scientific test equipment in grad school nobody wanted to touch that was using an old Cisco hub, but my memory is fuzzy.
I've definitely never seen any old Thicknet networks in use, though I have seen Token Ring in active use semi-recently. I never learned anything about Token Ring other than I remember old CompTIA mentioning it existed.
I do have a fondness for seeing this old tech that was way out of my reach or just barely too old to encounter back in the day being used, it's so neat to learn about. ClabRetro on YouTube is one of my few "drop everything and watch a new video" creators.
This reminded me of my first encounter with Thicknet (and thinnet) just a few years ago.
We'd never gone over it when I went to college in 2012 but come 2022, I found myself learning about it by pulling it out of the ceiling during a decluttering project.
Made some sense, though. Being a TV station, coaxial cable was probably very plentiful when it was installed. We saved a couple of the vampire clamps at the time just for laughs. I wonder if they're still in the office.
Back in 1999 I was still doing the crawling-under-desks thing. I was at a tobacco companies sales office... and they had a wallplate with ports labeled token ring, ethernet, and appletalk (looked rj11, so phonenet?). A year or two later I worked at their central location in Virginia in a warehouse decommisioning old equipement, got to see some FDDI and HIPPI hardware. Never saw it being used though.
And at my previous job, in a quite normal office, there was a wallplate labeled DEC with the weirdest jack I've ever seen. I couldn't visually identify it, and I thought I knew everything. 1.5" inches wide, with a plastic spring-loaded cover that came down from the top. Maybe only two conductors in it, not normal rj-style pins at all. I'd guess something Vax, but I thought those were always ethernet since the early 80s, and anything older still wouldn't likely have survived until 2025.
Hubs were pretty common when I went to college (2003-2007), but not for the school's infrastructure. They would lend them to students because the rooms only had a single network port. Not sure how long that use case lasted; I imagine by now renovations have happened and rooms have multiple network ports (and the dorms would have wireless in any case).
You were one generation before me, I started college in 2007. By then it was all those super cheap little 4 port Netgear switches most people reading this have probably owned at some point. I don't remember us having official dorm wifi but we were allowed to have our own wireless routers, probably an IT nightmare by today's standards. I recall we also all got publicly routable IP4s.
Hubs were valuable at our CS dept because you could sniff traffic easily with accurate timestamps
And still people use wi-fi...
My first exposure to networking was a 10base2 setup I got from a closing business. Setting that up to run an ipx/spx network which I could run doom/descent/quake on was my very first intro to networking. Was so cool to see computers across the room communicate back then. Something very satisfying about the way those coax connectors twist in and click.
> My first exposure to networking was a 10base2 setup I got from a closing business.
Exact same! My neighbors (still friends to this day) managed to put their hands on lots of used 10base2 network cards and coax cables / Ts and terminators from a company that closed shop.
So we put those cards in our PCs and created a LAN and we had a coax cable going from our bedrooms (my brother and I) to the neighbors' house. On the third floor! We contemplated using utility cables in the middle of the street to hook the neighbor across the street but decided it wasn't wise (we were teenagers so we still contemplated the idea). We also fancied the idea of drilling a hole between the two houses: my neighbor's father was fine with it and provided us with a really long drill but my parents didn't want to hear about it.
And we didn't have the proper dedicated BNC connectors crimp tool though (and had zero idea where to buy one and were too broke anyway), so we became masters at building coax cables with BNC connectors by using regular pliers.
Them little orcs and humans from Warcraft 2 would then transit through the coax cable from house to house (so around 1995).
tolkien-ring network
First college computer lab was coax 10base2. Got introduced to MUDs on that one. And yeah, coax was unwieldy to work with, but we weren't pulling cables out on accident (at least not without taking the whole case with it).
TokenRing was already an academic curiosity when I was getting my CS degree in the mid-90s. The CS department by that point had already switched to twisted-pair Ethernet. By the time of my first job in the late 90s, we used FDDI for our web farm's backend network, but we tore that out with the arrival of gigabit, so FDDI didn't last too long either.
Token Ring topology was included in Microsoft's Networking Essentials certification at least as late as 2000. I started my first tech job in 1998 at a large bank and they still had production token ring networks.
I worked with Prime minicomputers when I was a late teen. They introduced RingNet, a token ring product, in 1979.
Fun fact: each system had an IN and an OUT coax port, and the coax cable was about as thick as a thumb. If a system died, a small hardware box short circuited, tying the IN and OUT coaxes together. This kept the ring intact, but it would also make that segment of the network longer and there were strict distance limitations between nodes. If the distance got too big, the entire ring broke. Also, if any coax cable got damaged, that also lead to a ring break. Prime had a special diagnostic program they could use on multiple nodes to try to find ring breaks.
Token-passing networks were interesting for hard real time system analysis, as they featured deterministic access control instead of collision detection. They made possible extending Rate Monotonic Scheduling theory to the network layer as well. The priority work (priority ceiling protocol, multi level token priority schemes) survive today in token bucket rate limiting.
Interestingly, the CAN bus features deterministic access control through collision detection, using a bus topology rather than a ring topology. I've had bad experiences with CAN, but I believe they're due to hardware bugs and/or driver bugs rather than the underlying design.
Arbitration in CAN is still one of the coolest things I've seen at the hardware layer.
Being able to handle a collision without needing the higher priority message to skip a single beat seemed like magic to me when I first learned about it
the token bus combines the one-user-at-a time of a bus with the potential-token-loss of a ring.
IEEE 802.4 token bus standard was withdrawn
polled buses with a master are pretty common when you don't mind the performance hit and central-point-of-failure. i believe cable networks work this way.
Interesting, I'd never heard of token-bus networks, or IEEE 802.4, which was indeed a token-bus standard.
CAN isn't a polled bus, doesn't have a master, and doesn't impose a performance hit. However, on a bus network, every device on the bus is a single point of failure: if it starts jabbering, the network stops working.
everything can be deterministic until you lose the token.
YouTuber clabretro has numerous videos installing and configuring token ring:
https://youtu.be/4UWibKEjyY8?is=rpvYv31UIB3DLIep
I love watching this channel as a 20-something who never had a chance to see/touch the hardware he's working with. Great way to show the real-world applications of this tech rather than just reading historical summaries in textbooks.
Some of it is very telling about how close we were to having a completely different tech ecosystem. Have you seen his video on Sun thin clients? They had this tech working back when PCs were $2k a pop. They had this cheap thin client that had all the peripherals and authentication, plus a very rich java runtime and libraries, but due to mismanagement the tech never got off the ground.
IBM POWER is another odd one. They are still going strong because their shit Just Works so anyone doing Serious Business just keeps using them. They have hot pluggable CPU and ram ffs! Compared to cloud computing they are tiny, but they aren't going anywhere anytime soon.
> Coax connectors for Token Ring did exist, but were more common on Ethernet networks.
I don’t disbelieve the author here but anecdotally the only time I’ve worked with coax BNC plugs for networking is with TR. And it be never seen a TR network in the wild that wasn’t coax.
It’s interesting to read that other cable types existed for TR.
Conversely, the only time I've worked with coax BNC plugs for networking was with 10BASE2 ethernet (aka "thinnet" and also "that stupid broken thing").
That was just one long, shared transmission line that happened with RG-58 coax. Each end was terminated with a 50 Ohm resistor. There was no central brain-box or hub. All nodes were peers.
Nodes could be inserted on an existing run by bisecting the coax at any random point, terminating the bare ends with BNC male connectors, and plugging those into a BNC T connector which ultimately connected to a NIC. Doing this brought down the entire network segment in unison until it was reconnected.
In fact, any discontinuity in the transmission line brought down the entire network segment.
So when Bob in Accounting decides to move his PC to a different spot on his desk and breaks the transmission line? The network dips out completely, and without any good tools for a IT folks to pinpoint exactly where that break happened.
Fun times. (Except: It actually was kind of fun for homelab-type stuff when 10BASE-T twisted-pair networks were still kind of expensive. Adding more nodes just meant installing more connectors and maybe extending the line. Now we just add another magical non-blocking N-way $15 gigabit network switch wherever, or skip it all and use wifi.)
---
I've never worked with Token Ring at all. If I ever even saw it in the wild, I did not recognize it as such.
I do remember the flame wars about it, though.
Thinnet is known as 10BASE2
https://en.wikipedia.org/wiki/10BASE2
The "T" in 10BASE-T signifies "twisted pair", which isn't the coaxial in thinnet.
Yeah, it is. I even Googled it to make sure I got the hyphenization (or lack thereof) right, and then I still successfully screwed up the actually-distinguishing part.
Corrected. Thanks!
I think we spent more time in the computer lab in middle school running NET SEND and then imperceptibly loosening the coax on the back of our machine to take the network down than doing the actual exercises we were supposed to do
Thanks for sharing. That sounds a lot like my experiences as well so now I’m wondering if I’m misremembering those networks as TR (it was a generation ago).
I think you are misremembering.
Coax Ethernet with BNC connectors was absolutely the norm for LANs from the late 1980s to mid 1990s. Then, slowly, UTP took over, as 10base-2 was replaced by 10base-T, then 100base-T, then switches instead of hubs, then Gigabit Ethernet: 1000base-T.
I worked with Token Ring in the '80s and never ever even heard of a coax version. It was D9 connectors on the card to fat IBM connectors on the MAU, as pictured here:
https://en.wikipedia.org/wiki/Media_access_unit
I think you're switching TR and thinnet Ethernet in your head.
All the tokenring in use at the bank I worked at in the mid-90s was twisted pair.
I faintly recall that some of the older systems used the 9pin connectors.
And the first home ethernet I set up was BNC.
Short version:
Token Ring was a networking alternative to Ethernet.
With Ethernet when a node wants to transmit it just does so, then backs off and retries if there was a collision (someone else trying to transmit at the same time).
Token Ring avoided collisions by passing a "permission to transmit" token from node to node in circular (ring) fashion.
> With Ethernet when a node wants to transmit it just does so
Depends on the Ethernet variant. Older versions have collision detection as you describe. More recent versions are point-to-point full-duplex and thus collision free, with the outlier of 10base-T1S, which uses a virtual token ring for collision avoidance.
Token Ring functions a bit like a ham radio net in my understanding
A very nice trip down memory lane. The big banks bought token ring because nobody gets fired for buying IBM, and Ethernet looked so amateurish compared to the heft of TR cables and chunky spaceship MAUs. Hermaphrodite connectors, remember those? I knew a bank that ran TR at 4mbps, connecting its head office buildings nearby and was considering an upgrade to 16 - it seemed mysteriously fast in those days. But then Ethernet came along, and SCO Unix servers on 486, so it was relatively easy and affordable to get a department of 30 people up and running with PCs as terminals. Those were the days…
> and unlike Ethernet, all the cards on the ring had to talk at the same speed.
Essentially that was the same for Ethernet too, everything on the same leg needs to talk at the same speed, though IIRC a different speed device wouldn't take down everything, it would just cause a significant excess of collisions due to the out-of-sync cross-talk and not be able to talk to anything itself. It wasn't until switching hubs and full switches were around that you could mix speeds on the same leg (with a switching hub you'd actually have two legs, one at each speed, bridged, with a full switch you have a separate collision domain per port so can mix speeds quite freely).
Right? I remember struggling with this, especially with multiplayer lan gaming
Oh, flashback to learning queueing theory decades ago!
In principle, token ring was faster than Ethernet (which was CSMA/CD in the day): the mathematics of token ring were exact and analytic, and, when solved, gave higher speeds and better reliability than Ethernet, the mathematics of which were so complex (think three-body-problem) that analytic solutions were available only for bounding equations, which resolved to slower than token ring.
In practice, Ethernet was much faster.
A classic "the model is not the territory" problem.
[flagged]
Related: vampire tap
https://en.wikipedia.org/wiki/Vampire_tap
There's a Minecraft YouTuber called TangoTek who's building a deck building game called Decked Out 3 in Minecraft. That is using a token ring system to control events and messages sent in the game to avoid conflicts [1].
[1] https://www.youtube.com/watch?v=R18APaR7mTQ at 28:54.
Better be careful, or the IBM lawyers will hunt him down for patent infringement. Maybe he should preemptively license it to be on the safe side.
https://fridayfun.net/dilbert-1996-05-02
This link works better for me: <https://dilbert-viewer.herokuapp.com/1996-05-02>
Ah old Dilbert. Shame the author went crazy.
I just read through his Wikipedia entry and didn't see that he'd gone ''crazy'' unless you mean his political views? He's dead now anyways. I linked to that comic because I had bosses like that way back when.
>I just read through his Wikipedia entry and didn't see that he'd gone ''crazy''
If you want to see how Scott Adams has gone crazy, watch e.g. Behind the Bastards episode(s) on him:
https://www.youtube.com/watch?v=8nyEkHqP65c
I think you mean "went" and I wish there was a very brief plain text summary rather than a long video that I don't have time to watch. Anyway, I enjoyed the token ring comic and don't have strong feelings about the cartoonist now that he's gone.
Summary: Something that coulda been really cool & awesome, if not for crappy decisions made by the Dark Lord...er, I mean IBM.
Lot more details: https://en.wikipedia.org/wiki/Token_Ring
Token Ring was always a bad idea, even before IBM screwed it up. Tanium still does this crap and it's awful.
At a 2600 conference long ago, we built a cluster of public X11 dumb terminals, so hackers at the conference could browse the web and such. We hardened the server against local user attack, and the clients booted from a read only solid state media (aka, floppy disk) with no port access. But how do you harden a physical network against a thousand hackers? Solution: token ring. We made the network so ancient that nobody would have the hardware to hack it that weekend. Worked great.
I feel like I learned about this in every tech class I’d taken since the 90s but never once saw one in effect.
I only ever worked on one token ring network in the 90s - the client used mostly IBM hardware, and I guess that's what the consultants who set them up installed. It worked fairly well, but the main thing I remember was the switches were really, really noisy.
Every other network I worked on was ethernet, with the exception of one arcnet network. Which also worked really well, except it was very, very slow.
When I started my first job in the late 90s, they had Linux floppy disk based thin clients running over coax token ring.
The cable was laying right under my desk. When I touched it with my feet, there would be a disconnect because the coax t-connector was not sitting firmly enough on the network card of my device. All unsafe work would be lost.
Fun times.
Now we have time slicing for high reliability Ethernet (IEEE 802.1Qbv). I have never worked with it and don't properly understand it, but I think that instead of passing a token around, there are time windows for high priority devices to transmit.
Low priority devices must stop transmitting in the window, and switches will interrupt any transmissions not allowed.
Fun fact:
When I was an intern at Merck in the late 1990s, the networking team was switching over from Token Ring to Ethernet.
It was so much work that they were openly recruiting anyone who was an employee or intern to come in on a weekend and help out. They eventually offered to pay people extra and a couple of my intern buddies took up the offer.
Novell over token ring was pretty popular back then
The first time I read about this was about 35 years ago. I read about it in a computer science textbook my father owned. I was just a child so I never thought much about it, except that the name was pretty cool.
That textbook was sort of system textbook which describes a lot of low level stuffs such as networking, hard disks and such.
There is also a Tolkien Ring network, who has the One Ring, talks all the time.
It was on the GCSE Computer Science curriculum in England until very recently!
IBM were still using them in 2004.
https://web.archive.org/web/20261016122640/https://dfarq.hom...
FYI cable networks (upstream) are essentially token ring.
In fact, broadcast ethernet is essentially token ring, except instead of a token it just grabs the wire. Only one machine can grab the wire at a time. Switches hide that from most people.
No, DOCSIS is nothing like token ring. It doesn't use tokens, it uses channels for up/down and TDMA/CDMA.
No, 'broadcast ethernet' is not token ring. There is no token. It's not a ring. Switches do hide the details, but it's hiding collision domains, not token passing.
Why is the title written in the past tense? ARC156 is still being used in 2026, Automated Logic’s WebCTRL BAS system uses it for communication.
https://www.automatedlogic.com/en/media/AAR_CS-aug2023_tcm70...
The last token ring standard appeared in 2001 and there has been a huge amount of network-technology development since then. Tokenring is tiny and incredibly niche. It may as well be dead.
Ethernet got its latest in 2022 and another is in the pipeline.
I was something I learned about in college and then never encountered again. Until today.
Dang, this took me back to the Artisoft LANtastic vs Novell Netware days...
I really did not think I would hear about token ring again
Can't wait for some articles on ATM, X.25, and Frame Relay. :)
When I started working at Network World magazine in the late 1990s Token Ring was still a technology in our coverage area.
It didn't last. Gigabit Ethernet began popping up around the time of the dot-com bubble bursting. That was it for Token Ring.
I find this guy kinda annoying. I mean, yeah...seems like he did some networking back in the day, but everything he posts is just enough off (or ia Google-correct, not Been-There-correct) that I can't quite take him seriously.
The main reason Ethernet won the war was because it was cheaper.
He should have stopped there. That was it. And because it was cheaper, less complicated to build the chips, 'good enough' networking, a bunch of folks jumped on the bandwagon. And here we are.
If you had a 16 megabit network and brought up a system at 4 megabits, it took down the whole network.
Well, yes, everyone who has ever mentioned TR tells some variation of this. And if you left the terminator off of the coax ethernet, you brought the whole network down. But TR got smart MAUs and ethernet got UTP hubs around the same time and problem solved.
And hubs were a lot cheaper than smart MAUs. Sooo...
Ethernet was junk if you went with the cheapest options all around, but you could make it acceptable if you invested some of the savings in improved technology.
Just like TR. If you were deploying dirt cheap 8228s after 1990 or so, your network still sucked. You didn't do that.
But the other reason Ethernet won was because of scale.
Yes, there was a thing called 100Mb Token Ring. Someone might have even shipped some number of them. I never saw one in the wild, and I don't know anyone who claimed they had who didn't work for IBM or Madge (which, btw, wasn't an 'obscure' network company at the time and made more than just TR cards).
That's because basically everyone who did Token Ring (IBM most of all) had decided that you wanted to scale TR you were going to ATM (for a bunch of mostly political and marketing reasons). Which, unlike TR, was in reality a very bad idea, but that's another story.
I've never even heard anyone even suggest 128Mb Token Ring, and I suspect someones LLM is hallucinating. There was a gigabit Token Ring standard (IEEE 802.5v), but I feel pretty confident saying noone ever shipped it. Like I said...ATM was gonna be the future.
Ethernet used a 15-pin connector that looks quite a bit like a PC joystick connector. Token Ring used a 9-pin connector that looks like a 9-pin video connector or an Atari style joystick plug.
Ok...I'll just be snarky and point out there are ethernet cards that use a 9-pin D-sub connector to reuse a TR wiring physical plant. I have several.
Coax connectors for Token Ring did exist
Citation needed. I've seen coax for 802.4 industrial networking stuff, which is not TR. I'm betting more hallucinations.
ARCNET was another competitor from the same timeframe. It, too, was a contender for a while because it was backed by Datapoint, a pioneer in office automation using small computers.
Arcnet was a contender because it was even cheaper than ethernet and it was (and still is) widely used in industrial environments. And some popular early PC file sharing systems (e.g. LANTastic) would use Arcnet.
> Datapoint, a pioneer in office automation using small computers
In 1970, Datapoint created a desktop computer, the Datapoint 2200, with an 8-bit TTL processor. Intel created a microprocessor version of this processor for Datapoint, called the Intel 8008. Datapoint decided to stick with TTL, so Intel ended up selling the 8008 as a general-purpose microprocessor. Intel improved it to make the 8080, then made a 16-bit version, the 8086, as a stopgap product until the iAPX 432 micro-mainframe was ready. The iAPX 432 was a disaster and that's how you ended up with x86.
Ken...I know you tore open an Intel 82586 ethernet chip and an IBM Tropic TR chip. It might be interesting to do a comparison of the complexity of those two, being that they're roughly contemporary.
Yeah, when I saw 'Datapoint' I almost went there too. :-)
> Coax connectors for Token Ring did exist
Apollo Domain computers had a token ring which used coax. RCA phono plug connectors. We had a pair of those at the aerospace company for testing. Worked OK, but too weird. We ended up with Sun workstations.
At various times in my early career, I was involved with strange networking hardware only of historical interest now. Apollo Domain (sort of like Unix, but not.) "Thick Ethernet" 10BASE5. Later "Thin Ethernet" (Yes, you can mix thick and thin Ethernet without an impedance mismatch, but you may need to use a time domain reflectometer to check. The adapters were built to tolerate TDM pulses.) Hyperchannel, 20Mb/s which required 6U of rack space for the interface. (We had a pair, surplus from another project, but didn't do much with them.) DECNET triaxial cable for connecting things called the DMC-11.
These were all coax, and they were all true "baseband", that is, the signal was meaningful down to DC. No differential signals, and no scrambler to make the number of 1 and 0 bits roughly equal. You can't use decoupling capacitors or transformers. They were all susceptible to noise and ground loops. The coax shield should only be grounded at one point for the whole chain. Otherwise, noise in the ground comes through as noise in the cable. All this was no problem in a place full of RF engineers, who knew all that, and considered hooking up a TDR or a scope to a data network normal operational setup. And in an RF R&D facility, where copper busbars an inch wide were run in parallel with electrical ducts to solidly ground out noise, ground was solid.
So coax worked fine at Xerox PARC, it worked fine at Stanford (although inter-building coax didn't work all that well), it worked fine in single-room installations, and it worked fine for organizations which understood the limitations of non-differential baseband over coax. The coax technology thus needed too much basic electronics knowledge for general office use.
Modern Ethernet is pretty much idiot proof at the electrical level. It's differential signals over a pair, immune to common-mode noise and ground level differences. If you plug standard cables into standard interfaces, it will work.
Which is why coax Ethernet went away.
I ran a lab of Apollo 3500/4500 era machine in college (thanks for poking that memory). None of them had RCA phono connectors for token ring, but they did have a really weird connector that looked like it had 2 coax(ish) connections, along with some pins, in a D shell. Think something that looks like a 13w3 video connector, but only 2 signals and fewer pins. Are you thinking of an earlier generation of Apollo machines?
We had some early models which had two ordinary coax lines going in. Unplug either and the ring went down. They probably converted to something with one connector, and maybe a mechanism to close the loop if unplugged. That's a basic problem with token ring - anything removed must be jumpered out.
We just had two loaners for a demo, so we might have been given an old model. We were evaluating workstations, and had various devices for a few months at a time.
I saw quite a bit of rare early hardware over my career. Much of it was rare for a good reason.
Yeah, that makes sense; I never saw, for example, a DN100. Apollo machines were kinda weird. Domain/OS was very weird.
I saw quite a bit of rare early hardware over my career. Much of it was rare for a good reason.
Indeed. We should swap stories over a beer some time.
> But TR got smart MAUs
I wish I had a smart mau but she is standing on the windowsill mau-ing at the part that doesn't open while the other side is wide open. In the rain. Which is getting inside while she maus to get inside.
Maybe she's configured for 4Mb and the other side is 16Mb. Check the bridge between the lobes and make sure it's passing...er...traffic. :-)
He is writing about tech since 1991, but I read a few articles today and most of them seem to repeat the same thing 3x in the first paragraph which is a bit off putting. This article doesnt but I agree, its ‘just enough’? Anyway cool to have all this in one place!