Times like this make me miss the IRC days, I was just able to reproduce a bug in an semi-open-source project, and Discord went down right in the middle of me sending my findings. Now there's nothing I can do about it. I can only wait.
I don't exactly know what you are comparing. No popular IRC network came anywhere near what we would find acceptable in terms of reliability today. It was an absolutely (wonderful) wildfire.
Netsplits, where the entire IRC network would "split" into two (or more) effectively independent networks because some a link between two servers went down, were extremely common. I don't know if daily, or weekly, but common enough to be perceived as normal and expected in any case.
In the earlier days of IRC, netsplits were sometimes used for channel takeover. If someone was on a split off part of the net where there were so few people in the channel that they could obtain op status, they could kill and ban the "legitimate" ops when the nets joined back together.
The difference was it wasn't one global server for everyone. I think that's why the past feels like it was more stable.
Now, aws or cloudflare gets a hickup and half the internet is nuked.
The old internet was far more federated so doing something else meant to me "Welp, anandtech is down, let's go to pcper, digg, tomshardware, slashdot, etc"
Sure stuff would go down, but it would be just that small community rather than most of chat for the internet.
Yeah, netsplits were really common; nickserv and/or chanserv not working for long periods making popular channels a hell without ops.
I think the centralisation is the issue, I could connect to a different IRC network with a community around the same topic/game. When Discord is down there's nowhere else to go.
I never seen long-running problems in big server-federations like DALNet. Our local "big" IRC servers were generally down for 10 minutes at most. They were not empty either.
Simple services recover faster. Federated infrastructure is much more resilient. We had slower computers, more considerate coders, and simpler software; so everything was snappier, even with 56K modems.
For example, navigate to https://git.sr.ht/~bayindirh/. No scripts, pure HTML. running on a single server. Served instantly.
This is possible. We, as in the world, just ignore it for shinier stones.
Now, a small VPS in an AWS server lapses for 5 seconds, and half of internet is toast. Centralization for the PWN!
There would have been a server split, and half of the people would be chatting among themselves wondering why the some people suddenly got disconnected, while being unaware that there might be a server problem because they can still continue to chat. The other half of the people would think the same.
IRC is distributed and federated. Not only are there countless networks, each network has countless servers, and each group of servers that are up and can see each other can operate on their own, all the way down to a single server, or up to any subset up to all.
When a peering connection goes down and the network splits, maybe some people in the group disappear, or maybe from your point of view everyone else disappears.
Maybe the remaining subset of other users is already good enough because it's enough to continue what you were tallking about and who you were talking with, or if not, you have the option to just try some other servers until you find where everyone alse is. Were "server" is an actual seperate instance of the server software operated by an independant person, hosted on whatever kind of hardware or vm they set up, connected to whatever network they are on, not what Discord calls a "server".
Even if the entire group of say freenode servers goes down somehow (even though that's not really possible) there is still undernet and 400 other nets. Even without prior coordination it would be essentially trivial for the users to all just go looking for, or create on the spot, the same channel on some other net, and basically everyone finds each other again almost effortlessly. And that's if something unbelievable actually happens, let alone the normal minor breaks that actually happen once in a while.
This is entirely different from being wholly at the mercy of the single entity Discord.
Freenode had full-network outages periodically. ddos attacks, infrastructure failures etc. and when those happened, the practical experience was the same... people waited it out. Nobody coordinated a mass migration to undernet or stood up alternative servers for a few hours. (It took much bigger issues - social/organizational/political, not technical - to catalyze the mass migration.)
You're making an argument about the virtues of decentralization - and I agree, decentralization is great! Just in practical reality, freenode (not IRC itself) had exactly the same failure mode as we just saw today.
AWS makes it annoying to be resilient as AZs aren't transparent to their users, so I'm more surprised some were prepared for it.
It seems to me these day people are OK with AWS going down and just blaming it on AWS rather than on themselves for not being prepared for big outages.
"Oh, nothing we can do because AWS/Cloudflare is down"
I predict in the future many will blame the poor overall quality of software and the poor uptime of services on AI, as if things weren’t terrible before AI.
You mean all the AI "slop" that's finding and writing new kernel exploits every day? And submitting hundreds of previously-unknown security bugs in critical software?
Times like this make me miss the IRC days, I was just able to reproduce a bug in an semi-open-source project, and Discord went down right in the middle of me sending my findings. Now there's nothing I can do about it. I can only wait.
I don't exactly know what you are comparing. No popular IRC network came anywhere near what we would find acceptable in terms of reliability today. It was an absolutely (wonderful) wildfire.
Netsplits, where the entire IRC network would "split" into two (or more) effectively independent networks because some a link between two servers went down, were extremely common. I don't know if daily, or weekly, but common enough to be perceived as normal and expected in any case.
In the earlier days of IRC, netsplits were sometimes used for channel takeover. If someone was on a split off part of the net where there were so few people in the channel that they could obtain op status, they could kill and ban the "legitimate" ops when the nets joined back together.
It was so much fun placing some eggdrops on servers that usually split to takeover channels.
I feel like I have different memories about the instability of Internet services in the past than some people do.
Common IRC servers were not without problems. I think it was just more common to shrug it off and do something else until the problems went away.
The difference was it wasn't one global server for everyone. I think that's why the past feels like it was more stable.
Now, aws or cloudflare gets a hickup and half the internet is nuked.
The old internet was far more federated so doing something else meant to me "Welp, anandtech is down, let's go to pcper, digg, tomshardware, slashdot, etc"
Sure stuff would go down, but it would be just that small community rather than most of chat for the internet.
Yeah, netsplits were really common; nickserv and/or chanserv not working for long periods making popular channels a hell without ops.
I think the centralisation is the issue, I could connect to a different IRC network with a community around the same topic/game. When Discord is down there's nowhere else to go.
Ah yes. chanserv and nickserv hiccups were bad. I remember that now, but they were not as catastrophic as outages we see today.
I never seen long-running problems in big server-federations like DALNet. Our local "big" IRC servers were generally down for 10 minutes at most. They were not empty either.
Simple services recover faster. Federated infrastructure is much more resilient. We had slower computers, more considerate coders, and simpler software; so everything was snappier, even with 56K modems.
For example, navigate to https://git.sr.ht/~bayindirh/. No scripts, pure HTML. running on a single server. Served instantly.
This is possible. We, as in the world, just ignore it for shinier stones.
Now, a small VPS in an AWS server lapses for 5 seconds, and half of internet is toast. Centralization for the PWN!
If there was a netsplit you just bunched together on one server. It was more decentralized and a bit more reliant in a way.
How would that have been different in the IRC days?
There would have been a server split, and half of the people would be chatting among themselves wondering why the some people suddenly got disconnected, while being unaware that there might be a server problem because they can still continue to chat. The other half of the people would think the same.
Don't forget everyone flooding both halves with "z0mg net split!"
:P
Netsplits were fun. Especially if you were on the splitting part. You could get to know new people you got stuck with.
> "z0mg net split!" :P
The 00's were an interesting time in internet culture.
Internet slang like this disappeared almost completely once the whole world got access and platforms rooted out all the weird and niche communities.
I mean, when freenode would go down it was more or less the same thing, no?
Not remotely.
IRC is distributed and federated. Not only are there countless networks, each network has countless servers, and each group of servers that are up and can see each other can operate on their own, all the way down to a single server, or up to any subset up to all.
When a peering connection goes down and the network splits, maybe some people in the group disappear, or maybe from your point of view everyone else disappears.
Maybe the remaining subset of other users is already good enough because it's enough to continue what you were tallking about and who you were talking with, or if not, you have the option to just try some other servers until you find where everyone alse is. Were "server" is an actual seperate instance of the server software operated by an independant person, hosted on whatever kind of hardware or vm they set up, connected to whatever network they are on, not what Discord calls a "server".
Even if the entire group of say freenode servers goes down somehow (even though that's not really possible) there is still undernet and 400 other nets. Even without prior coordination it would be essentially trivial for the users to all just go looking for, or create on the spot, the same channel on some other net, and basically everyone finds each other again almost effortlessly. And that's if something unbelievable actually happens, let alone the normal minor breaks that actually happen once in a while.
This is entirely different from being wholly at the mercy of the single entity Discord.
You're arguing against a claim I did not make.
Freenode had full-network outages periodically. ddos attacks, infrastructure failures etc. and when those happened, the practical experience was the same... people waited it out. Nobody coordinated a mass migration to undernet or stood up alternative servers for a few hours. (It took much bigger issues - social/organizational/political, not technical - to catalyze the mass migration.)
You're making an argument about the virtues of decentralization - and I agree, decentralization is great! Just in practical reality, freenode (not IRC itself) had exactly the same failure mode as we just saw today.
im surprised the aws outage hasn't been bigger news today https://www.cnbc.com/2026/05/08/aws-outage-data-center-fandu...
I'm getting Access Denied - You don't have permission to access "http://www.cnbc.com/2026/05/08/aws-outage-data-center-fandue..." on this server. (with an Edgesuite error code)
It was one AZ. Kinda surprised those guys are built in a way where a single AZ failure takes them down.
AWS makes it annoying to be resilient as AZs aren't transparent to their users, so I'm more surprised some were prepared for it.
It seems to me these day people are OK with AWS going down and just blaming it on AWS rather than on themselves for not being prepared for big outages.
"Oh, nothing we can do because AWS/Cloudflare is down"
https://news.ycombinator.com/item?id=48057294
Likely due to aws incident.
Can update to the direct link: https://discordstatus.com/incidents/4hpm4454hxtx
matrix.org alive and well
IRC unite!
It's not a novel opinion but I'm tired of things being fucked all the time.
It's gonna get a lot worse with all the AI slopcode that is about to be pushed directly to production.
I predict in the future many will blame the poor overall quality of software and the poor uptime of services on AI, as if things weren’t terrible before AI.
You mean all the AI "slop" that's finding and writing new kernel exploits every day? And submitting hundreds of previously-unknown security bugs in critical software?
What are you saying here? An LLM helped humans do something right once therefore it's perfect to use in every other situation too?
llms aren't doing any of that. Some smart people are using llms find those vulns, there's a huge difference.
Your definition of "do" seems different than mine.
Can't wait for Cloudflare to go the way of GitHub :)
Isn't consolidation great.