65 comments

[ 3.8 ms ] story [ 136 ms ] thread
For anyone who doesn't know, IRC users have a _lot_ of inertia. It's common to stay logged in and idle for weeks/months/years at a time. The bandwidth used is negligible, and IRC natively doesn't have backlogs unless you're logged in the whole time.

So I'd assume these numbers include tons of people who just have no idea anything happened yet.

It also includes people like me who are just signed in to both until my last couple of groups switch out.

The Libera users count is probably decently representative of recently-active people, just because they'd have had to take deliberate action to sign in there within the last month.

The Freenode numbers, though...

That can be seen in the volume numbers. Libera seems to already have a lot more chat volume even though it still has only 2/3 of the users.
I used to have a detached screen with irssi sitting on a server for a very long time after I actually had stopped checking in on it.

The only reason it's not still running is because eventually the server that I had the detached screen sitting on was decommissioned.

(Ok, admittedly it was probably rebooted a couple of times between that I stopped attaching my screen session and when it was decommissioned. But the server used to be up for hundreds of days at a time without reboots so the point still stands that I "was" on IRC for a long while even after I had stopped "being" on IRC. I don't even remember anymore if I was connected to freenode or IRCnet or both though.)

I connected to freenode as a tourist because of all this stuff and left myself logged in. Otherwise I haven’t been on for years. I doubt I’m the only one who is just there to see what happens long-term who will drop off at some point soon.
It's kind of funny to think of those users that reconnect to their bouncers in a few years from now, see that everyone else left the channel and think: "WTF has happened?"
I find some dark humor in the fact that someone, somewhere, decided to pay a big chunk of cash for Freenode... and now that same person must be aware there is a website called isfreenodedeadyet.com tracking its inevitable demise, as a direct result of their purchase (and subsequent bad acting).
Truly magnificent isn't it?

It's also a testament to the beauty of IRC. "we don't like the change. So here's a viable alternative".

We could say something similar happened with WhatsApp, but the difficulty of moving everyone to another platform is so much harder.

The difference is that

* IRC users have the technical acumen to move.

* IRC users aren't strangers to setting things up and configuring them, so the switching cost isn't as big

* There's a shared ethos about IRC, but not so much for an app that reaches everybody on the planet. You can convince one demographic much easier than all of them.

Not just "technical acumen" - the clients are designed to connect to multiple networks and treat them equally. Literally nothing about my user experience is different now that I'm no longer logged into Freenode.

Non-technical users have the same sort of experience e.g. switching cell phone providers (provided your area has number portability), switching between subreddits, etc., and those moves definitely happen.

The difference is that

* IRC is interoperable and easy to change servers, unlike apps with proprietary protocols. If Yahoo started doing horrible things with their email, then people who notice would switch to GMail or another host.

On a related note I wonder if many people ended up migrating to Discord instead of to another IRC network.

For example the SerenityOS project used to advertise on their site that they had a channel on freenode but now they link a Discord instead.

http://serenityos.org/

That’s a good point. Honestly, if I had a popular open source project right now, I would have matrix bridged to discord and maybe irc if people asked for it.
Don't try to suck off IRC users too hard... The protocol and clients have been designed with this in mind. it's not like IRC users are so much more intelligent than others...

I feel IRC is better, but it's not because of the average intelligence of users & opers

Awww. Fellow IRC users feeling attacked... What?? We're not the smartest users of all chat protocols because we're using the oldest & most open??? Some people use others because its easier & has the most users regardless of its safety & privacy & our digital choices don't 100% correspond to intelligence?? OMG.
Downvotes on HN are unpredictable at the best of times. This, however? Sure-fire way to get them.
I welcome them in this regard lol. Bunch of hackernews people sucking themselves off? Love it. There is no way to assure a bunch of down votes than to show ycombinator folks that the flavor they insist is that of pure umami is actually that of their own penis. now you have the conundrum... Do I downvote this guy and prove his own point? Or do I upvote him and admit I was sucking myself off with what I believe is the potency of my own mind?
Please don't do this here.
As a friend of mine wrote in a WhatsApp group that we are trying to migrate to Telegram, "which data about me can they [WA] steal that I care about." So she's not installing Telegram.
People did leave WhatsApp. If you believe Telegram's numbers, at least 25M WhatsApp users migrated after the privacy policy change. That's more than 1000x as many people as have left freenode. But still a drop in the bucket for WhatsApp's user base.

It's more of a testament to how niche IRC is that a 20K user migration is significant.

> If you believe Telegram's numbers

I have a hard time believing anything from Telegram because of the owner. But that number does seem feasible.

> People did leave WhatsApp

I don't think people joining Telegram directly leads to people actually leaving WhatsApp. I joined Signal, but am still in WhatsApp because most people are still on there.

It is delicious! Painful too, and unnecessary, but since it has happened, why not gawk away?
It's not really someone, somewhere: it's rasengan, who is here on HN!
pocket change
you caused quite a stir and made a lot of people upset, was it worth then?
heh
The clown prince of Korea, everyone.
hackernews folk.. When they're down voting its the most righteous. When they get downvoted it's never because they're wrong.
Not just a stir, he pissed on a working, useful, helpful thing and destroyed it. It's shameful.
At this point I'm not sure one could take any of his words seriously. That he continues trolling HN users is just a sign of worsened derangement.
Pointing out that it's cheap for you, while you couldn't get it right, isn't a good look.
Youre so pathetic. You steal a network when nobodys watching, steal channels that have no allegience to you or your service, and then act salty when your users leave.

Your reputation before all this is why i never signed up for PIA in the first place, and after this just confirms what a true dunce you are.

You can't buy respect with all the pocket change in the world.

You can make yourself look like a fool though, and you have.

Pocket change you made a huge deal about and called ‘significant money?’
Off topic, but I really love DataDog, I just wish their pricing was more transparent.
Is this as relevant a comment in this topic as me saying I like pizza, or am I missing something?

Btw I used datadog like 5yrs ago when they had a decent free plan for one server and I loved it aswell.

Transitioned to frontend dev recently so I have no use it for now though.

Edit: okay so the site uses datadog charts, now I get your comment.

I like schadenfreude as much as the next guy, but some of these graphs are just plain awful. The registered channels for freenode shows them plummeting, while the actual difference is about 0.5%.
The ones with the big "NOT TO SCALE / EASILY MISLEADING, PAY ATTENTION TO AXES" labels in their titles?
That's interesting - I've never seen a Datadog dashboard used like this. Quite powerful.
I used to run one of the more populated channels on freenode before one of they killed nickserv registration without notifying anyone. As a result I lost chanserv access and the oper/freenode staff member I contacted “couldn’t prove I was the actual owner”. I think that freenode has suffered from chronic mismanagement for years.

I professionally worked with some members in the channel who attested to my identity and true ownership of the channel and I offered to provide a signed pgp message using my registered email to prove that I owned the channel using a domain name that I own which matched all of my information (basically a true cryptographic proof of identity).

Instead of this, the oper made me announce my intentions to “re-take over” the channel to the other members saying that if nobody contested then he’d return ownership. I did this for about four months every week. Then the oper decided to give owner perms to his friend saying that I wasn’t active enough to administer the channel. After his buddy took over the channel, there have been no administrative actions taken (ie no mode changed) because there haven’t needed to be any? It’s almost as if I knew how to administer an IRC channel and configured it in a way where it didn’t need to be touched? Either way it was pretty much a slap in the face after spending years growing a community for their platform. If anyone is a member of this channel and wants to start fresh ##embedded channel somewhere else, happy to do so. I hope others do the same.

Looks like the former staff member resigned recently. Good on you buddy. But you still have a special place on my no fly list :).

If I'm following this all correctly, the former staff member who decided to give your channel to their friend is staff for the New and Better(tm) replacement to Freenode that all the open source projects are being encouraged to move to. You know, the one that is obviously so much more trustworthy because all the previous Freenode staff are behind it.
I'm never going to go on libera.chat... I use IRC networks, but I don't even use Freenode.
If you never used Freenode, why would you? Libera is just good ol' Freenode minus its old trademark.
I used to use Freenode...
Use Libera if and only if you want to communicate with the people who used Freenode. (Actually, if you want to communicate with the people who currently use Libera; there are people who use Libera who never used IRC before, and a some old Freenode users who have followed their communities elsewhere.)
I think that both, freenode and libera will end up dead. I sure hope that Libera stops spamming soon though...
hey freenode staff here. SPAM SPAM SPAM. BUY OUR STUFF. WE CAN SELL YOU CHEAP BITCOINS OR WHATEVER.

notice that just because I said I'm freenode staff... that doesn't actually make me freenode staff.

I don't understand how to read some of these graphs, especially the Globally Online Users Trend. It says "NOT TO SCALE, PAY ATTENTION TO AXES", but the numbers don't even seem to add up.

The Freenode Users line is around ~48k right now according to the hover-card and sits at 48k on the graph axis as expected. The Libera Users line represents ~31k users according to the hover-card but sits at 54k on the axis. I've seen stacked graphs before, but 48+31 isn't 54 so I am not sure if that's what it's trying to show.

Maybe I need more sleep :(

The left axis* has the online user count scale of freenode (46k - 56k), currently at 48k.

The right axis* has the scale for libera (20k-36k) currently at 31.2k.

It's a really confusing graph so maybe another kind of graph could represent this data better.

* fixed!

It doesn't need to be a different kind of graph, it just needs to use consistent axes. It's very weird to have the left and right labels be different when the units are the same.

It'd be one thing if it were two different quantities being measured, but we're explicitly meant to be comparing Freenode to Libera.chat, so they should really be plotted on the same axis.

Oh I see. The right axis doesn't render properly and I assumed it was just some kind of glyph: https://i.imgur.com/t4CeS1u.png
The picture look like maybe the labels on the right axis are rendered with too thick fonts so that they are unreadable.

Do you have the raw data from which the graph is produced (without JavaScripts or pictures)?

The graphs are pretty chaotic. Thankfully the headline one seems unambiguous and is what one might intuitively want to view.
The trend graphs are deleted now. I don't think they were stacked, but dual-axis. It wasn't very clear.
Nice visualization...

Please, which tool(s) did you use for the charts?

It looks like it's only a matter of time until we are looking at the demise of Private Internet Access (a VPN service owned by the person that bought Freenode).