Perhaps, but it’s a network services bot, which is very much a part of most IRC networks. Did you know they connect to the network as a server, not as a regular client like most irc bots?
I think the author, Neozeed, is the person back in the day when I asked them to teach me how to "hack", they instead taught me what telnet was and that there was a larger network beyond AOL. It lead me to what I am today. Thanks.
If I may, I should point out that it says "I’m a guest poster" at the top pof the post. The author is not actually neozeed, but rather me. I'm very grateful for the opportunity I was given.
Maybe I should've made the authorship clearer in my submission, though.
Changed jobs a few times, changed cities, then continents and then finally settled down. But I'm much older than you!
Started a blog over 10 years ago as a place to record my crazy experiments and to remember obsolete stuff. Not that I think many people care about ancient Unix, Novell or OS/2... I never really have much time to go back to the old chat.taucher.net thing... With the wife / kids / work it's hard to find the time.
When Gmail went public I thought we were all doing aliases, meanwhile everyone went with real names for some reason. And some Australian football player with the same name and his rabid fans grabbed all the variations of mine... Oh well.
Ahh, memories -- memories of waiting in a queue to download something only to forget to enable "auto-accept DCC send" and have it time out some time around 2 AM!
Thanks for inventing XDCC :) IRC and online chat was a lot more fun 15-20 years ago.
irc is so useful that i cant believe it is dying, and aside from freenode perhaps even dead.
so what killed the IRC? its not like theres any other alternative out there. slack seems to come close but its still very different and for a different purpose.
Gamers moved from Quakenet to Discord, open source projects moved from Freenode to Slack, and that’s it.
Luckily, most projects are left on Freenode, as an open source project relying on a proprietary service is something few people can stomach (outside of some people that already accepted the commercialization of the net)
We use IRC sort of interchangeably with Slack at my work. The nice thing about IRC is you can host it on-premise and Slack doesn't have all of your company's confidential information. The downside is that it doesn't come bundled with a pretty webapp with GIFs and emojis.
At work we have an onsite Zulip instance which meets both those requirements for us (including SSO integration). It's surprising to me that there aren't more places using it.
Not sure why you're getting voted down. I'm looking for something to replace hip chat at the moment (we replaced slack with hipchat as slack didn't work well for us either), so I'm genuinely curious why a suggestion for riot is unpopular. Anyone care to comment?
What part of slack didn't work out for you? We're using slack at my workplace - and I think it's okish. Not a fan of it being a web app - but not frustrated enough yet to try the irc bridge with weechat (which is on my todo-list. Seems like a sane way forward, the last irc client I used was BitchX, and I don't think I'll likely to trust such an old c codebase with anything if I can help it..).
Now that everyone "knows" what slack is, I wonder if it's time to blow some dust off of Apache (née Google) wave...
The Slack IRC bridge works pretty well, IMO. I use it with irssi. The only negative I can think of is it joins you to every Slack channel you're in every time you reconnect. (Including #general, which you can't permanently leave.)
(I guess the other caveat is: it's a 3rd party hosted application; an internet outage or service outage could take it down, and it's potentially vulnerable to massive information leak in a way that your company isn't.)
Try the wee-slack plugin[https://github.com/wee-slack/wee-slack], it uses the Slack API and I have been using it on my Android phone(using Termux) and my Linux laptop, works great :)
Unfortunately, I suspect Slack use at my company grew largely out of IT's unwillingness to spin up a web-facing IRC solution, and their unwillingness to expose internal services to the web. The latter is very convenient for sales or on-call employees. As a result of that security posture, we instead share tons of company secrets on a 3rd party service. It's kind of a weird local maxima.
To me, that's not a downside but a positive feature. At work, I connect to slack via the weechat wee-slack plugin, and it's great to filter out all that gunk. I want plain text chat, and this way I can still get it.
For me, it is the fact that most people connect with multiple devices, and want to have a smooth transition between them, and to be able to see history on all of them.
There is no easy way of doing that with IRC, which kills it for me.
> There is no easy way of doing that with IRC, which kills it for me.
Yes, there is! This is exactly what Quassel is made for.
Just run the core anywhere (we’re looking into offering it hosted, too), connect with the web ui, the desktop client (Linux, Mac, Windows - uses Qt5, and is only a few megabytes, uses almost no RAM at all. A true native client), or the Android client.
As many devices as you want under one account, a smooth transition from one to any other, shared history (just scroll up - all devices remember where the others left off, and you always see exactly the same on all of them).
With some minor extra effort, you even get nice and fast fulltext search – this video is recorded on a 2$ OVH VPS, searching through 3 years and 6GB of logs in 65ms: https://s3.kuschku.de/public/quassel-rest-search.mp4
I'm glad you are working on that, but "Just run the core anywhere (we’re looking into offering it hosted, too), connect with the web ui," really doesn't count as easy to use -- now everyone on my IRC channel needs to run a server which is is accessable to the internet, worry about security updates, both for the server and your software specifically.
Whereas with slack, I just do "make account, login, done". Until it's as easy as slack, I can't pursade people I work with to move over.
> now everyone on my IRC channel needs to run a server which is is accessable to the internet, worry about security updates, both for the server and your software specifically.
Well, the beautiful thing about this is that you can run one quassel core, for dozens, hundreds, or potentially even thousands of users.
LDAP login is also supported, and I might add later on automated connecting to a pre-configured network.
You can probably imagine that this is very useful already.
Every IRC community I used to be a part of even a year ago has made the jump. It's just more convenient for all concerned- the only holdouts are people clinging to their CLIs and people objecting for ideological reasons.
I wish it hadn't come to this, but IRC's just stagnated for too long. The barrier to entry's too high, using it on mobile devices is a chore, the feature set's extremely limited, etc. Had IRCv3 managed to get off the ground and made some strides towards creating a modern user experience maybe two years ago, Discord wouldn't have had such a vulnerable market open for it to stroll in and snatch up.
Discord provides seamless cloud message logging, mobile notifications, intuitive clients, a huge ecosystem of easy-to-use bots, and integration with games that just works without the user having to set anything up. Even non-gaming-related communities have been switching to it for convenience's sake.
I just wish this wasn't a battle of decentralization vs centralization. Discord seems cool and all but channel operators lack reasonable amount of ownership over their channels and Discord touches every message that comes through its network.
Unfortunately it seems to be an irreversible trend: Decentralized/federated services with many compatible client options are dying and being replaced by many competing and incompatible centralized services, each with few (or one) client options. It's like programmers are all constantly asking the question: "How can we make the Internet worse?"
If E-mail were invented today, there would be 5+ competing protocols, owned by 5+ competing companies, you could only use clients written by those companies, and you wouldn't be able to send E-mail to someone on a different service than the one you were on. Kind of like the current text chat app cluster-fuck we have to live with.
Odd argument. I run an irc network. Quite a large one. My funding is less than 0.00001% of that. It doesn't take much to out compete monetarily. Even if a large amount of your total money goes to marketing.
A community I'm a moderator of recently set up shop in Discordland, due to its superior user management stuff. Both I and another person independently the similarities to IRC; I said that we're re-inventing IRC practices from the ground up.
I don't understand this statement. One finds IRC clients everywhere. One literally has to know just one command at the beginning: /join. And then if you Google "IRC commands" or something like that, you get the rest of them easily (if your client lacks in the help area).
> the feature set's extremely limited
I don't understand this mindset. Why should a program have an extra-large set of features? Is the value of a program measured by the number of features it has?
> Had IRCv3 managed to get off the ground and made some strides towards creating a modern user experience maybe two years ago, Discord wouldn't have had such a vulnerable market open for it to stroll in and snatch up
Apples and oranges. Discord did have serious opponents named Teamspeak, Skype, Mumble and other VOIP clients - but certainly not IRC. And if my memory serves, the first thing they did was to make sure to dislodge those programs from the gamers community.
I’ve been a software engineer for several years now and I still have trouble with MiRC and Adium. I tried Irssi and eventually gave up with it. The problem with IRC is that even the “simple” clients never just work. There’s always some little infuriating quirks that you have to dive super deep into configs to figure out if it’s even possible to resolve without some sort of PR to the repository.
Well, quassel for example doesn’t even have configs, everything is done in simple ways via the UI, and everyone, including the devs, uses that. It’s rocksolid, just works, and if you use the core/client split version, you can also connect a webinterface and mobile client to it.
"IRC is like email. An old boring tool that just does its job. I'm not having fun with it, I'm not feeling special when using, so I need a 'modern' alternative."
Yet I find myself using the web interface of Freenode when I need it, instead of installing a client in my Linux machine with a single command, any of the ones I used for a long time.
> Is the value of a program measured by the number of features it has?
I'd say by the number of features reliably shared by everyone else. This is why XMPP never took off and basically Google killed it.
> Apples and oranges.
Discord specifically targeted communities that used IRC, at least initially. The main difference with the other chat services is that you can enter a room from a browser, so it's almost friction-less to bring someone. IRC natively doesn't support that but many servers could and did (and I think it's the reason IRC remained somewhat popular).
For the record, I don't like Discord for its proprietary and centralized nature (despite calling the communities "servers"). Instead I'll be supporting Matrix and Riot as soon as Riot issue #4488 is fixed.
Having setup and attempted to administer an IRC server, I'd say that's where the barrier exists for someone wanting to create a new community outside of the huge Freenode network.
Which is a) very wordy. B) appears to essentially say “only certain kinds of groups are welcome”. C) doesn’t have a simple way of doing so instead being “reach out and discuss with us and then maybe we’ll let you set up shop here”
> One literally has to know just one command at the beginning: /join. And then if you Google "IRC commands"
If your UX requires needing to google for syntax to type into your program in order to use basic features of the app, there is absolutely no way you are going to take off with people that are used to programs like twitter, gchat, skype, and facebook messenger.
I grew up on IRC and have a lot of fond memories of it. I don't particularly miss all the constant rolling brown-out connection issues and functionality/syntax that differed depending on what server you were connected to. It's no wonder someone came in and easily stole the thunder out from underneath IRC.
Engineer types might have been keen to how transparently most IRC clients exposed the protocol, but I suspect for the vast majority of users IRC was tolerated, not embraced.
I don't think Discord deserves all the credit either, I really think it started with slack being adopted within developer communities and open source projects. Discord got the gamers, Slack (initially at least) got the developers.
I think it may be partly that it was largely created and used by people who don’t care about ease of use and appearances, maybe partly because it kept out the masses. Make it hard for new people and eventually your old users are gone with no one to replace them.
Personally I try to keep most of my workflow in the terminal but I haven’t been able to find an irc client I like and the rooms I join are always silent. I don’t the etiquette and have zero incentives learn
Change aversion killed IRC. A lot of the opers on the big networks are extremely conservative and default to poo-pooing ideas because it's simpler.
IRC with good mobile/push support could have taken off, but using it from a mobile device is painful because of the constant drops and the need to have your device responding to ping pongs every few seconds kills your battery.
> IRC with good mobile/push support could have taken off, but using it from a mobile device is painful because of the constant drops and the need to have your device responding to ping pongs every few seconds kills your battery.
You mean, something like IRCCloud, or what I’m moving Quasseldroid to right now?
The push notification systems used by Android and iOS only allow the same person that can authorize updates for an app to send push notifications to this app.
This means either every IRC app would need to operate an open gateway for notifications, or you’d need to give every IRC server the same certificate that you use to sign your app updates on the Apple AppStore.
Push notifications on mobile devices are designed to only work for situations where the same company controls all servers, and apps.
Random irccloud/quassel anecdote for you: when I started working on what is now irccloud, I was originally just going to write a (mobile-)web front end for quassel core, since I was using Quassel on KDE at the time (and had been for some time). This predates Quasseldroid I think, at least I don't remember there being a good mobile app for quassel back then.
The protocol between the quassel core and clients looked to be some sort of serialized Qt objects (iirc), so I abandoned that idea and wrote the irc client backend in erlang, with a json protocol.
There is still #xkcd on foonetic, #ludumdare on afternet, as well as #lobsters and #proglangdesign on freenode. They may not be active 24/7, but there are good conversationa there if you lurk for awhile.
On the bright side from all the proprietary locked in sell-your-data platforms that were mentioned as IRC successors, Gitter and Matrix are popular and open, and the later is federated. I'm not aware of any major feature IRC has the Matrix protocol doesn't provide, and it has all the niceties of Discord without being the proprietary product of a for profit company.
Oh, memories. We also made a lightweight channel service to take the load off of Q. I think it was a little before newserv came into production. We designed it over generous amounts of whisky in a suburb to Stockholm. It's on github (https://github.com/quakenet/lightweight).
I love trips down the memory lane to, what I call "the old internet" (before facebook etc), like this.
I also had to check if "my" bot was involved somehow. I wrote a C-bot called VladBot in the early 90's, stopped working on it somewhere in '97 but it's seen all kinds of reincarnations (good and bad) since then.
I'm also known as MadHacker, and I'm the original author of one of the other Q variants on the list (Q3). It's good to see people still interested in the history of IRC, especially as it seems to be slowly dying, sadly. I've spoken to the author about the various sets of source code for QuakeNet's services, but the versions linked are the best we can find just now. If there's anyone out there with old copies of any of the Q variants, it'd be good to hear from you.
-- MadHacker@quakenet
77 comments
[ 3.3 ms ] story [ 60.4 ms ] threadMaybe I should've made the authorship clearer in my submission, though.
But I'm just hosting this post, it's not mine. The rest of the blog ramblings are.
Started a blog over 10 years ago as a place to record my crazy experiments and to remember obsolete stuff. Not that I think many people care about ancient Unix, Novell or OS/2... I never really have much time to go back to the old chat.taucher.net thing... With the wife / kids / work it's hard to find the time.
Neozeed .AT. gmail.com
When Gmail went public I thought we were all doing aliases, meanwhile everyone went with real names for some reason. And some Australian football player with the same name and his rabid fans grabbed all the variations of mine... Oh well.
Thanks for inventing XDCC :) IRC and online chat was a lot more fun 15-20 years ago.
so what killed the IRC? its not like theres any other alternative out there. slack seems to come close but its still very different and for a different purpose.
Luckily, most projects are left on Freenode, as an open source project relying on a proprietary service is something few people can stomach (outside of some people that already accepted the commercialization of the net)
Now that everyone "knows" what slack is, I wonder if it's time to blow some dust off of Apache (née Google) wave...
(I guess the other caveat is: it's a 3rd party hosted application; an internet outage or service outage could take it down, and it's potentially vulnerable to massive information leak in a way that your company isn't.)
Unfortunately, I suspect Slack use at my company grew largely out of IT's unwillingness to spin up a web-facing IRC solution, and their unwillingness to expose internal services to the web. The latter is very convenient for sales or on-call employees. As a result of that security posture, we instead share tons of company secrets on a 3rd party service. It's kind of a weird local maxima.
There is no easy way of doing that with IRC, which kills it for me.
Yes, there is! This is exactly what Quassel is made for.
Just run the core anywhere (we’re looking into offering it hosted, too), connect with the web ui, the desktop client (Linux, Mac, Windows - uses Qt5, and is only a few megabytes, uses almost no RAM at all. A true native client), or the Android client.
As many devices as you want under one account, a smooth transition from one to any other, shared history (just scroll up - all devices remember where the others left off, and you always see exactly the same on all of them).
With some minor extra effort, you even get nice and fast fulltext search – this video is recorded on a 2$ OVH VPS, searching through 3 years and 6GB of logs in 65ms: https://s3.kuschku.de/public/quassel-rest-search.mp4
All of this under GPL.
Whereas with slack, I just do "make account, login, done". Until it's as easy as slack, I can't pursade people I work with to move over.
Well, the beautiful thing about this is that you can run one quassel core, for dozens, hundreds, or potentially even thousands of users.
LDAP login is also supported, and I might add later on automated connecting to a pre-configured network.
You can probably imagine that this is very useful already.
Every IRC community I used to be a part of even a year ago has made the jump. It's just more convenient for all concerned- the only holdouts are people clinging to their CLIs and people objecting for ideological reasons.
I wish it hadn't come to this, but IRC's just stagnated for too long. The barrier to entry's too high, using it on mobile devices is a chore, the feature set's extremely limited, etc. Had IRCv3 managed to get off the ground and made some strides towards creating a modern user experience maybe two years ago, Discord wouldn't have had such a vulnerable market open for it to stroll in and snatch up.
Discord provides seamless cloud message logging, mobile notifications, intuitive clients, a huge ecosystem of easy-to-use bots, and integration with games that just works without the user having to set anything up. Even non-gaming-related communities have been switching to it for convenience's sake.
If E-mail were invented today, there would be 5+ competing protocols, owned by 5+ competing companies, you could only use clients written by those companies, and you wouldn't be able to send E-mail to someone on a different service than the one you were on. Kind of like the current text chat app cluster-fuck we have to live with.
an IRC network developed entirely by volunteers can't really compete with that sort of funding
I'd wager that most of the value gained was in marketing.
A community I'm a moderator of recently set up shop in Discordland, due to its superior user management stuff. Both I and another person independently the similarities to IRC; I said that we're re-inventing IRC practices from the ground up.
Now, you can get some of that in IRC via a bouncer, but it's nice to just have it baked in.
Of course, it is a centralized system, with all the issues present there.
I don't understand this statement. One finds IRC clients everywhere. One literally has to know just one command at the beginning: /join. And then if you Google "IRC commands" or something like that, you get the rest of them easily (if your client lacks in the help area).
> the feature set's extremely limited
I don't understand this mindset. Why should a program have an extra-large set of features? Is the value of a program measured by the number of features it has?
> Had IRCv3 managed to get off the ground and made some strides towards creating a modern user experience maybe two years ago, Discord wouldn't have had such a vulnerable market open for it to stroll in and snatch up
Apples and oranges. Discord did have serious opponents named Teamspeak, Skype, Mumble and other VOIP clients - but certainly not IRC. And if my memory serves, the first thing they did was to make sure to dislodge those programs from the gamers community.
"IRC is like email. An old boring tool that just does its job. I'm not having fun with it, I'm not feeling special when using, so I need a 'modern' alternative."
Yet I find myself using the web interface of Freenode when I need it, instead of installing a client in my Linux machine with a single command, any of the ones I used for a long time.
> Is the value of a program measured by the number of features it has?
I'd say by the number of features reliably shared by everyone else. This is why XMPP never took off and basically Google killed it.
> Apples and oranges.
Discord specifically targeted communities that used IRC, at least initially. The main difference with the other chat services is that you can enter a room from a browser, so it's almost friction-less to bring someone. IRC natively doesn't support that but many servers could and did (and I think it's the reason IRC remained somewhat popular).
For the record, I don't like Discord for its proprietary and centralized nature (despite calling the communities "servers"). Instead I'll be supporting Matrix and Riot as soon as Riot issue #4488 is fixed.
https://freenode.net/groupreg
Which is a) very wordy. B) appears to essentially say “only certain kinds of groups are welcome”. C) doesn’t have a simple way of doing so instead being “reach out and discuss with us and then maybe we’ll let you set up shop here”
If your UX requires needing to google for syntax to type into your program in order to use basic features of the app, there is absolutely no way you are going to take off with people that are used to programs like twitter, gchat, skype, and facebook messenger.
I grew up on IRC and have a lot of fond memories of it. I don't particularly miss all the constant rolling brown-out connection issues and functionality/syntax that differed depending on what server you were connected to. It's no wonder someone came in and easily stole the thunder out from underneath IRC. Engineer types might have been keen to how transparently most IRC clients exposed the protocol, but I suspect for the vast majority of users IRC was tolerated, not embraced.
I don't think Discord deserves all the credit either, I really think it started with slack being adopted within developer communities and open source projects. Discord got the gamers, Slack (initially at least) got the developers.
and /password or was that /msg notify something or ??? If you get it wrong, congrats, you've just published your password to the world.
IRC wasn't even very good compared to other cli oriented programs such as the Unix shell or muds/moos.
Personally I try to keep most of my workflow in the terminal but I haven’t been able to find an irc client I like and the rooms I join are always silent. I don’t the etiquette and have zero incentives learn
IRC with good mobile/push support could have taken off, but using it from a mobile device is painful because of the constant drops and the need to have your device responding to ping pongs every few seconds kills your battery.
You mean, something like IRCCloud, or what I’m moving Quasseldroid to right now?
"Something like irc cloud" should have been baked into the ircd/protocol so that there's a lower barrier to entry.
The push notification systems used by Android and iOS only allow the same person that can authorize updates for an app to send push notifications to this app.
This means either every IRC app would need to operate an open gateway for notifications, or you’d need to give every IRC server the same certificate that you use to sign your app updates on the Apple AppStore.
Push notifications on mobile devices are designed to only work for situations where the same company controls all servers, and apps.
Things like this are the reason slack and discord ate irc.
Most users simply won't sign up for 3 services just to chat.
The protocol between the quassel core and clients looked to be some sort of serialized Qt objects (iirc), so I abandoned that idea and wrote the irc client backend in erlang, with a json protocol.
It's planned to be replaced it with protobuf relatively soon, though.
[1] https://xkcd.com/1782/
http://irc.netsplit.de/networks/statistics.php?net=QuakeNet
I also had to check if "my" bot was involved somehow. I wrote a C-bot called VladBot in the early 90's, stopped working on it somewhere in '97 but it's seen all kinds of reincarnations (good and bad) since then.