IRC - An example of the power of open decentralized protocol

Posted by Sami Tikka on January 7, 2017

IRC was created by Jarkko Oikarinen in August 1988 in Oulu, Finland. Next year IRC turns 30 years old, so it has certainly demonstrated some serious staying power in the fast changing landscape of internet technologies. What has enabled the protocol (and service) to withstand the pressure from social media and more recently from commercial and more polished alternatives, like Slack?

Unique niche

For starters, IRC seems to fill a certain niche on the online discussion. In corporate usage, social media alternatives like Yammer fill another niche, which is a walled garden board style place intended for sharing and discussing pieces of media, like short texts, images or documents. IRC on the other hand is not structured at all, it’s a free for all chat box having little to do with threads or topics. Facebook enjoys solid usage as social media of choice among all kinds of less informal groups and connections for topical discussion, but it is also intended for topical discussion opposed to IRC.

Lately Slack (proprietary), Mattermost (open source) etc. have emerged as more polished alternatives for corporate users to have private conversations in chat like manner. They are both centralized and rely on central server to handle communications. They are meant for corporate users, who want to be able to control access to the communications. IRC on the other hand is decentralized, you can also set up your own private server if you want and at the same time connect to all the other channels available on the IRC network(s).

Technological staying power

My favourite philosopher Nassim Taleb expanded the original Lindy effect domain, suggesting the staying power of idea/concept/technology can be predicted from it’s current age. If a communications protocol and network like IRC has been around and in use for 30 years, we can pretty safely assume it’s going to be around in 30 years still. Another testament of IRC’s staying power is the fact, that IRC via it’s many networks has not died completely, even though corporate usage has more or less shifted to Yammer/Slack/HipChat/HipsterServiceDuJour. I think IRC has been favored as a communication tool by technologically oriented people, who appreciate the open, decentralized aspect of it and do not require or even avoid polish and bling.

Looking ahead

Another IRC related software from Finland, my favourite IRC client irssi, recently turned 18 years old, and released version 1.0.0 (btw this is why people should adopt a sane versioning scheme like semver). Irssi’s original developer, Timo Sirainen, seems no longer active on the project, and it has been successfully taken over by other developers.

Now that IRC seems to have proved it’s staying power, an effort has been started to modernize the IRC protocol to include stuff like registration and authentication (which is now handled by custom services, for example nickserv in freenode), support for custom message metadata etc. Draft for IRC protocol specification v3 is online at With these modernization efforts, IRC seems to be capable to support custom extensions built on top of the original, solid protocol. Slick service like Mattermost built on top of IRC would be pretty solid competitor in corporate communications space. With support for custom message metadata, even service like Twitter could be built on the IRCv3. Decentralization is at the heart of internet, and I hope IRC manages to stay relevant far into the future by protocol upgrade.