The difference is that IRC is a protocol, not a "social network". You're comparing apples to oranges. Of course centralized companies are going to have more traction than a decentralized protocol-- the proof is in the points about netsplit.de stats reflecting the "whole of IRC" (of course statistically speaking, a few large networks are enough of a sample size to make a relevant judgement about the "whole").

Compare IRC to something it is like. Saying you should add video to IRC is like saying you should add hefty new features to the SMTP protocol. Protocols serve specific purposes, they're not really made to evolve. Evolution in terms of protocols usually means the birth of a completely new protocol (compare Gopher with HTTP). If someone wants video + "IRC", it would be best to implement a new protocol-- that way implementors of the IRC protocol need not be bogged down by implementation details of video support, which they may not care about.

Video could be added to DCC as an optional extension, but adding "optional extensions" to a protocol generally just makes for low adoption rates and bad user experience when having to deal with different clients.


- argv[0] on EFnet #mIRC
- "Life is a pointer to an integer without a cast"