the reason why

• I know BITNET RELAY since 1988.
• In that same year I wrote a conferencing system for VM, which in its small way somehow followed the principles of today's PSYC.

  ---

• I know IRC since 1989.
• I know its weaknesses since 1989.
• The impact of its weaknesses on the service quality has grown exponentially.
  I've seen it...
  • in the days when you had to use TELNET as a client to IRC from VM.
  • when I wrote a gateway from my VM conferencing system into the IRC. As in those days on VM it was technically harder to implement a client, I had to re-implement the whole IRC server! It was much easier then, though.. just numeric channels, no channel operators.. In the end it turned out being too slow.. heh.. it was written in VM REXX!
  • during the IRC war in 1990, as I attempted to run a server on the liberal side of the IRC.
  • as I installed the first two IRC servers in Italy and spread the word of IRC. I did the maintainance of the IRC server in Rome in the following years and was involved with the european coordination and routing until the EBIC was founded.
  • while being IRC administrator at the University of Oldenburg since later in 1990.
  • while writing and running the only ever IRC to BITNET gateway (which behaved like a BITNET RELAY on the bitnet side, only a little nicer). Had to take it off though, because of its resource consumption. But it *could* be fixed now, with the newer VM releases.
  • while working on ircII:
    • I introduced the CTCP ACTION protocol into IRC, which is the one that lets you do things like:
      • LynX is really hungry now.
      The commands to send actions with I named /me, /describe and : (in ircII command mode).
    • I introduced the ircII scripts which are shipped with every ircII.
    • Several other things about ircII.
  • while writing the rxIRC client to IRC.

  ---

• As MUD administrator (first ChatLand, then Nemesis) from 1990 to 1993 I've seen the advantages and disadvantes of using MUD as conferencing system, as well as it's suitedness for object-oriented handling of networking. I specially added support to the MUD driver for that, and interfaced the MUD into IRC.
• Then I have learned about WWW and HTTP, which made me think of using URLs for people - allowing for static addressing which is extremely contrary to IRCs approach - and redirections to solve the user location problem, which then arises.

  ---

I have been planning IRC protocol replacements since 1989, since I first glanced at its protocol spec, but never got down to actually do it, just like everyone else in the field.