A brief history about IRCSTAT
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Version 1.00-000 was created sometime in 1991 and was based on a program 
named "analog" by Kai 'Kaizzu' Kein{nen - kmk@cc.tut.fi. The enhancement 
that I did was to sort out on domains instead of hosts. This was done by 
some regular expressions.
Prior to modifying Kaizzu's program I had allready written a lot of 
shell/awk/sed scripts to count users and domain usage of the ircd. But
I had to do something "new and improved" when the logfiles also included
start- and connectime.

Espen Aneling (then operator/administrator of the ircd in Oslo) found the 
program usefull and gave me a lot of ideas on new features. 

Somewhere around 1.10 the IDF file was added - is was not a Good Thing to
hack the program everythime you wanted to change / add domains.

New ideas where added, a few admins began to use the program and around
1.58 I rewrote the bastard program and it became faster (40%), and version
2.00 was born. 

New features was added and a few new tools (totdays and newdoms) where 
written when I had a some spare time on my hands. Around 2.65 I got
problems with what to name the various releases - up to then I changed
the version number of the mainprogram when I messed with the other
programs or manpages. This was not desirable so version 3.00 was born
a dark winternite in january 1995. Prior to release 3.00 I added yet
more tools: farnear, dycount, timecount and the Configure script. 
Version 3.00 is the most mature release ever (and the best release).
All the programs had manpages, the README file was cleaned up and
there where even included a directory with examples of output!

I must confess that ircstat was more "fun" in the earlier days due to it's 
hackerish nature and lack of proper documentation.

The following persons have had a great influense on the program and I
owe them a great deal: Espen CH Anneling, Nicolai Langfeldt, Vesa Rukkonen 
and Eldad Yahel. Out of these people Nicolai is perhaps the person who
knows the inner workings of these programs as good as I do.

People have asked me on the version numbering and how it's done - esp
when I released the PI (3.14) version of the program. Versions are
numbered as follows: A.BB-CCC. The "A" part is major version-number.
Only complete rewrites will change this part. The "BB" part is "features",
thus when I add a new feature or a new bit of software this number 
increments. The "CCC" part is in fact bugfixes. This number is incremented
since the first "A" release - thus since I released version 3.00-000 the
number 3.15-008 indicates 15 new features/codechanges done and 8 bugfixes.

Bergen, may 31, 1995.

