Netrek sociology study

[sheldon@visi.com]

I asked James Cameron the other day if I could get a copy of the logfiles that they had been keeping on continuum to see what sort of data could be extracted relating to the eject command.

I'll be the first to admit, that I am fairly biased on this as I think it's a ridiculous feature that should have never been added to the code.  INL is one thing, pickup is quite another.

Also being a techy geek I have a hard time trying to explain something to someone in a non-arrogant manner which seems incredibly obvious and trivial to me.  So I apologize ahead of time for any argument I'm going to cause in r.g.n. :)

But I would like to see some statistics which either validate or contradict my contention that the eject command is used the majority of the time not to eject abusive players, but to eject newer/twinky players.  

Now getting to the heart of the matter...

Unfortunately the logfiles that Cameron and Tanner have been keeping on continuum do not contain enough detail.  All it shows is a date/time stamp, the individuals userid(i.e. ssheldon@#amic.visi.com) and what was sent.
It doesn't say who it was sent to, and we also do not have any of the daemon generated messages(i.e. who killed who, who entered, who exited and most importantly who was ejected)

 Even so I believe I can generate some semi-useful information from these particular logs.  I'm going to write a short program to parse through them and generate the following information:

 - how many eject commands are issued per day.
 - which userids have sent eject, and tally per id
 - try to mine out the reason for the eject.  On some ejects they might have been 'tagged', but in those cases where they were not I'm going to try to look at the 10-20 message prior and after the eject was issued to identify a reason.  From there can generate some stats on that.

 Cameron mentioned generating better logs with watchmes.  Might be better.  I also had in mind a couple of additional patches to log a couple examples of "abusive" behavior which may justify an eject command.  i.e. tractoring mates off planets, orbiting a third space planet with no activity, blowing up on teammates, etc.  I would have to think those through more.

 Anyway, I'll see what I can generate in terms of results from the current log files, and publish those.  I'm interested in any discussion regarding what the best way to pull statistical results might be.

Steve