I'm sure many of you thought this day would never come, but slowly but surely, I've been working on the October 2004 Top Ten. Under normal circumstances these results would have been announced in November 2004. It is now March 2007. That makes this 28 months late. The September 2004 results were published two months late, and I was mortified by that. This is so much worse.
But after the disastrous email meltdown of October 2004, which caused missing data in both September and October of 2004, I was so disheartened. I tried my best with the September info, but I was still late. After that, after being late for the first time in the history of the email top ten, it just seemed to be of little point. Once it was broken, it was broken. So I continued to work on the reconstruction of the October 2004 email results, but rather than just spending the hours straight through to get it done, I did a few minutes worth each month. Sometimes an hour's worth on a good month. But I let other things take priority. I should have made this an absolute priority from the beginning. If I had done that, this quite possibly would have been complete shortly after the September results, but I did not. There were a lot of other things going on these past two years. But with a little bit done each month... and a bump in priority in February 2007 so I did a bit each day instead of a bit each month... we are finally at the end of that road. The October 2004 Top Ten is done!!
So here we are two and a third years later. Wow, that is a long time.
In any case, I won't repeat now the details of what happened in October, or describe the techniques I used to do as good a reconstruction of the data as I could manage... I described both in the September 2004 results. The only thing I will add was that the manual matching of log files to mails I had through other means was difficult due to time stamp differences between the various sources of reconstruction information I had available. There were a lot of human judgment calls involved. Between this and the log file reliability (not great), I am sure that I both missed some messages for which all evidence is just lost forever, and I may also have inadvertently double counted some other messages.
But even given both those things, I am convinced that this reconstruction is as close as can possibly be managed given the information that survived after the crash. And given the known emailing patterns of most people who send me email, most likely the ordering of the top 10 would not be affected by any missing or double counted emails. In any case, it is the best I can do, and it is time to close this sad chapter of the Top Ten and start catching up to real time.
As for October itself... Rebecca was by far and away the winner... almost entirely due to her effort to forward me copies of emails I lost that she also had copies of... a major part of my reconstruction for both September and October. Thank you Rebecca.
For the most part, the rest of the list reflected fairly normal traffic patterns compared to previous months. Matt falls right off the list after having made a big effort to capture the #1 spot back in September. Ron also slides off the list, although he didn't have as far to fall, having been in the #10 slot.