A letter to my friends, May 31, 2000.

My two weeks at the United Methodist General Conference

I've promised many of my friends a recap of my time at the General Conference of the United Methodist Church in late April and Early May. I've put my thoughts to keys and bits and hope that you wont mind this impersonal newsletter kind of format. Many of you will not be interested in parts of this missive. I won't be offended if you skip the parts you find boring. But if I tried to write to each of you about the parts of conference that each of you may be interested in, I'd probably not be done by the time I got to General Conference 2004 in Pittsburgh, PA.

You can find a brief history of my involvement in the Petitions Entry and Tracking System (PETS) and also some information on the history of my connection to the General Conference at http://www.angelinn.net/john/petsinfo.htm . You can also find the General Conference website at http://www.gc2000.org . The 17,474 web pages that Susan and I created to track the activities of the general conference start at http://www.gc2000.org/pets/default.asp . You can also look at http://umns.umc.org/gc2000news/stories/gc034.htm for a story about the computer use at the General Conference that includes some information about me along with a quote or two.

General Conference:

I was pretty sure I was going, but I hadn't seen any airline tickets yet. I called 1 week before the conference and found out that the conference travel agency had "messed up" a large number of all the domestic tickets. The GCFA (General Council on Finance and Administration) had their hands full trying to get all of the problem tickets resolved. They got a price of $1800 for my round trip ticket on Continental. They asked if there was anything cheaper and got a $900 price for a 19 hour San Jose, Los Angeles, Las Vegas, Dallas/Ft. Worth, St. Louis, Cleveland nation wide tour. Thank God that reason overcame cost concerns and four days before the conference I had my tickets.

I began my two weeks at General Conference by running late to the airport. I was only mildly late until I stopped at a McDonalds for breakfast. After watching the third customer complain about getting the wrong food, I should have just walked out the door, but my mind was on the conference and my stomach was on empty. I was very late when finally I jumped in the car with a big bag full of... the wrong stuff.

I've sometimes found that a horrible start can signal a wonderful trip. So I was hopeful about the whole General Conference when I got to San Jose International and discovered the huge wait in line at the counter was because my flight to Houston was canceled. (Not the second hop from Houston to Cleveland, just the first one.)

The airline that sold me my E-ticket, Continental, had booked me on an America West flight for that first hop and they told me that I'd have to get tickets for the entire round trip printed at the America West counter before they could even try to book me on a different flight. They did mention that there was no real hurry because they didn't have anyway to get me to Cleveland until the next day anyway!

Almost 2 hours later, I got to the front of the line at America West, and getting a clerk that had just been yelled at in a very impolite way, I decided to make light hearted small talk. She seemed very relieved that I didn't seem intent on making rude judgments about her intelligence, sex life or friends. She very quickly explained that her airline had rebooked my travel on TWA. She printed out my tickets and sent me off to my third airline check-in line. On the way, she gave me a coupon for a free meal at the airport, a 5 minute calling card, an information packet and a customer comment card.

There was no wait at the TWA line, and I found two smiling people fighting over which would get to help me! I found that my new ticket was for a completely full plane to St. Louis. The very nice lady that had won my business decided to help the gate out by seating me in 1st class right away so they wouldn't have so many people to shuffle later. When I asked, with my best innocent expression, whether I'd have that same seat all the way to Cleveland, then she talked to her supervisor, looked at how much I'd paid for my tickets (full price because the were purchased only 4 days before) and seated my in First Class on the second leg of the trip as well!

Ah, there's nothing like being pampered in First Class to change your mood for the better. I didn't even mind when the student jetway driver took eight minutes to get the jetway lined up with the parked plane in Cleveland. No joke! The jetway moved in, then out, then left, then rotated, then right, then in, then out, for a really long time. Eight full minutes doesn't seem like a long time when you are roasting a turkey. But when you are standing in the isle of an airplane with no air moving, listening to cheerful PA announcements about it being just a couple of minutes more, eight minutes seems like a very, very long time.

The pilots were furious. They'd made up time after being held in a taxi-way queue in St. Louis, and had actually pulled into the gate about two minutes early. But on time landings are measured by when the first passengers step off the airplane, and our fist passenger was six minutes late.

The next day I got to the convention center and found the local computer rental people frantic about getting the network running and all the PCs and printers in place. It seems that they had been told over and over that they would not be allowed to unload their trucks themselves. It seems that the local teamsters union would suffer irreparable harm if they allowed someone else to ever unload a truck. So the rental folks sat for four hours waiting for the teamsters to unload the trucks. After several escalations to the convention center management, the union guys finally wandered by and announced that they didn't have time to unload the trucks and that the computer rental people should do it. The rental people explained that they had not brought enough people to unload the trucks because they had been told, several times, that they wouldn't be allowed to do it. So finally, the teamsters unloaded the trucks, but piled all of the equipment in the lobby instead of taking it to the various rooms throughout the three city block complex.

The network setup was delayed about 1 full day because of this union vs non-union mismatch of expectations.

Still, I got the server, backup server and web server running on Saturday, although the network wasn't fully functional until Monday.

Tom and Richard from the UM Publishing House had volunteered (or been volunteered) to be my assistants with PETS and the PCs this year. They arrived on Sunday and soon had all of the PETS PCs gathered into our make-shift training room and loaded with software.

I'll skip several discussions with the house electrician, but finally we had some holes punched in the walls so we could run extension cords into the adjacent rooms and get all of the PCs and half of the printers running that we needed for our training class.

Training went pretty well on Monday afternoon and by Monday evening we had the PCs back in the Legislative Committee Rooms. I went to bed Monday feeling very optimistic about the operation of the PETS system and my other activities.

The General Conference officially began at 2pm on Tuesday, and by then the folks in my office had figured out that we had a big problem.

Most of you know that PETS stands for Petitions Entry and Tracking System. And most of you know that any United Methodist can submit a petition to the General Conference. Probably, many of you also know that the General Conference is obligated to act on each and every one of those thousands of petitions, once. And only once. That is why PETS is so helpful. It tracks all of the legislative activity having to do with each petition through the course of the conference.

What few of you probably know is that the original petitions (the actual letters received by the Petitions' Secretary) are handed out to the elected leaders of the Legislative Committees at their training session on Tuesday Evening. A second copy of each and every petition is held by the Petitions Secretary in case something happens to the working copy while it is in the hands of the committees. The text of each petition is also printed in the Advance Daily Christian Advocate (ADCA) which is given to each delegate and alternate, and is available for sale to visitors of the conference. Finally, the text of each petition, as published in the ADCA, is available on the web site www.gc2000.org/pets.

I knew that there would be a lot of work to convert the PETS web writer module to make the petitions fit into the 2000 web site. It looks much, much different than the 1996 web site, and uses Active Server Pages (.asp) instead of flat HyperText Markup Language (.htm) files. So I was working on web server changes when other people who also work in the Secretary of the General Conference's office started finding severe problems with some of the petitions.

By 2pm on Tuesday it had all of our undivided attention.

By 3pm we characterized the problem with the petitions like this: About 10% of the 1900 petitions either did not match the cover sheet (which is printed by PETS) and therefore didn't match the contents of the PETS database, or they didn't match the text printed in the ADCA. These were not minor typos, these were cases where the text in the ADCA was about a completely different subject and a completely different paragraph of the Discipline than what the petition addressed! No one involved in finding these problems could even believe that such problems could exist. The more we looked into petitions with problems, the more petitions with problems we found.

By 4pm there was a crises. Several of the core team (Gary Graves, former acting Petitions Secretary; Susan Brumbaugh, webmaster; Randall Partin, Assistant Webmaster and Research Lead; and Gere Reist, Coordinator of the Calendar) had independently come to the conclusion that a complete audit of the petitions would have to be undertaken before we could hand them out to the Legislative Committee Secretaries. But it was worse than that. Elana, the person who had been acting as Petitions Secretary for the previous year, refused to help with a review of the petitions. She announced that it was too much work, and that we should just explain to "everyone" that a "new person" had put the petitions together and made some mistakes. Then she walked out of the room.

The team was floored. Here is the person who is supposed to be in charge of the activity, and she just decided to not even help other volunteers to try to fix the problems she had caused. I went to update the Secretary of the General Conference, Carolyn Marshall, who was on the floor for the opening worship service. When I walked back into the room I was confronted with several people who were very upset at Elana's attitude and comments. It was clear that something needed to be done to fix the leadership crises before we could even start to fix the petitions crises.

With Carolyn's permission to speak on her behalf, I walked into the Daily Christian Advocate (DCA) office and had a quick talk with Rich Peck, the editor of the DCA. Rich is also the person who had hired Elena as a contractor a year before to fill in for the elected petitions secretary, Sheila, whom he had moved to other assignments within the DCA. I told Rich about the kinds of problems we were seeing with the petitions, I told him we expected about 200 of them to be wrong, that we were going to have to conduct a full audit and finally conveyed Elana's attitude top him. Rich was very concerned that we not hurt Elana's feelings. I gave him my opinion that the conference secretary's office was not a self-esteem workshop and that we needed competent leadership to get through this crises. I told him that we needed Elana's full commitment to help fix the problem or that we needed a resignation, and that we needed it by 6 pm.

Within 20 minutes Rich explained that Elena would be needed to work full time in the DCA office, and that we'd need to provide leadership from the conference secretary's office to fill the gap. I thanked him and went back to the work room to explain the situation. I named Gary Graves as acting petitions secretary (again -- he had filled in for the previous petitions secretary who had become a full delegate and couldn't do both jobs in 1996.) Some people were disappointed that Elana wouldn't have to face up to all the problems she caused. But I was very glad to have her out of the way whether or not she ever admitted to her part in our disaster.

Like most disasters, this one went from very bad to much worse. By the time the committee leadership training happened at 8pm we had to announce that we would were going to have to do a full audit, and that we'd not be able to give them any petitions until their first legislative committee meetings at 2 pm the next day. We also told them that we'd get all of the petitions to them by the end of the next day. We didn't know it at the time, but we were lying.

We started finding more and more amazing problems. Petitions were attached to each other and under a single cover sheet that weren't related at all. We found about 30 petitions that shared only the word "retirement" in common. Some were asking for mandatory retirement, some against it, some changing the age, some about ministers, some about bishops, some about judicial council members, and all were stapled together under one petition number as if they were duplicates of each other. We started noticing that petitions were missing. On one of them the document attached to the cover sheet was an e-mail message saying that three documents were attached and asking them to be assigned petition numbers. No sign of the three documents. Then we started noticing other missing petitions. Between the seven of us and our knowledge of petitions sent in by our own annual conferences, we found that 20% of them were completely missing.

We quietly started telling annual conferences to check the ADCA for their petitions and resubmit them to us if they were missing. We didn't want to broadcast that kind of announcement very loudly because we were afraid that we would start to get new petitions from all over that had never been sent in, and we were already buried.

Then the bottom fell out. We discovered that the two supposedly identical sets of petitions (the working set and the reference set) bore little resemblance to each other. We hadn't figured that out yet, and we were actioning the problems on only the reference set. We were then printing new cover sheets for the petitions that had passed the audit or had been fixed, having volunteers attach the cover sheets to the working set and sending them out the door to get the committees started with their work. It was devastating to discover that about 10% of the 1100 petitions that we thought we had completely fixed had not been identical between the reference set and the working set, and we now had some 100 problem petitions floating around through the conference process!

We had to pull all the petitions already assigned to the committees in groups to review and update them, and we had to change our audit procedures to additionally check for differences between the two petitions sets. This was essentially starting the audit over after getting just over half way done the first time.

It took us four days and 525 volunteer hours to get all the petitions fixed, corrections printed in the DCA, corrections made to PETS and to the web site. Some of the committees didn't see all of their petitions until the Saturday that they were supposed to be all done with their work. The total number of problems was: 212 Petitions incorrectly reported in the ADCA or incorrectly entered in PETS. 71 petitions which were incorrectly aggregated with unrelated petitions, 94 switched or incorrectly attached to original cover sheets, 31 that should never have been given petition numbers, and somewhere in the neighborhood of 400 petitions which were never assigned a number and never made it to the conference.

Some problems still slipped through. I'll give you the two examples that stick out most in my mind. A petition with an incorrect Discipline Paragraph reference was routed to the Financial Administration Legislative Committee when it should have gone to Faith and Order. It's intent was to change the definition of what kind of sexual behavior is "contrary to Christian Teaching". In the phrase "practicing homosexuality is contrary.." it wanted to replace the word "homosexual" with "anal and oral sex". The title as given by Elana was "Change homosexual to anal and oral sex." The chair person of the Financial Administration legislative committee, while trying to hand out petitions to the various subcommittees, held that one up in the air and asked the 80 members of the committee (in front of 200 visitors): "Does anyone know why we have a petition that wants to change homosexual to anal and oral sex?" It was quiet for a moment and then a man in the back of the committee stood up and asked "Do they need a receipt?"

It became a rallying cry in our office for the remainder of the conference. Often when someone asked an unanswerable question, the reply would be "Do they need a receipt?"

The second problem was much more serious. Under the 1996 rules regarding retirement of judicial council members, an election had to be held to replace one member who had completed 4 years of an 8 year term, but who was too old to complete his second 4 years. The election was very political and took several ballots to elect someone. (Many people expect the judicial council to be quite important over the next four years as the division within the church over the role of homosexuals becomes more litigious.) The person finally elected holds considerably different views than the retiring council member.

At the same time a petition flowed through one of the legislative committees which would eliminate the early retirement requirement for the judicial council. The physical petition included the sentence "This action will be effective immediately." The ADCA did not contain that sentence. It is unclear whether the members of the legislative committee knew that sentence was in the petition, but it is clear that they didn't print a correction in the DCA. So when the conference voted to approve the petition, most of the people voting were unaware that they had just voted to invalidate the judicial council election they had just spend an hour or more doing.

You see, the judicial council members' terms end on the last day of the conference, so while the election was complete, the newly elected person had not officially taken office yet. When the petition with the stealth language was approved, then the vacancy ceased to exist and the election was void. This was argued long and hard on the floor, and was even submitted to the judicial council for ruling. The ruling came back as I've just described.

The end result was 2 1/2 wasted hours of all the delegates time on the floor of the conference (actual cost to the conference of around $150,000 - 200,000) -- All because the contractor who had been assigned to prepare the petitions for the conference couldn't be bothered to be careful, and her work wasn't reviewed by the publishing house staff or by the elected petitions secretary.

While all this was occupying all the available time of those around me, I had a few other kinks to work out with PETS. The first was that a change in the outline of the Book of Discipline caused my program to incorrectly assign two calendar items to the regular calendar when they should have been on the consent calendar. These two dealt with paragraphs numbers that had been part of the constitution of the church when I wrote the PETS software but have since changed. This one was fairly easy to find and didn't really cause me too much trouble.

The second problem was much more insidious. We discovered after the first batch run of calendar items that several had a negative one instead of a valid, lettered subparagraph identifier in their extended calendar numbers. (If this doesn't make sense, don't worry about it.) I found the errant code and thought I had it all fixed after the first batch run. But then the second batch run messed up exactly the same way.

For those of you who haven't been there, the batch run is how I spend my evenings at Conference. I start preparing for it around 6pm, usually run it around 8 or 9 pm and sometimes get it completely fixed somewhere close to midnight. This is the program module that takes all of the petitions that have been voted on in Legislative Committee and turned in to the Coordinator of the Calendar, and processes them. It splits them up into three consent calendars (concurrence on Disciplinary changes, concurrence on non-Disciplinary actions, and non-concurrence) as well as the normal calendar based on the number of people who voted against the recommendation of the committee. Items on the consent calendar are essentially rubber stamped without discussion on the conference floor, so it is very important to not let items into the consent calendars that don't belong there. On the other hand, items that are left pending at the end of the conference are "unfinished" and no action is taken, so an item that should have been placed on the consent calendar and for some reason isn't is likely to have the opposite action as if it had been correctly placed on the consent calendar. In either case, there is tremendous pressure to get it right.

On the technical side, the batch module is the only one that cannot easily be backed out of, restarted or run again. To add to the pressure, any item that hasn't been entered into the batch process queue before the batch process runs, will not be printed in the DCA for another day, delaying possible action in the plenary session by another 24 hours.

So each evening I spend all of my energies making sure that all of the potential items are in the queue, I action any items that failed any of the validity tests to be added to the queue, I double check all of the entries for removal from consent calendar by signature of delegates, and I hold my breath. Then I start the batch module, and watch without blinking while it runs through its various, irreversible steps. Finally I check as much of the output of the batch module as possible before turning the final output file over to the DCA staff.

That output file becomes 20 to 60 pages of text in the next morning's DCA. They had a hard deadline of midnight to get the camera ready pages to the printer so that the 2 or 3 thousand copies of the DCA can be delivered to the conference center by 6:30 the next morning and placed on all the delegates desks by the time they arrive for the 8:00 am opening worship service.

Any problem with the batch process potentially delays the entire DCA printing and delivery activities. If you've ever worked for a daily newspaper, you understand the time pressures of publication deadlines.

All of that information is background to set up my second evening of playing with the batch module. I was sure I had fixed the subparagraph code, but I still had subparagraphs that were negative one when they should have been an upper or lowercase letter! I manually fixed the DCA output file and started to manually update the PETS database when I discovered that they were changing back to negative ones in the database just after my manual update!

I started digging into the code and found, to my horror, one by one, that NONE of the PETS modules handled alphabetic subparagraphs correctly! I was up until 7:30 am fixing and recompiling all but two of the ten PETS modules.

That first Thursday was my longest working day: 22 1/2 hours without a break. Across the two weeks, my shortest working day was 16 hours (not counting the 2 hour lunch I had that day.) And in all those days, Sunday through Saturday, then Monday through Friday (actually Saturday morning at 6:30am) I only had five meals outside the convention center. All the others were working meals where I ate in a meeting or in our computer room and worked while I ate.

The reason the conference is staffed primarily by volunteers is that absolutely no one would do this kind of work in this kind of environment for money.

The batch module ran without a hitch until the second Tuesday of the conference. Then my worst fears were realized. It halted with a fatal error code that I'd never seen and didn't even recognize. An emergency was forming because I had started the batch module later than normal that night. The code showed the error was in a place that seemed impossible to me. The program was unable to open a normal ASCII output text file that it had already opened and closed a couple of dozen times during the batch process. The file was on the local hard disk (C: drive), not on the network. There were no other users, no weird security parameters, absolutely nothing that should have prevented my program from being able to open that file again.

To give you some idea of how bizarre that error is, the various PETS modules use exactly the same text output file opening routine many times as they write every text file, every report, every cover page, every proof page. It is used thousands or even tens of thousands of times per day during conference and is unchanged since the first PETS program I wrote under DOS 2.11 in 1987! How could it fail?

Well, I don't really know for sure, but I suspect a combination of really bad power and a really cheap rental computer. The computer I was using was one of those build your own PC by adding a motherboard, processor, disk and power supply in a no-name case kind of junky box. I'll admit that I'm spoiled by usually having very high quality HP computers around me. And I'll also admit that it makes sense for the General Conference to rent the very least expensive computers available. But what I had on my desk while I was running the batch module was probably the least robust computing environment I've ever personally worked with. And it was running in an environment where the mains power had been measured to drop as low as 87 Volts AC when the Xerox machine in the next room started up.

So, Murphy's Law struck, and the one and only time I'm ever likely to see that particular fatal error, it caused the batch module to stop when it was irreversibly part way done. I knew that I had intended to make the batch module self correcting if it failed part way through, but I had never written very much of the code it needed to do that. I hoped against hope that it had failed in a place that it knew how to recover from, but it wasn't to be.

I asked Susan to drop what she was doing and look over my shoulder while I tried to manually back all of the partial database entries out by editing the binary database files directly. Then I had to reset all of the text files which are changed or moved as part of the batch process. I reset all the partial completion indicators, and then tried it again about 90 minutes after it had failed the first time.

Right off the bat it produced critical error messages that indicated that it couldn't find and move some text files that it was expecting. So I let it run to completion, and sure enough, there were significant sections of text and important text files missing. I immediately figured out what I had done wrong, and started a second time to manually reset the database and text files in preparation for a third run of that evening's batch module.

The DCA office had started calling on a 30 minute basis to ask me for time estimates about when I thought I'd be able to complete the batch process. They also called the printer, and for some extra money were able to move the hard deadline back to 1am.

The second time through the manual reset operation, Susan and I were able to do it in about 60 minutes. We kicked everyone else off of PETS to make it easier for us to make the changes without loosing updates by other workers. I prayed a heartfelt prayer and launched the batch module for the third time at 11:15 pm -- over two hours past my hard deadline for submission to the DCA.

It seemed to run OK, so I was mystified, and more than a little panicked when the output file looked exactly the same as the first time I had run it! It only contained the first 12 items of about 250 and then the file simply ended!!! I did a directory listing to see how big the computer thought the file was, and it also showed a 17k file when I was expecting something around 200 - 400k. I felt like I was about to dissolve into tears when I noticed the time on the output file was 8:45 pm not 11:15 pm. I was looking at the output file from the first batch run, not the recent one. I hadn't deleted the output file because I assumed that a subsequent run of the batch module would overwrite it.

But it hadn't! Why not? Then I remembered, the module actually wrote the file to a different place and then copied it to where I was looking. The copy had failed because a file already existed with the same name. But the file it was going to copy from still existed! I looked there and found a file with the right time stamp and the right length. I opened it in an editor and felt relief flood through my body to see that it seemed complete and correctly formatted.

I did a cursory check and then sent it to the DCA office at 11:35. I was still at the DCA office when it went out the door in Rich Peck's hand at 1:05 am while he sprinted for a car to drive it to the printer. While that day had only been a 19 hour day, I was absolutely and completely exhausted. I was drained body, brain and soul. I've got no idea how I got to my hotel room that night. But I do know that without Susan, Gary, Gere, Mom, Randall and Mike Cunningham of the publishing house, I would not have survived that evening, and the calendar items would not have been printed and in the delegates hands the next morning.

Looking over my shoulder through all of the trials from the first Friday through the second Wednesday was a systems analyst from CGI in Texas. His company will be creating the RFP (Request for Proposal) against which the Commission on General Conference will solicit bids for the replacement system for PETS. I'm hoping to call this new system Legislation Tracking System (LTS), because it will be so much more than a Petition Entry and Tracking System. And in all honesty, I feel somewhat proprietary about the name PETS.

I don't know what the analyst's impression was of the Tuesday night mess, but I did see in his notes that "all system activities must be reversible (with appropriate capability)" and I look forward very much to using (and not having to write) a system that meets that goal.

On a more personal level, I learned many things.

First, don't put wall warts (plug in transformers) inside dress shoes in a suitcase that will be handled by airline ground crews. My only pair of black dress shoes showed up in Cleveland having been cleaved. Well, at least punctured in a fair impression of a power socket. By the time the conference started the material between the slits had burst and if not for my new black socks, it would have been an embarrassing sight, even from quite a distance. It took me until the second Tuesday (8 days after the blowout) before I had time enough to replace them. My new shoes are not quite as comfortable yet, and they were the most expensive shoes I have ever purchased, but I was able to get them within walking distance of the convention center during a short lunch break.

Second, when you represent an office that must remain impartial and detached from the political nature of the conference activities, it can be very difficult to sit still at the front table on the platform of the General Conference when people you admire and respect are standing to show support for a position that you hold dear. I was in that position on the final Thursday morning of the conference. There had been various demonstrations throughout the conference, and over 180 people had been arrested the previous day outside the convention center for sitting down in the main driveway of the center during the morning hours as delegates and visitors were trying to gain access to the building. Also there had been a suicide attempt in the plenary session (a potential jumper from the visitor's balcony) just a couple of minutes before. Emotions were raw throughout the convention center, and my emotions were completely engaged. See http://umns.umc.org/gc2000news/photos/0511photos.htm for some pictures of these events.

The demonstrators on Thursday pushed their way into the bar of the conference (past the marshals, who are security guards of the mildest sort) and held a peaceful stand-in that sought not to prohibit the conference from operating, but to put faces on the people who would be most directly affected by the actions of the conference that morning.

The presiding bishop called a 20 minute recess in order to try to either accommodate the demonstrators or clear them. The supporters of the demonstrators' cause (about 1/4 of the delegates and perhaps 1/6 of the visitor's gallery) continued to stand and sing. Those of us at the Conference Secretary's table on the platform continued to sit and I was painfully uncomfortable. One bishop, Janice Riggle Huie, came over to the Carolyn's table and began a conversation with Carolyn and me. This provided our opportunity to stand in order to talk directly with her. It gave us an excuse to stop sitting (which felt to me like an implied agreement with those who opposed the protesters) and stand in a way that wasn't obviously in support of the protesters. It allowed me to maintain an image of impartiality without feeling like I was in defacto disagreement with the demonstrators. It may sound to my friends who aren't familiar with the conference as inconsequential, but I will never forget Janice's perceptiveness and kindness.

I learned that my recorder sounds wonderful in the all-marble great entry and stair case of the Cleveland convention center at 1 or 2 or 3 am when no one else is about. There was only one night in the entire two weeks that I left the convention center before ALL of the delegates, visitors and volunteers had been helped out and all the exterior doors had been chained but one. All of those other nights, I managed to play a little bit on the recorder on my way out of the building. All of the notes in my range seemed to sustain equally. I would guess that grand entry has a reverb time of 6 seconds. I would have given quite a lot to have some of my musical friends with their instruments and voices there to record some music in that amazing sounding space.

Other random thoughts:

I appreciate very much that Hewlett-Packard paid my salary during my time at the conference, allowing me to save my vacation time to be spent with my family. I've been pretty up front with the conference about my employment at HP, and I hope my work for the conference reflects well on the company.

Illness seemed to strike my family throughout the conference. My mom wasn't feeling good early on in the conference. But she recovered and was up to helping in the Secretary's office much of the time she was there. It was good to have her available to help during the petitions crises.

Dad got ill the last evening of the conference. He walked into the computer room carrying a waste basket and explained from a distance that he had very recently used that basket in an undignified way and was headed to the hotel. He asked me to get Mom and all of their stuff back to the hotel when she was ready to go. He also gave me a map to where Mom was sitting in the visitor's balcony. Even when he's between hurls, Dad is a pretty good orienteer and cartographer. I was able to find Mom without any problems even though she was in the middle of a thousand people in a section of the convention center I'd never been in before. Fortunately, Dad recovered by the next day. Actually, it was really fortunate since their first flight was delayed, they missed their connection in Chicago and didn't get home until 3:30 am on Sunday!

And Joy had sick children on her hands in my absence. It was particularly hard on Joy because she couldn't even reach me reliably to let me know that she though she might have to take Elizabeth into the hospital. Our good friends Nancy and Dave Fries came to the rescue and supported Joy in my absence. While I didn't have a lot of time to just sit and think about family and home, it was still very unsettling to know that my baby was very ill and to not be able to do anything about it.

On my day off (the Sunday between the two weeks of conference) I was able to have breakfast with my parents and then join the Susan, Randal, Gary, Quinn and others for an afternoon at the Rock and Roll Hall of Fame. I learned that I have a passion for one-hit-wonders. Those bands that put a song I love at the top of the charts and then never did anything worthwhile again. It is unbelievable how often that has happened. And it is amazing how many wonderful songs are the only wonderful thing from some no-name group. It was even surprising to me how many of the songs I grew to love in my high school Youth Fellowship group (YF) were of that ilk. Do you remember "Put your hand in the hand"? Do you remember anything else from the group "Ocean"?

I spent a fair amount of time the last two days of the conference meeting with various people to figure out what we would do differently in 2004 to make sure we didn't have any repeats of the problems of 2000. I'm currently about 27 pages into a 100-150 page presentation that describes in detail what the processes, responsibilities, rule changes and reporting relationships ought to be for the Legislation Support section of the Secretary of the General Conference's office. Many good ideas such as a mandatory pre-conference audit of the petitions material, and creating petition packets or folios to ensure physical control of the integrity of the paperwork came from the group as we met formally and continued to talk informally about the conference processes.

During one of those conversations with Victoria who will replace Rich Peck as editor of the DCA in 2004, she asked if she could have the petitions sent back to Nashville in case questions came up while the Book of Discipline editing process continued after General Conference was over. We responded that the working set and reference set had to go to different places after the conference, but she should have the Disaster Recovery set in her office at Nashville which we estimated should be at least 80% accurate based on our work with the working and reference sets. So she asked Elana where the off-site (Disaster Recovery) petitions set was stored and found out that Elana had never made that third set! For the first time in the organizational memory of the conference we had no fallback set of petitions in case of fire or other disaster in the convention center. If the various electrical problems had created a fire in our work room (and the computer server room) at the start or end of the conference, we would have lost everything. Computer data, all original and hard copies of the petitions and calendar items; everything.

We tried to be very diligent about keeping backups of the electronic data "off-site" during the conference process. We made CD-ROMs and DAT tapes every day and stored them in Richard and Tom's hotel rooms. On that last overnight push after the conference ended at midnight on the final Friday, I copied all the files from the PETS server and the web server to my portable, and started a copy over the internet to the Angel Inn Server at my home. At 6:30 Saturday morning, I went to my hotel to get 2 hours of sleep before I had to get up to catch my flight home.

Richard and Tom got to the convention center on Saturday morning (they didn't pull the all nighter like Susan, Randall, Gary, Gere and I did). They copied everything from the primary server to the backup server and then made a complete image backup of each machine. They boxed them up, put them on a pallet and headed back to Nashville.

When the pallet with the servers arrived at the publishing house in Nashville by truck. All of the boxes were empty! Both servers and ALL OF THE BACKUP MEDIA HAD BEEN STOLEN!!! When I returned from my family vacation after the General Conference I called Richard to find out why I couldn't access the PETS server remotely. I then heard the story of the stolen equipment. Richard asked rather anxiously if I had backed up the files somewhere. As it turns out, the file copy to my server at home had failed over the internet, and the ONLY electronic copy of ALL of the work of the General Conference was on my laptop computer! I quickly copied it to my server, and set permissions to allow the publishing house to copy the files from my server to a server in their data center. I also got everything on a DAT tape here in California which will soon live in Dad's fireproof safe.

It is unbelievable that so many processes and procedures broke down throughout this General Conference. The problems we had to deal with make all of our problems of the previous conferences seem trivial. And the number of times we flirted with real and complete disaster still make me reel. As bad as so many things were, we were still lucky beyond any expectation that we barely missed the complete loss of the conference information.

This is weighing heavily on my mind as I work on the suggested policies and procedures as well as the RFP for the Legislative Tracking System for General Conference in 2004.

As I think about the parts of the conference about which I have not written, there are so many little things that I'd like to share. But so many of them are tied closely to the people and processes and history of the conference that there is no reasonable way to tell you about them. I'll have to settle for offering my apologies for my inability to write about the conference as well as I'd like. And I'll offer to answer any questions you have about the General Conference, PETS, my role there, or anything else you find of interest.

Peace,

John.