I first worked on a CDF project back in about 1981, years before the Tevatron turned on. The University of Illinois wanted to make the central muon detectors, and Lee thought we could use a simplified version of the Aachen tubes built for UA1. I fiddled around with some extrusions and learned how to make a cosmic ray telescope to trigger the readout and discovered that electrolytic capacitors could sometimes explode (a CAMAC-based prototype TDC we borrowed went fiss/bang and caught fire). At the end of it all I didn't get really good results, and the design was eventually chucked anyway because the drift time was too long. The beam bunch crossing time was going to be longer than the drift time, which would mean we'd have mixed in hits from muons from different collisions, which confuses the reconstruction.
Later on the Tevatron went to more frequent bunch crossings, and as we went up and up in luminosity we started getting multiple interactions per bunch crossing, not nearly as many as at the LHC, but still a lot. So we wound up with hits from different events anyway, but by then we had a much better handle on how to reconstruct tracks and stubs.
I took a break from the drift chamber work to work on my thesis experiment, and then went to Wisconsin as a post-doc. There I worked on UA1 and CDF again: not monogamous, as Carlo Rubia complained about my boss and his team. Since then I've been either full or part time on CDF--1985 to 2011. Funding for my part of the project ends this year along with the beam.
The beam is off, and it is the end of an era--but the data processing goes on, and student analyses and theses will keep going for several more years after that. And some profs will decide to have a look at something they never got a chance to finish; something that the LHC can't do as well. My group was charged with providing software releases to last another 5 years; that are supposed to be robust enough to be ported to the next generation OS with minimal labor. I think we're getting it.
I still have some directories with software for the online system--something to monitor the HV for the Wisconsin muon detectors which I was delegated the job of maintaining after the author left, and so on. Time to clean up.
I worked a little on a few analyses, but kept getting pulled off to fix something, so I didn't actually contribute a lot to the physics. Some, yes; but not a lot. Kind of a gloomy thought to end on, but that's life--I spent time keeping things running and helping people get their jobs going and keeping the data running. And writing code to help make it work, online and offline. It's my fault if I didn't push harder to do the interesting stuff. (OK, some of it is interesting. The W-mass work is useful but not wildly exciting.)
I took shifts back when we had a walk-through that checked the voltages on the HV supply panels and the gas percentages in the mixer in the gas rack (had to climb down into it) and that got you back to the control room in time to do it again. And I took shifts when almost everything was automated, with software to diagnose what was wrong and blink warnings at you (I wrote some of it), and the biggest headache was going through the run validation checklist. VAXes and terminals tuned into Linux and multiple displays. Back in the day dumb terminals were so expensive that Fermilab bought Macintosh computers and ran terminal emulators (the old Macs).
Meetings in the pump room were awful--the pumps downstairs made it very hard to hear anybody. A new room was built later--and still called the pump room. The old one turned into offices which had enough cubicle walls to muffle the noise.
Fermilab got larger, and more and more bureaucratic and rules-bound. Some of the stuff turned out, on inquiry, to be implementing Federally mandated rules, which seemed to appear in new crops every year. Three years ago it seemed as though somebody had gone training-happy and was designing mandatory training courses for everything up to and including garbage disposal (I'm not kidding). It has been quieter since--perhaps the users rose in wrath.
The trailers where almost everybody worked have gotten shabbier and emptier, and nobody is replacing the computers anymore--which is kind of a hassle when you are required by Fermilab policy to upgrade to a new OS that maxes out the memory in the old machines.
I didn't go to the party--there were some family things to do, and I've not been a huge fan of big parties anyway.
UPDATE:
Store 9158 was the last, and run 312510. Final total: CDF acquired 9.977 fb-1. Also see the Fermilab Tevatron timeline.
It is funny that one of the milestones is listed as "First 8mm tape used to record data from a high energy physics experiment." I was somewhat involved in that, agitating for their use. I started a discussion group on the main VAX cluster about the subject, and after a few months the computing division took ownership of it away from me. I don't know if it was because it was getting large and they wanted to manage large discussions, or because there were some rather frank comments about the computing division on it. A few months later they started introducing the tapes