Sometimes I walk around the dish in moonlight. The 300-meter dish is impressive, even overwhelming, during the day. At night, under the soft light of the Moon, and with wisps of fog in the bowl, it takes on another dimension. The huge dome and massive feed support structure, suspended hundreds of feet above, seem to float weightless in space. Its an inspiring scene and helps to puts things in perspective. Tonight, Im thinking about how I came to be here, at this telescope in the moonlight.

But that was not all; "suitcase SETI" used the latest integrated circuits to perform Fourier Transforms, the mathematical technique that splits the radio spectrum into channels, and was controlled by one of the first desktop Unix computers, which also analyzed the data and presented results. With these and other advances, Horowitzs system marked the dawn of SETIs "modern era."

While Horowitz was building "suitcase SETI" on a fellowship at Stanford University, I was just a few miles away at NASAs Ames Research Center. As a research associate, I studied pulsars with Ivan Linscott who was also working with colleagues at Stanfords STAR Lab on a proof-of-concept system for the NASA SETI Program. These connections led to my writing a few lines of code for Suitcase SETI. Then, when our observing runs at Arecibo overlapped, Ivan and I collected data on "giant pulses" from the Crab Nebula pulsar and helped with Suitcase SETI observations. It was a very busy time at Arecibo, but very rewarding. It was during this visit I discovered the effect of moonlight and fog on the dish.

Meanwhile, back at Stanford, the prototype of NASAs dedicated SETI signal processing system, the Multi-Channel Spectrum Analyzer 1 (MCSA 1) was taking shape. Much bigger than a suitcase, it occupied most of a six-foot tall electronics rack and used the latest chips to generate channels with even greater filtering capability than the smaller system. The greater performance of MCSA 1 would allow NASA scientists to search with extreme sensitivity for ETs signals against the background noise of strong radio signals from Earths own technology.

I was interested in how the signals would be detected, especially ones that change frequency with time. Astronomers searching for pulsars have to deal with similar problems so I proposed to apply pulsar search techniques to SETI. I received a fellowship to work with the NASA SETI Program. For several years we worked on developing and testing signal detection algorithms in the lab and eventually in the field at NASAs Goldstone Station in the Mojave Desert. Moonlight on the desert is pretty but its not Arecibo.

During this time, the SETI Institute was formed (in 1984) and by 1989, Institute scientists brought the MCSA 1, along with a control computer and signal detection software, to Arecibo. That search was the most sensitive SETI search ever conducted, yet it was most useful as a test of designs and techniques being developed for the even larger Targeted Search System (TSS).

Drawing on lessons from the 1980s, the TSS used cutting edge technology. With incredible filters and exquisite signal processing performance the high-speed custom computer systems of the TSS analyzed data in real time allowing results to be evaluated immediately by sophisticated software. Easily the most capable search system ever assembled, in the first few minutes of operation the TSS did more searching than all previous SETI programs combined. What a thrill it was for my colleague, Dr. John Dreher and me when we operated the TSS during those first minutes a decade ago.

Now I am at Arecibo again. The moon is nearly full. The SETI Institutes Project Phoenix team is gathering here and at the Jodrell Bank Observatory in England. For us, this is a time to look forward and back. This will be the last time we use the TSS. It has served us well for ten years. We await the arrival of its successor, the New Search System. The NSS is designed to evolve as new technology becomes available. It is designed with the future in mind. However, until it moves to the Allen Telescope Array in late 2004, the NSS will continue the search here at Arecibo, listening for ET in the moonlight.