I never met him, but we had quite the correspondence going when he was developing OPUS-RS. I had stumbled across some information about the OPUS-RS beta test site, so I started submitting some of my files. He found out about it an emailed asking how I had heard about it and what was I doing. I explained I was just experimenting and he asked if I wanted to experiment more. Over the next year or more he sent me beta OPUS-RS (RSGPS) executable files and instructions and I loaded the various versions on my PC and proceeded to experiment. I found multiple ways to make it not work and found out when and what kind of data was not acceptable. I am rather reliably sure that some of the final user aspects of OPUS-RS were at least spurred by my tests and findings. Early beta OPUS-RS would accept submissions of 5 minutes to 4 hours. I showed that throughout a 24 hour day there were sufficient 15 minute intervals were satellite PDOP was at a higher value than allowed for a good solution. Even with our full constellation today one can still find a few higher PDOP periods during the day. At the other end I surmised that the single atmospheric model correction value they were using was not as adequate as recalculating a correction value for shorter sub periods. I belief that had some affect on their current 15 minute to 2 hour file limitations. Because RSGPS allowed me as the user to adjust certain assumed input limits and constants, I managed to blow up some final solutions. I believed that ability amused Dr. Schwarz. I also had the ability to look at the solution process as it progressed and was amazed that the RSGPS program could sometimes resolve 100% of all ambiguities with as little as 4 seconds of data. For the majority of my test solutions I was relying on NYDOT network CORS stations and had a large enough network to test solutions with primarily N-S or E-W networks and found no significant differences.
We also discussed L1 only OPUS solutions, which I thought RSGPS was entirely capable of doing. After Charlie fully retired their was no individual with sufficient impetus to move that project further along. Also other staff persons where not inclined to humor me with later copies of RSGPS. After Charlies retirement I did manage to prove to myself L1 only data was adequate by taking L1 receiver RINEX files and calculating L2 appropriate values and then submitting to OPUS-Rs for solutions, but I never got around to automating that process.
I will miss his stimulating interest.
Paul in PA