My first connectathon was in 2004. I've now been to 10 of the last 11 North American connectathons. That first connectathon was held in a hotel in San Diego. I remember the hotel well, because that's about all I ever saw that week. I forced myself to go outside for an hour one day just to get sunlight. Our network ran at a blazing 10Mbps. We had one tiny little pipe to the Internet, through which, if someone turned on their VPN, shut down the entire network. Network was spelled with two o's that week.
I shipped three tower configured boxes and a monitor to that event. My ADT feed failed, and our support phone lines were down back at home, and VPN didn't work, so I had to rewrite my HL7 ADT feed from scratch in Java. I did it in a day. That was before I knew about HAPI, or I could have done it in much shorter time frame. I was testing Request Information for Display (RID), and needed the ADT feed so I knew which patient to generate my CDA (Release 1.0) document for.
My favorite Connectathon related experiece was actually during the HIMSS Interoperability Showcase. It was when, due to the overflow around the NIST booth due to the premature invention of XDS, that we had to move the table. You can read more about that story here.
Besides the accidental invention of XDS, the biggest contribution of Connectathon for me has been the "Social Interoperability" that happens on the Connectathon floor. This is where people exchange ideas about how to fix problems we haven't yet solved, and talk about where the organization, and our respective national programs need to go. The brain trust that is here this week always has been incredible.
The biggest change I've seen are the signs of profile and vendor maturity with respect to interoperability and interoperability testing. This year I spoke with many teams who, by the end of day 1, had completed not just their basic network testing, CT tests, and document sample uploads and viewing tests, but many had also completed several other transactions, including XDS.b and BPPC, two of the harder profiles to test. And nobody has yet run me down for ATNA assistance. I guess the FAQ has done its work.
I look forward to another decade of connectathons, but this time in the sunlight.