On Monday I attended the Linux on Wall Street conference. The overall theme of the conference was leveraging open source (and Linux) to power mission-critical business applications, particularly financial services applications. Many of the big boys from all spheres were there (customers, vendors, press, and analysts). I had the opportunity to present a keynote session entitled “Open Source in High-Performance Trading Systems”.
My talk began with a discussion of how popular Linux and open source have become over a relatively short period of time, where the market is today and is projected to go, and what lessons there were for all of us in this era of mass collaboration and the “architecture of participation”. Who of us a few years ago would have imagined that the terms Linux and Wall Street would be in the same sentence, let alone a conference title?
I also drew the comparison between the early rise of the internet and the current trends going on in open source development. Does anyone remember folks who said, “What would our company ever use the internet for?” I recalled the “Sustained Inattentional Blindness for Dynamic Events” study, or what I called the “Gorilla Experiment”. For those of you who don’t know about this, it’s an experiment that was put together by researchers from the University of Illinois and Harvard.
Volunteers were asked to watch a video and count the number of times that two teams of people passed a basketball back and forth to one another. So, in the video, you had these two teams throwing the ball back and forth, and the volunteers were asked to count the number of passes that were made. The video runs for about 45 seconds and at about 30 seconds into the video, a person dressed in a gorilla outfit comes into the video, right into the middle of the action, stays there for about 9 seconds, and then leaves.
Here’s the amazing part. 50% of the volunteers never saw the gorilla. They were so intent on counting passes that they missed the gorilla walking in. To me, it’s this whole idea of conscious unconsciousness. Applied to the open source development model, on some level, people are aware of it, but on another level, they are choosing to do nothing about it. Details of the “gorilla experiment” can be found here.
We then dove into the details of the financial services market, the challenges that companies face with their legacy infrastructures, and the need they all have to offer more “portalized” experiences for their clients. As customers, we all want to access our account information from wherever we are: brick, web, mobile, etc. And, we don’t care what applications, databases, or operating systems are running the systems. These portals need to serve as windows into services on the network and back-office functions. Of course, with their legacy heritage and various M&A’s, most financial institutions are struggling with modernizing their environment and achieving this sort of services oriented architecture (SOA).
But, these problems are solvable with the proper phased approach. And, this is one of the areas that the Open Solutions Alliance (OSA) is working to address. In their common customer view prototype (which will be demoed at the upcoming LinuxWorld in San Fran), a point-of-sale legacy application will be integrated with an open source ERP system, an open source CRM system, and a bunch of other open source components. The customer information will be captured once (at POS) and automatically propagated into every system and every database. The reference architecture will be published for anyone to use, and the entire project is completely transparent … anyone can come to the OSA site and contribute.
Back to Linux on Wall Street, I also flew in a client of ours from the UK to speak about their success. Redmayne-Bentley is one of the UK’s largest independent stockbrokers. And, Michael Wheeler, their CIO/CFO equivalent, is one of the nicest people you’ll ever meet.
Redmayne-Bentley was running a legacy Cobol application on SCO Unix and essentially was running out of horsepower, and of course, had very little opportunity to innovate. And, this was their main trading system! So, they were looking to modernize the application environment. They have had a very good relationship with Unisys (kudos to our sales lead there Richard West), and we guided Redmayne-Bentley to an open source modernization solution leveraging three ES7000 systems.
To make a long story short, the migration to the new system was completed with zero down-time, they now see a 10x improvement in overall performance, their overnight processing is now down from 13 hours to just 1.5 hours; and they’ve had 100% reliability and availability since the migration (zero seconds of downtime, other than the normal software maintenance updates).
Yes, Linux and open source have made it to Wall Street, and yes, they belong there. Mass collaboration is changing the world, and Linux and the open source development model have helped usher in this phenomenon. Here’s to the road ahead.
Finally, to my faithful readers, I would be remiss if I didn’t offer a comment about the Roosevelt Hotel, the venue hosting the event. Those of you who follow my blog already know my perspective on the Roosevelt. Beautiful hotel, lousy rooms.
So, not only was my room as cramped and suffocating as the last time I stayed there, but I must have also gathered some negative karma that resulted in my leaving my cell phone charger in the room. When I phoned the next day to check on whether anyone had “found” it, they told me that the current occupant indicated that nothing was plugged into the wall. I don’t suppose they ever considered asking the housekeeping staff.
