drbowen wrote on Sunday, April 03, 2011:
This is kind of a long post.
Jack Cahn did bring up some interesting topics that effect all of us.
Tony McCormick and I had a great time at the POSSCON in Columbia, South Carolina from March 23 through March 26, 2011 evangelizing about OpenEMR (on our on nickel by the way). There were about 500 attendees and this was the second year for their Healthcare track. There a surprising show of open source healthcare leadership at the conference.
Highlights:
Wednesday evening there was a “Birds of a feather” meeting in the IT-ology Executive conference room. This meeting was literally standing room only. David Sparkman the Chief Program Officer of the National Cancer Institute posed a question to the group and sparked a lively round table discussion among the attendees. The CaBIG program used by the NCI has developed into a mature program that the NCI is now seeking to open source and promote a community around CaBIG. CaBIG is the suite f tools used by some 70 different projects at the National Cancer Institute. The round table discussion ended in a debate between the “charismatic leader” vs. the Apache committee leadership approach. Jay Batson of Acquia, Inc. (Drupal) championed the charismatic leader point of view while Jim Jagielski the president of the Apache Software Foundation expressed the “Apache way.” This conversation was moderated Iain Sanderson. Opinions were offered by Dan Housman of Recombinant, Dan Demaggio of Animoto, Sam Bowen of OEMR, Dan Russler of Oracle and Mark Gunnels of HSSC. David Riley advised forming a separate foundation. Ned Thurman was concerned that the existing software was not really open source since it had been written with public monies that the licensing was really “Public Domain”. David Speakman appeared a bit overwhelmed by the magnitude of the response. The problem will be taking a large bureaucratic organization and develop cells of open source culture within the context of the whole.
Wednesday night the healthcare group adjourned to the “Flying Saucer” where William “whurley” Hurley of “Chaotic Moon Studios” paid for beer in exchange for ideas to start Android Apps. The goal was to start a new open source project with multiple coding plans before morning. There were 42 people in the this and “whurley” got about 50 ideas for new apps. Whurley gave an open source “unpresentation.” He asked the audience what questions they had about “open source” and developed his presentation “on the fly.” Whurley has a lot of experience with developing open source communities and offered a lot of insight into his prior work experience art starting open source communities.
Thursday, Sam Bowen and Tony McCormick gave an “open source presentation” with a combined presentation. The presentation seems to have gone over very well. We covered the history of the OpenEMR project, its basic software language and backend, and the growth of the project. Dr. Bowen covered the Internationalization and the major contributors to the project. Tony McCormick covered the recent Meaningful Use project and the recent ONC Modular certification. There were a number of good questions concerning the project that caused us to run over our allotted time somewhat.
Sam Bowen, was invited to join an expert panel to field questions about the future of Open Source. He joined Dan Russler the chief of Health Informatics at Oracle, Jim Jagielski the President of the Apache Foundation, and David Speakman the Chief Program Officer of the National Cancer Institute.
There was a lot of emphasis on the “Open Source Process,” how this is becoming more and more main stream. Now having contribution and participating in an open source process is considered a marked positive on resumes. The traditional training model for software developers has not emphasized the collaborative tools that now make open source such an effective process. Working in an open source process teaches collaboration, communication, working in social groups. Deb Bryant described large numbers of government projects that are adopting open source development methods to lower costs and increase quality of the software.
I personally used this opportunity to pick the some of the brightest brains in Open Source today. I specifically spent a substantial amount of time speaking with Jim Jagielski, President and Co-Founder of the Apache Software Foundation. There are several things that came out of this conversation.
1) It’s OK to make money off of Open Source
Sometimes this means charging for software development. Sometimes it means making money off of sales and support. We all like to refer to this is “The Red Hat Model.” Obviously, with a market capitalization of around $65 billion, the Red Hat Corporation has done very well off their investment in time and labor.
When I first started with the OpenEMR project in 2003 I did so as a volunteer to help promote the use of OpenEMR. I have had a lot of developers and companies ignore me, tell me to shut up, or offer to let me pay them (uh usually $10-15K) to enable my ideas. Well , I am still here and most of my ideas have been incorporated into the project. My goal then, as it is now, is to get as many users as possible to start using OpenEMR. My thought process is that users mean money to software developers. The more users we have the more developers we would have, then more companies, and finally even more users. This all means increasing investment by multiple entities into the OpenEMR project.
2) Some seek to get a temporary advantage by not releasing code back to the project.
The biggest danger for these folks is usually the advantage is only temporary and frequently the “advantage” is never actually realized. An open source project tends to grow rapidly and fill in these holes so that holder of the “advantage” finds themselves on a dead fork. These guys usually don’t make money as quickly as they expected and fall out of the project after a while. In my mind, a successful open source project kinda grows like a huge kudzu vine. It covers everything.
3) I don’t think you can make “weasels play nice.” (personal opinion)
Either they “get the concept” of Open Source or they don’t. Those who do “get it” contribute their code back to the project. This is why these holes keep getting filled in. No reason, Jack, Art, Mukoya, and others can’t put some social pressure on those who are not “playing nice.”
4) 90% of the work is being done by professional developers who are being paid by a company to do the work in all of the successful large projects.
5) GPL licensing.
The GPL v2 and v3 are “Strong Copy Left” licenses. They are very restrictive with what you can, and cannot do, with the code. This is the strongest threat to those who would like to hide their developed code “for market advantage.” It is easy to screw this up and the Free Software Foundation likes….LOVES…. to go after offenders. Curiously, these licenses don’t require attribution.
It is important to understand the GPL licensing. You other arm-chair legal fanatics check my language here (wink-wink Stephen-Smith).
If you write OpenEMR classes, modules, derivative works, or libraries using OpenEMR GPL’d code, you must release your software (classes, modules, derivative works, or libraries) under the GPL license. The GPL license in the software licensing world is referred to as “viral” because of this behavior. If you, for instance, write a script that makes database connections using the OpenEMR database connector class, then you must release the script as GPL. If you want to write software that you want to keep private (for whatever reason), don’t start with an OpenEMR script, class or module.
The project has encouraged, but does not require, donating the code to the non-profit. This is done to ensure that we have an attribution to the original copyright holder and to have uniform copyright in the source code. One of the biggest problems an open source projects face is the uniformity of licensing. Much of the code has just been created without any attribution. If you donate code, please, at the very least, put your name on it.
6) Changing to a less restrictive license would be very difficult.
These easiest way to change to a less restrictive license is to write the whole thing over from scratch. This sounds a bit “loony” but there are reasons to consider this. Very large successful projects such as Mozilla and Apache use permissive licenses that allow much easier development of “privatized code”. The main requirement is that there has to an attribution and the attribution cannot be removed. The main advantage of using less restrictive licensing is the projects tend to attract development investment more quickly. The Apache Software Foundation now has Seventy different projects. The ASF requires the developers to donate the code to the project and it has to donated under the ASF license.
7) On professional suicide.
What you should not do, is use someone else’s code and put your name on it. You will not be defended by OEMR, OSMS, Free Software Foundation or anyone else, if you do this. Software written by someone without attribution should be left that way. Don’t read other proprietary code and then start your own similar module. The chances of cross contaminating your thought processes are too great.
I would love to have a sit down, face-to-face with Stephen-Smith. He is obviously very knowledgeable and passionate about the licensing issues. My main goal is to make sure that developers can contribute and feel comfortable with what they are doing. The traditional openness of our project is a bit at odds with what is best for the project from a legal stand point. What is best legally is to only accept donation, under one predetermined license, similar to what Apache Software Foundation does.
8) Charging for copies.
The board of directors actually approved this quite a while back, but the way things are set up under Modular Meaningful Use, this won’t work. Individual practices have to attest that they are using “Meaningful Use Features” with CMS and there is no way the non-profit can affect this in any way.
9) Public-Private partnerships.
I don’t have any problem with setting up public-private partnerships with any single entity providing services to the OpenEMR project. Out of fairness to all the contributors to the OpenEMR project, OEMR would have to offer similar terms to any company who wants to set up a similar deal. Referrals should be made in an equitable fashion.
Several entities fall into this type of arrangement: E-prescription services, laboratory services, billing services all come to mind. The end user will be open to select the vendor of their choice.
All of these services involve sending or receiving protected health data from an end-user clinic or practitioner with third party entities who are invariably using proprietary software. (notice I am not an attorney, but tend to rush in to offer my lay opinion any way). To me these services make useful legal DMZ zones that keep the OpenEMR GPL’d software from directly interacting with the proprietary services of say SureScripts, Quest, LabCorp, Spectrum, BC/BS, Aetna, Cigna and others. These almost invariably involve signing non-disclosure agreements and other documents that are the antithesis to any GPL’d project (legally and socially).
The typically way to accomplish this is to interact with another service over TCP-IP protocol (not system calls).
OpenEMR 
Open Source permissive licensed engine (like a Java Servlet, Tomcat server, MIRTH) Proprietary Vendor Service (Uhh….like SureScripts).
So Shameem, I ave no problem with your proposed agreement as long as you understand that if others want to participate will will need to do this in an equitable fashion. If yours is the only billing company then you get first crack. The only requirement that I can think of from the OMER perspective is that we need to be fair when more than one vendor is involved, transparent, and vendors are contributing all of their code back to the project will get higher priority than ones who don’t.
10) Grants
Grants are hard. OEMR only recently received the 501©(3) designation. Hopefully this will make things “easier”, but they still take a lot of time and effort, especially federal grants. There is only one of me and I do all this mostly as an unpaid volunteer. If you are interested in helping with the grant writing process or can do some background work on what agencies may be willing to help fund this project please let me know. I really need “doers” not just “talkers.”
11) Hey! You can still donate!
Can’t argue with this one.
Sam Bowen, MD
http://oemr.org
The OpenEMR logo and any trademark actually belongs to me, personally. I had this created back in 2004 and have been letting the non-profit use it under the Free Document License with no enurement to me.