No announcement yet.

open-source and HL7

  • Filter
  • Time
  • Show
Clear All
new posts

  • open-source and HL7

    I'm confused about an apparent incompatibility between HL7 and open-source licensing. Maybe someone here can help.

    HL7 Inc. is protective of its copyright on standards such as v2 and v3 messaging, the RIM, etc.

    Confusingly, the HL7 bylaws prohibit individual members from using HL7 protocol specifications to develop software products for healthcare information exchange:
    (from HL7 bylaws section 03.02.02)
    The use of the protocol specifications by an individual member not associated with an organizational member in the development of software products specific to the exchange of healthcare information or the interoperability of such information is strictly prohibited and may be grounds for revocation of membership.

    However, the HL7 bylaws also say the information content of HL7 protocol specifications is in the public domain and promoted for use by any interested party:
    (from HL7 bylaws section 02.03.02)
    The information content of HL7 protocol specifications shall be in the public domain and be actively promoted for use by any interested party. The protocol specification documents and other productions thereof shall be the intellectual property of HL7. HL7 expressly reserves sole right to publish and sell the documentation of its protocol specifications, and shall exercise all applicable copyrights to said materials.

    The bylaw quoted above is the only HL7 written policy I've found which grants permission for anyone other than HL7 Inc. to develop software using HL7 protocol specifications. Beyond this, I haven't yet found a clear written description of HL7 policy toward software implementations of HL7 standards.

    Because the HL7 bylaws exhibit an apparent lack of internal consistency, I'm not quite certain how to interpret them.

    HL7 also distributes software and other content on its GForge site (, but the license associated with that content is not open-source.

    By my understanding, Mirth Connect has its own implementation of HL7 protocols.


    1. Did Mirth developers obtain explicit permission from HL7 to develop software based on HL7 protocol specifications? Is permission, or HL7 membership, required for development and distribution of this type of software? What about redistribution of essential HL7 artifacts such as XML schemas? (How is HL7 compatible with open-source?)

    2. Does HL7 have a clear, written policy that describes how software developers, who may or may not be HL7 members, may use HL7 standards in software?

    Information to help answer these questions would be greatly appreciated.

  • #2
    no response on this?

    It seems like a valid set of questions; I wonder why nobody from Mirth Corporation has replied?


    • #3

      it is not possible to prevent or prohibit reading and writing of data bytes which can be observed with the simplest of tools, and producing a parser which understands the information so observed.

      this is the basic tenet of reverse-engineering.

      around 1998, microsoft claimed that the SMB specification was "copyright" (which it isn't copyright MS, at all, it's actually copyright The Open Group). the extensions added by microsoft included some error messages. microsoft claimed copyright of those error messages.

      so, andrew tridgell wrote some code which generated SMB error codes, starting at 0 and increasing by one in each network packet. he then captured the resultant traffic using Microsoft NETMON.EXE and dumped the results to a flat text file.

      then, he wrote some code which parsed the output of NETMON.EXE, looking for the verbose description. the verbose description included the all-important string associated with the error code number.

      by outputting that string along with the error code, he was able to bypass the stupid stupid attempt by microsoft to claim "copyright" over the error codes, because the information was obtained in a manner that's compatible with "free speech".

      it just so happens that the resultant smberror.h is _entirely_ compatible with microsoft's nterror.h codes.

      then, there was the stupidity of a company trying to claim for years that RC4 was their copyright. all that happened was that people renamed their algorithms to "arcfour" and called it "RC4-compatible".

      you _cannot_ stop people from producing interoperable systems, and you _cannot_ patent software in Europe (and will be able to do so when hell freezes over: i've spoken to someone from the EPO and they and all of their colleagues will let software be patentable in Europe over their dead bodies).

      all that happens, if you try to enforce "copyright" over a language is that you look like a dick. and in the case of free software vs a corporation _especially_ where health is involved, the party who is attempting to "control and restrict" has the hippocratic oath to take into consideration.

      think about this: by attempting to "restrict and control" the flow of information, that company would be preventing and prohibiting doctors from being able to help their patients, because they would be prevented and prohibited from using anything but the most expensive restricted and controlled and licensed software, and many doctors simply cannot afford that.

      so the restriction and control of information formatting, through the claiming of "copyright", is in direct conflict with the doctor's ability to fulfil their hippocratic oath.

      think about that for a minute, and you begin to understand why there is silence on this issue.



      • #4
        The background of bylaws section 03.02.02 is the following

        You have to be a member to HL7 if you want to get the published specifications. A couple of different membership levels exist depending of the purpose the member has.

        A company has to become an organisational member to use the specifications for generating software. As an individual member, you may use the standard documentation for personal usage, which of course is not redistribution of software.

        This model is quite logical, otherwise companies would choose a personal membership for one of their employees to get the standard as a cheap solution.

        It's quite normal that you have to be a member of a standards organisation to have access to the standards.Remember that the development of the standard is based on the money gained by the membership fees and a lot of voluntary work.

        Last but not least: The bylaws do not say that it will be a copyright fraud producing HL7 compatible SW. It says that you might loose the right to be a personal member of HL7 if you use the standard documentation for things other than granted.

        One last word on HL7's gforge:
        The software beeing there is not for creating HL7 compliant interfaces. The GFORGE contains tools beeing necessary to develop and maintain the HL7 standard.
        Last edited by piet; 08-11-2010, 06:56 AM.