Re: [unrev-II] A Cusp: License issues for released code & Legal Notice

From: Paul Fernhout (
Date: Tue May 16 2000 - 06:30:23 PDT

  • Next message: Eugene Eric Kim: "Re: [unrev-II] A Cusp: License issues for released code & Legal Notice"

    Eugene Eric Kim wrote:
    > On Mon, 8 May 2000, Paul Fernhout wrote:
    > > "Indemnification" as I understand it means, you pay the other person's
    > > bill if anything goes wrong.
    > Not quite. It means that if anything goes wrong because your code
    > violates someone else's IP, neither BI nor Stanford are liable (i.e. they
    > can't be sued.) In that situation, BI/Stanford would still have to stop
    > using that code. Indemnification cannot be used as a loophole to avoid
    > paying IP licensing fees.

    Based on previous discussions with my IP lawyer, indemnification as
    broadly construed is in effect a warranty to ensure the customer for the
    Intellectual Property can use the IP and will not have any additional
    legal costs associated with that use. What you say is correct, BUT you
    are ignoring who *is* then liable in this case (the contributor!) as
    well as the implications of that liability given however Stanford/BI
    have chosen to use the contribution (i.e. Stanford/BI's loss in not
    being able to use the code, and the cost of redressing that loss).

    Example: Some company contracts with me to develop a software package
    for displaying compressed images. I unknowingly infringing someone's
    patent on displaying bitmaps. The patent might even have been issued ten
    years after I supplied the code, despite my own prior art -- it doesn't
    matter. Someone sues the company. Unfortunately for me, I have agreed in
    writing to a broad indemnficiation claue similar to the one in the
    Colloquium's "permission to use". The implications of indemnification
    are as I understand them in this case are:
    A) I have to pay all the company's related legal bills and any
    B) I have to fix the problem.

    Since the indemnification clause is so general, it does not specify:
    A) That I as the contributor have any control over the legal defense, or
    B) What acceptable remedies are for fixing the problem.

    Thus, the worst case implications are:
    A) The company hires expensive incompetent legal defense which I have to
    pay outrageous fees for,
    B) because they are incompetent the case is lost in the worse possible
    way (despite prior art), with the worst possible settlement,
    C) I have to pay the worst possible settlement,
    D) I still have to fix the problem. Since there is no specification I
    can replace the code, the company may not want to accept a replacement,
    and this may means negotiating a licensing fee with the owner of the
    patent, or alternatively, if the patent owner will not license, paying
    the company whatever it wants so they stop using the code. The company
    may have done anything with the code because there are no distribution
    restrictions in the agreement. They may have sublicensed it to many
    other developers, embedded it in millions of set-top box computers,
    based their business model for going IPO on exclusive rights to that
    code, etc. Thus, the costs of "fixing" the problem could be exorbitant.

    The only hope in that case is that an unsigned indemnification agreement
    of this scope would not hold up in court (which I believe may well be
    the case).

    In practice, one might expect some good will and a better than worse
    case outcome, but I think it could get *that* bad. That is why last
    time I "indemnified" a client for something (only after being enticed by
    a large sum of money for the project), there were specific limitations
    insisted upon by my IP lawyer, in effect:
    A) Indemnification only applies if I control the legal defense,
    B) I could replace the code with other code which accomplishes a similar
    end goal, and
    C) The use and distribution of the code would be limited to the
    companies internal operations for purposes of the indemnification (i.e.
    I could estimate in advance the scope of the risk).

    Perhaps someone on this list with their own convenient access to their
    own good IP attorney might want to check with legal counsel to see if my
    statements on indemnification are correct in a worst case situation.
    This is important because you are putting your company's assets on the
    line if you participate as an employee (not an individual). By the way,
    has anyone else had their company's legal counsel review the agreement
    who can comment on it?

    When I have seen unsigned or clickthrough indemnification agreements,
    they are usually the other way around, written to protect the code's
    author, as in: "You may use this code or service which the author
    supplies, but you indemnify the author if your use is legally actionable
    for whatever reason." For example, the mailing list service (egroups) we
    currently use has such a "other way around" indemnification provision in
    its terms of service:
    (By the way, from that agreement they gain some rights to the list
    content as well, so I guess one could go to them as well as BI/Stanford
    for some rights to this content, as long as it is distributed only to
    "members of the list" however that gets defined.)

    (I am not a lawyer, so consult with your own lawyer for legal advice
    before making decisions based on anything written here...)

    The deeper emotional issue is what a one-side "permission to use"
    agreement for the Colloquium "feels like" to any open source developers
    lurking out there.

    -Paul Fernhout
    Kurtz-Fernhout Software
    Developers of custom software and educational simulations
    Creators of the Garden with Insight(TM) garden simulator

    Win an iMac computer from The Apples in stereo and other
    FREE great stuff at!

    Community email addresses:
      Post message:
      List owner:

    Shortcut URL to this page:

    This archive was generated by hypermail 2b29 : Tue May 16 2000 - 06:35:12 PDT