Public Access to Perforce?

Matthew Dillon dillon at
Wed Aug 18 19:58:35 PDT 2004

    It should be noted that the copyrights on some of the networking pieces
    in DragonFly are really a non-issue.  For some reason I just couldn't
    convince Jeff to go with the standard 3-clause (which is what DragonFly
    has adopted as its 'official' copyright).  The 4-clause is annoying,
    but no less open source then the 3-clause (and that's Kirk's opinion,
    not mine!).  There is a ton of 4-clause stuff in all the BSD's including
    FreeBSD, inherited from copyright statements attributed to authors other
    then UCBerkeley (for which the UC Letter does not apply and thus the
    clause cannot be removed without tracking down said authors and getting
    their permission).

    The compromise I reached with Jeff was to separate his copyright from
    the 3-clause UC copyright and the 3-clause DFly copyright (all 
    appropriate copyrights are separately listed in the files), and to
    sunset the extra clause in a year so it doesn't create an issue down
    the line.

    I know there are some hard feelings between some of the principles.
    The hard feelings are something I cannot fix, I can only say that I did
    my best to accomodate everyone.  I sure as hell have no intention of
    ripping it out, it's primo code that fits the DragonFly model to a T,
    and we have already reaped *huge* benefits for GigE TCP streams out of
    it.  A 4-clause for 1 year is a small price to pay for such excellent

    In the ongoing Saga of all things BSD, this is a tiny blip, and not
    worth arguing over. 


    In anycase, I don't know why Bosko is bashing DR for it, DR has nothing
    to do with this particular issue.  And DragonFly as a project has
    adopted the 3-clause for its official copyright so it seems rather unfair
    to blast the whole project for what is basically a personal issue between
    two developers.  I will remind the FreeBSD principles that code borrowing
    is a cornerstone of open-source, and it should not be begrudged in any
    fashion... and it goes in all sorts of directions.  We wouldn't have
    a USB stack (you OR me) without NetBSD, don't you forget it!  Nor is 
    DragonFly in any way merely a copy of FreeBSD-5.  The greatest
    similarities will be in the device driver code, because we are doing
    our best to leverage the device work from FreeBSD-5 (and BUSDMA itself
    is not a FreeBSD-originated concept, I'll remind people).  But there are
    *MASSIVE* differences throughout the whole kernel, and that is not a 
    word that I use lightly. 


More information about the freebsd-current mailing list