Subject: Re: [shapj@us.ibm.com: Re: EROS license]
From: Ian Lance Taylor <ian@airs.com>
Date: 26 Jun 1999 13:03:52 -0400

   From: shapj@us.ibm.com
   Date: Sat, 26 Jun 1999 12:33:06 -0400

   >You said that Cygnus made a binary only release.  That is what I was
   >disputing.  Did you actually get that software directly from Cygnus?

   Hmm.  If I said that I must have written badly.  The releases were indeed source
   releases, and they did indeed come to us direct from Cygnus.

You said this:

    I also am inclined to permit binary-only distribution during beta/testing
    phases, provided that this is time limited.  Companies like Cygnus has done this
    on several occasions (e.g. where there is an NDA on the hardware instruction
    set).  This seems to me to be a violation of the letter of the GPL, but I think
    that it's a good thing.  We certainly don't want terms that delay the beginning
    of a new compiler until the day the chip releases, when everyone else is already
    shipping theirs.

I hope you see why we Cygnites and ex-Cygnites were surprised at your
claim.

   Michael misrecalls, though.  We definitely received source releases that could
   only be given to us because we were under NDA.  This was prior to the
   publication of the spec or the distribution of any compiler by Sun.  I was
   *writing* part of the spec at the time.

I don't question that you were under NDA to Sun, and I don't question
that Sun only permitted Cygnus to distribute the code to you because
you signed the NDA.  My point is that Cygnus did not make a binary
only release, Cygnus did not restrict you from any further
redistribution, and Cygnus did not violate either the letter or the
spirit of the GPL.  Remember that the GPL does not require that
sources be distributed to everybody; it only requires that sources be
distributed with binaries and with no additional restrictions.  Sun
may have restricted you from further redistribution, and I agree that
if so they violated the terms of the GPL, but Cygnus did not.

Ian