Subject: Re: EROS license
From: Ian Lance Taylor <>
Date: 27 Jun 1999 22:48:41 -0400

   From: Russell Nelson <>
   Date: Sun, 27 Jun 1999 22:21:32 -0400 (EDT)

   DJ Delorie writes:
    > > I'm trying to convince Ian (and presumably others) that they benefit
    > > from encouraging authors to give up most of their proprietary rights
    > > in return for the ability to dual-license.
    > Except that Ian is describing the contributor's point of view, where
    > giving up rights to the original author doesn't usually benefit the
    > contributor (except, perhaps, very indirectly).

   Except that it may be the difference between the software being free
   or being proprietary.

   "May".  That's the key.  Ian is acting as if "may" is "may never".

No, I think DJ is on the right track.  There are different interests

1) License holder (typically the original author)
2) Contributor
3) User

In these terms, I think you are saying that the license holder gets
more benefit from a dual license than a non-dual license.  I think we
all agree that the user gets more benefit if the software is free
rather than proprietary.  I think you are saying that since the
benefit of a dual-license may be enough to tip the license holder into
using a free license, we should accept and encourage dual licenses,
because in the end we get more free software (or, as I might prefer to
say, open source software, but I suppose I should try to avoid
starting that argument again).

I can understand and accept all of that.  However, I add a premise
(which you may also believe): a straight free license is better for
the user, and better for the contributor, than a dual license.
Therefore, I believe that when it comes to contributing to a project,
given a wide choice of projects to work on, it's better to encourage
the projects which use a straight free license.

I hope nobody thinks that I think that a dual license is somehow evil.
I don't think it is.  I merely think a straight free license is
better.  Given limited time and a vast array of possible projects, I
allocate my time accordingly.

    > Can we match this up with real life experience?

   Well, sure.  Mozilla.

Are you counting that as a success or a failure?

Personally, despite Jamie Zawinski's comments, I consider it to be a
success, but I think it could have been a much larger success with a
different initial code base, and a somewhat larger success with a
different license.