The Invisible Hand

Matthew C. Weigel weigel+ at
Thu Oct 4 22:43:31 UTC 2001

On Thu, 4 Oct 2001, Russell Nelson wrote:

> But you *did* say "free software licenses according to the FSF".  But
> that's precisely what I was objecting to -- your implication that the
> FSF defines "free software" and that nobody else's opinion matters.

Please stop, and think about what you're saying for a minute.  If I had
said *just* free software, and used the FSF definition, I would be
implying that only the FSF's definition matters.  However, I qualified
it with "as according to the FSF" to clarify that not everyone agrees.

> Look at the GPL's requirement for three-year's distribution of the
> source code.  That's a constraint, and a cost.  Doesn't stop you from
> distributing any code, but it makes it less convenient.  Oh well.
> Neither does the APSL's publishing requirement.

It materially affects what you can and can not do: you can not develop
a derivative work that uses anything you are prohibited from
publishing, *whether you distribute it or not.*

> Again, we are not the FSF.  We don't have a charismatic founder whose
> word is law.

What started your entire response to me was my statement that RMS is
more open to criticism than the OSI.  So, charismatic or not, it
strikes me as *his* word being much more openn to discussion.

> I'll accuse you of any damn thing I feel you're guilty of.  And if I'm
> wrong, I'll apologize for it, too.

But you'll ignore any argument that you're wrong.  Since you've
continued to accuse me, let me defend myself.

I have been on this list from the beginning, and I have tried to do
whatever an outsider to the OSI can be expected to do.  At the time I
made that claim, the OSI had studiously avoided all messages on the
list asking, "what's going on?  does the OSI need help?"  It is no
accident that I referred to the refusal of the OSI to communicate as
the deepest problem.

> But in *general* in the open source / free software / whatever world,
> contributions are gladly entertained, suggestions carefully

Yup, and the OSI is such a sterling example of collaboration in action,
that offers of help, requests for direction, *everything* is ignored.

> Are you ever going to evaluate the three current licenses or are you
> going to wh*n* some more?

Ah yes, the three (wait, no two) licenses you claimed were submitted on
8/31.  Plus a third that has been discussed voluminously already.


For readability,

What legally constitutes portability fixes?  Bug fixes?  Is it a bug if
it reports your software as not compliant?

Clause 4 discourages sharing and using the code as a starting point, in
much the same way requiring pristine source + patches.  In a similar
way, I think it's OK but should be discouraged.

I think that this license is conceptually opposed to open source, being
that it is concerned with test suite software.  It jumps through a few
hoops to qualify as open source anyways, and it is interesting to note
the way in which it circumvents requiring that only 'standard' binaries
be distributed.

It probably qualifies as open source, but the authors should consider
their motivation for wanting the test suite to be open source.


(this one's HTML is fine, you might want to compare the differences in
the submissions and encourage submitters to follow the MOSL's method;
looking at the archive it's not obvious, but you presumably have the
messages themselves sitting around)

This license requires a rewrite for clarity, so that there aren't two
different 3(a)'s, 3(b)'s, and so that the 4(e) describing additional
rights you may choose to offer licensees actually exists (if it does
exist, I couldn't find it).  My recommendation would be, at least as a
first (new) draft, to do it in something like HTML that numbers them on
the fly.

Others have brought up the clause requiring Dutch Law, I'll defer to
other people on that.

The "4. U.S. Government End Users." clause also bothers me, I am not
comfortable with claims of fact like that within the license.


Oh yes, I've not made a peep about any of these licenses.  Oh wait, I
*have* discussed this license before.  It is clerarly open source, does
nothing interesting, but should be approved because it qualifies and as
a sign of good will.
 Matthew Weigel
 Research Systems Programmer
 mcweigel at ne weigel at

license-discuss archive is at

More information about the License-discuss mailing list