For Approval: ACE License

[John Cowan]

Forrest J. Cavalier III scripsit:

> I recommend withholding approval until all of the following are
> changed:

A meta-point: note that this license is actually in use for software
in distribution, and that the submitter has no authority to change it.
Rightly or wrongly, the fact that a license is actually being used
to distribute software (and not just proposed for such use) makes a
difference in how it appears to us.

> 1. The license includes a statement whiich I think is a restriction
>    on use....
>       You must, however,
>       include this copyright statement along with code built using DOC
>       software.

The BSD and other licenses already include such a clause.  It is not
a restriction on use of the software, and only a trivial restriction
on reuse.

> 2. The embedded references to URLs are a problem.  What happens when the
>    content at those URLs is replaced?  The license must stand on
>    its own before it can be approved.

The content at the URLs is not incorporated by reference or otherwise
part of the license, so it's irrelevant what it is.

> 3. I think all the text about being y2K compliant must be removed.

Well, this isn't a warranty, because all warranties are disclaimed,
so it's just a bare claim by the authors of the software, and as such
nonfunctional.  But it doesn't in any way contravene the OSD, any
more than the GPL's preamble does.

> 4. The license includes this statement, which seems to contradict the
>    possibility of commercial support, which is then claimed a sentence
>    or two later.

The first says that no support is provided with the software; the second,
that support is or may be available elsewhere.  No contradiction.  The
GPL makes similar statements.

>        Moreover, DOC software is provided with no support.
> 
>    The problem is that I need to get the downstream licensees to agree
>    to two contradictory things: a contract from me that says there
>    is support, and a license that says there isn't.

There is nothing in the license that says you must impose it on downstream
licensees.  You have to include the text of the license itself, but that
text is not applicable to your own software unless you make it so.
Nor is there anything resembling contract formation here: this is a
simple permissive license.

>    Adding one or two words to make it clear that there is no support
>    from WU, UC Irvine, etc would be fine, but that isn't what the
>    license says.

It's clear to me, certainly.

>    I think OSI should continue to withhold approval of licenses that
>    do not say what the licensors intended.  They should be forced to
>    clean up the language.

See above.

> 5. The patent and trade secret and liability indemnification paragraph
>    bothers me.   I don't know exactly why.  Is it legal?
> 
>    Maybe it bothers me because why should the license
>    identify a subset of contributors that are indemnified?

As I read it, it's unenforceable.  It is not for the infringer to claim
that he has no liability for his infringing acts, any more than a
trespasser can claim that he has no liability for his trespasses.
But it's harmless junk.  We have no remit to reject conformant licenses
because they contain harmless junk.

-- 
BALIN FUNDINUL          UZBAD KHAZADDUMU        cowan at ccil.org
BALIN SON OF FUNDIN     LORD OF KHAZAD-DUM      http://www.ccil.org/~cowan