[License-review] Please rename "Free Public License-1.0.0" to 0BSD.

[Richard Fontana]

On Tue, Oct 16, 2018 at 03:01:52PM -0500, Rob Landley wrote:

> When you were trying to convince SPDX to change its mind in 2015 (because you
> couldn't change yours), you said your official policy was to be compatible with
> SPDX. That implies you'd use what SPDX already says. I'm happy with that.

Let me attempt to address some of this. 

First, OSI and SPDX have very different purposes. You seem to think
that OSI decided in ~2005 that license proliferation was a bad thing
(true), stopped approving licenses (not true, though it probably
entered a long period of intentional resistance to new approvals), and
SPDX was formed to take up the slack (not at all true).

SPDX is not a license certification or license recognition
authority. It does not vet or approve licenses.

SPDX is first a standard for sharing data about software licenses (and
a working group for developing that standard). An important
side-activity (in some ways the most visible and unquestionably the
most successful aspect of SPDX so far) is its designation of standard
long and short identifiers for a set of a few hundred licenses. Those
licenses are a superset of the set of licenses that have been approved
by OSI. Inclusion of a license in the SPDX identifier list does not
signify approval. It signifies that someone asked the SPDX legal group
to add an identifier for a given license -- as you did with what ended
up being given SPDX short identifier 0BSD -- and the SPDX group
agreed.

Now, as for "official policy was to be compatible with SPDX": My
secondhand understanding is as follows: Sometime not too long before
mid-2013, representatives of the SPDX group and OSI appear to have
agreed to something like the following: (a) SPDX would make an effort
to distinguish OSI-approved licenses from other licenses in the
license identifier lists; (b) SPDX and OSI would coordinate such that
if OSI approved a new license, SPDX would be given notice so that it
could designate identifiers for the new license; and (c) OSI would
endeavor to promote the SPDX identifiers on the OSI website.

It is possible that (b) came later, and it is possible that (c) was
more specific than I am suggesting. (If Jilayne Lovejoy is on this
list she'd know.) There had been a partial effort to reform OSI
license page URLs so that the URL would make use of the SPDX short
identifier. 

When I got appointed to the OSI board and got somewhat interested in
license approvals and in the way licenses were organized and presented
on the OSI website, a few things became clear to me. First, the SPDX
"long identifiers" should generally not be promoted by OSI because in
many cases they deviate significantly from longstanding traditional
names by which particular licenses are or have been known. I'm not
sure OSI was actually promoting them. In any case the SPDX long names
see little to no real world use as far as I can tell, in contrast to
the short identifiers which have been remarkably successful [though
perhaps in a way that the SPDX group did not really intend].

Second, the attempt to use SPDX short identifiers in URLs was not
practical in some cases, because a "short" identifier might actually
be relatively lengthy and cumbersome compared to the actual or likely
common-reference name of the license, or might be typographically
problematic.

Anyway, there was no policy to be "compatible" with SPDX beyond more
or less what I've described above. 

Richard