[License-review] License drafting quality and process [was Re: Comment on MOSL and similar licenses]

Patrice-Emmanuel Schmitz pe.schmitz at googlemail.com
Thu Jun 6 10:34:08 UTC 2013

When I read Josh:
>I'm not convinced that there are any new "good" ideas in OSS
>licenses.  I take the extreme position that if we never approved another
>license, it would be a good thing.
It looks a little bit like if OSS licenses would escape to any progress
or Darwinian evolution as from date X (when?). If such biblical vision was
true, OSI should indeed focus on other subjects, bit I believe it is not,
However, it is obvious that the process of OSI license approval is
currently not the right one and should be improved:
- a more "formal" process, with a signed form including the full OSD,
- a short list of "usual licenses" to read by the requester before any
submission (not the 60 OSI-approved licenses, but maybe a selection of
about 10 (representing permissive, moderated copyleft and copyleft models),
- a formal declaration that the new proposed license is not in
contradiction with the above OSD,
- a mandatory indication of which "usual license(s)" is(are) the most
similar to the proposed new license,
- the proposed TXT,
- a mandatory explanation about differences between the above similar usual
license and the new one (why the need for a new licence),
- rationale about why the "issue" cannot be solved by adding a simple
additional agreement, compatible with the most similar licence,
- the (perhaps non mandatory, but strongly recommended) name of the lawyer
who has reviewed the above submission.

Another suggestion would be to deliver more rapidly a simple "certificate
of OSD compliance" which would be a first step. OSI approval would be given
only as a second step, i.e. once the steward bring the evidence (names +
URLs) that a minimum number of projects (100, 250, 500?) are published
under his license.  This first step would put an end to the current
frustration of refusing any new license. The advantage of implementing two
classes of licenses (first class with more of X projects and second class
with a simple certificate) would be the possibility to clean up the current
OSI-approved license list, moving some of the less used or old licenses in
a second basket.

Just my two cents...

2013/6/5 Josh Berkus <josh at postgresql.org>

> Clark,
> > What if the "discussion" part of the process were mandatory  -- the
> > first step would be a "need" validation on license-discuss by posting a
> > synopsis that outlines the "real world" licensing problem being
> > addressed and why other licenses are inadequate.
> > The discussion could then be focused on validating if this "need" is
> > consistent with the vision of the OSI.   If so, then the participant
> > could be *invited* to submit their license design and perhaps even
> > official legally prepared license text to license-review.
> I like this general idea, actually.  I think there would be significant
> non-proliferation benefit to having a multi-stage approval process for
> new licenses, an important part of which could be license-discuss.  If
> nothing else, the time that this process takes would discourage "crayon"
> licenses.
> Everyone: note that Clark is talking about discussion on
> license-discuss, NOT here.
> I'm not sure that his should be explained in terms of "needs" though,
> which is not part of the OSI charter.  Rather, it should be explained in
> terms of:
> (1) does the license fit the OSD?
> (2) is it sufficiently different from all existing licenses?
> The desires of the requestor in creating the license would come out in
> discussion of (2).
> > I'm reluctant to shut down good
> > ideas, just because the submitter doesn't have $10k spare cash to throw
> > at an attorney.  In other words, asking for varnish isn't going to help
> > when the wood is rotten.
> I'll take a couple exceptions to that:
> First, *at some point* the new license will need legal review by
> *somebody*.  We should not be approving licenses which no lawyer has
> ever taken a hard look at.  So either the legal review happens before
> submission, or it happens afterwards.
> Second, I take the tack that if someone *does* have the resources for
> professional legal staff to handle licensing, we *have* to handle them
> here because we can't ignore them.  They've demonstrated the resources,
> organization and resolve that they might go ahead and deploy their
> license whether or not we approve it, so we have a limited window of
> time to talk them into modifying it or (better) using an existing
> license instead.
> Third, I'm not convinced that there are any new "good" ideas in OSS
> licenses.  I take the extreme position that if we never approved another
> license, it would be a good thing.
> --Josh Berkus
> _______________________________________________
> License-review mailing list
> License-review at opensource.org
> http://projects.opensource.org/cgi-bin/mailman/listinfo/license-review

Patrice-Emmanuel Schmitz
Legal adviser
pe.schmitz at googlemail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-review_lists.opensource.org/attachments/20130606/22eedea8/attachment.html>

More information about the License-review mailing list