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

[Rob Landley]

On 4/5/21 9:54 AM, Richard Fontana wrote:
> On Mon, Apr 5, 2021 at 9:22 AM Rob Landley <rob at landley.net> wrote:
>> Could someone please point me to where in the archives this issue was raised
>> again and voted on again to change the name back without notifying me the issue
>> was once again in dispute?
> 
> So, to be clear, what happened is that the heading "Zero-Clause BSD"
> apparently got changed to "Zero-Clause BSD / Free Public License
> 1.0.0". It's not clear when this happened. I left the OSI board at the
> end of March 2019 (and also ceased having access to edit the OSI
> website). I don't *think* I would have made this change without
> remembering it, nor can I think of why I would have wanted to make
> such a change. I had come to be firmly in support of having
> "Zero-Clause BSD" be the single name for the license in question.

Thank you. As I've mentioned, the name was chosen strategically.

People have actually grumbled about the success 0BSD has had where their
preferred public domain equivalent license doesn't get approval from the same
corporate legal departments:

https://github.com/github/choosealicense.com/issues/805

In fact, there's evidence that success of 0BSD has just recently started to act
as a precedent/gateway for use of CC0 and The Unlicense within Android:

https://android.googlesource.com/platform/external/toybox/+/5d35ee244cee3

>> If there wasn't a second vote changing the name again, and "0BSD" is still the
>> acknowledged name for it, could OSI please remove all mention of the no longer
>> relevant name from the 0BSD page? It does not need a "historical" mention
>> because it was not what the license was called when it was created and is not
>> what the license is called now. It does nothing but cause market confusion (Free
>> as in Free Software Foundation, on the GPL side of GPL-vs-BSD axis, it must be
>> REALLY viral), and apparently if we don't remove all of this tumor it metastasizes.
> 
> I would agree that there is no longer any need for a "historical"
> mention (in the form of that "Note") because as you say it "is not
> what the license is called now." Back when the name was changed, it
> was in my view appropriate to have the note explaining the earlier
> (from the OSI website perspective) name.
> 
> I'd still weakly support the OSI retaining the first sentence in the
> note that says "Despite its name, Zero-Clause BSD is an alteration of
> the ISC license, and is not textually derived from licenses in the BSD
> family". I guess I'm seeing it as equivalent to calling MIT No
> Attribution (https://opensource.org/licenses/MIT-0) "BSD No
> Attribution" -- surely a lot of people would be annoyed by that.

Not that I've noticed? I haven't found any software developers annoyed by the
provenance of 0BSD, and the only person approving or disapproving the use of
software in corporations that brought it up was you. (And that wasn't even in
the context of approving the use of software licensed under it within your
organization, it seemed like an aesthetic complaint?)

20+ years ago the dominant context was the "free software vs open source"
debates, and these days too many of the younguns seem to have thrown software
copyrights in the same bucket as software patents, both "Too Dumb To Live" to be
opted out of until the Boomers die:

https://speakerdeck.com/benbalter/open-source-licensing-by-the-numbers?slide=41


Which is a real pain for the rest of us trying to use their code in the
meantime. I've been trying to promote public domain equivalent licensing as a
palatable alternative to "not licensing one's code at all because only
capitalist old fogeys care about that". (Here's a 13 year old Ted talk about why
so many people under 30 hate the concept of copyright:
https://www.youtube.com/watch?v=7Q25-S7jzgs )

Outside the legal profession, "GPL vs BSD" seems to be the common axis of "The
Free Software Foundation's viral stuff vs everybody else".

https://timreview.ca/article/67
https://fosswire.com/post/2007/04/the-differences-between-the-gpl-lgpl-and-the-bsd/
https://www.quora.com/What-are-the-consumer-differences-between-GPL-and-BSD-licences
https://news.ycombinator.com/item?id=5187235
https://www.reddit.com/r/linuxquestions/comments/biy57v/why_would_any_software_use_a_bsd_license_over_a/
https://opensource.stackexchange.com/questions/6353/difference-between-gnu-gpl-and-bsd

The differences within the two groups (among multiple "GPL-like" licenses, and
among multiple "BSD-like" licenses) tend to get lost in the noise, especially
for software authors who don't want to be lawyers. You yourself said in your
2015 message to the SPDX list that you thought it was an MIT license, and only
later corrected that to ISC.

That grouping was the effect I was trying to take advantage of. In terms of
decision making, "4 clause, 3 clause, 2 clause, zero clause BSD" tends to get
rubber stamped. Despite other public domain equivalent licenses such as CC0 or
MIT-0 having significant pushback from the same departments (when proposed
independently).

Functionally, the minimally modified version is a public domain _equivalent_
license which is technically a different category from the entire cloud of
public domain _adjacent_ licenses, and that's why I snuggled it up to a herd of
zebras and painted stripes on it so people professionally looking gift horses in
the mouth would go "zebra, I know what that's like, ok then" and hopefully
rubber stamp it.

Unfortunately, during Richard Stallman's much-publicized return to the Free
Software Foundation, OSI has essentially renamed 0BSD back to "Free Public
License 1.0.0" on its website. Not just severing the "4/3/2/0 clause BSD"
grouping, but placing the license on the GPL side of GPL-vs-BSD, I.E. on the
Free Software side not the Open Source side. This is a direct attack on an
otherwise successful marketing strategy.

The name "Free Public License 1.0.0" managed to have THREE marketing problems:

1) The word Free: Richard Stallman's biography was "Free as in Freedom", kinda
strong association there.

2) The word Public: the first big viral GPL-alternative attempt was the Open
Public License https://spdx.org/licenses/OPL-1.0.html which was prominent enough
Linus Torvalds released software (https://lwn.net/Articles/205624/) under it.
That word means nothing to your average 20-something but has exactly the wrong
associations to people deep enough into the license world to know more than a
couple prominent names (I.E. potential license approval decision makers).

3) 1.0.0 says brand new and untested.

> I'll grant you that more people today would be bothered by using the
> "BSD" label for a license textually derived from the MIT license than
> by using the "BSD" label for a license textually derived from the ISC
> license, and that there's a slightly greater historical justification
> for the latter.

https://www.isc.org/about/ says "ISC was founded in 1994 as Internet Software
Consortium, Inc., to continue the work of maintaining and enhancing BIND
following in the footsteps of CSRG at U.C. Berkeley..." And BIND stands for
Berkeley Internet Name Daemon.

The history's actually fascinating: in 1969 DARPA deployed the original BBN IMPs
based on a Honeywell minicomputer hardened for submarine use (the main appeal of
which to them was the bomb-proof sealed metal enclosure, keeping students out of
the hardware), but 10 years later that hardware was declared obsolete and DARPA
targeted VAX hardware for the replacement systems. They wanted to use Berkeley's
Unix port to run the thing. Darpa's network stack pegged the CPU keeping up with
a 64 kilobit network connection and Bill Joy wrote a replacement that used 2% of
the CPU to do the same, and that's how Berkeley Unix became the backbone of the
internet. This article covers it in detail:

  https://www.salon.com/2000/05/16/chapter_2_part_one/

ISC only happened because Berkeley shut down the CSRG (and the BSDi partnership
with Novell disappeared into a cloud of litigation with AT&T especially after
they screwed over Bill Jolitz...)

In terms of political lineage, The ISC license is somewhere between a nephew and
the cadet branch of the BSD lineage. It was explicitly a reaction to it, and
intended as a replacement for it.

Then OpenBSD circled around and adopted ISC for _its_ packages (as their
suggested template license), and I adopted it from there while explicitly
looking for "a BSD license". I also asked Kirk McKusick's permission (the
second-ish maintainer of BSD back at Berkeley, and the most senior active BSD
developer in that community) and he approved the name.

That's why I called it Zero Clause BSD.

> And, ultimately, most people probably don't care and
> just want a convenient way to refer to the license.

Indeed. I generally don't have to explain to people what a BSD license is. I
would have to explain what ISC or MIT is to most people who don't already have
fairly broad knowledge of license alternatives.

It's one of those "the market has two dominant players and the rest are lost in
the noise" things. Coke, pepsi, everybody else. Android, iPhone, everybody else.
As I said at the start: there was explicit marketing strategy in my attempt to
make a public domain equivalent license palatable to a broader population, and
that strategy was working.

Rob