[License-review] Request for OSI Approval: Tiwaz License, version 1.0 (New License)

Rob Landley rob at landley.net
Fri Jul 18 08:46:01 UTC 2025


On 5/28/25 15:20, Loncothad wrote:
> Opinions on the mentioned "flaws" come from open source community members I know who are concerned that Mozilla, as the steward of MPL 2.0, might exploit the "or later" clause (MPL 2.0 Section 10.2). Another community member developed a different approach to address this concern: the "Immutable License" (https://github.com/cull-os/carcass/blob/master/LICENSE.md). However, that license takes a completely different (more "distributed") approach, and its creator neither wishes to be a License Steward nor seeks OSI approval for it. This is where our views diverge, as I am seeking OSI approval for the Tiwaz License to establish it as a valuable tool for the wider Open Source community. 

What specifically are you trying to accomplish?

A dozen years back I did a public domain equivalent license:

https://en.wikipedia.org/wiki/Public-domain-equivalent_license

Which I got permission from Kirk McKusick to call a BSD license to ease 
adoption:

https://landley.net/toybox/license.html

At the time (2013) there wasn't an MIT-0 or John the Ripper license or 
similar, and Creative Commons was insisting their licenses (including 
CC-0) should not be applied to software, so I didn't think I was doing 
https://xkcd.com/927/ but even so somebody else came up with the same 
idea under a different name a year later and I had to argue about using 
a consistent name for it (because I'd run it through SPDX but not OSI). 
Hopefully that's still sorted.

Anyway, Android picked it up at the end of 2014 (when they started using 
toybox) and Samsung asked me to get it approved by SPDX and then I ran 
it through Github's approval process in 2018, and there's apparently 62 
thousand repositories on there (not including forks) using it now:

https://github.com/search?q=license%3A0bsd&type=Repositories&ref=advsearch&l=&l=

(Probably helps that it sorts alphabetically at the top of the list. :)

You can complain it's a copyright license that doesn't address patent 
law, contract law, trade secrets, that weird IP niche involving ASIC 
masks, DCMA circumvention exemptions, or anything except granting the 
right to use copyrighted material. But if you want simple public domain 
equivalent license instead of a complex legal thing: it's pretty simple.

> My goal with the Tiwaz License is to offer an OSI-approved option, within a familiar MPL-like structure, that provides this version immutability, thereby guarding against the potential "exploits" associated with "or later" clauses.

Anyway, my question is... what are you trying to accomplish with your 
license? What are your goals, why should people use _that_ one instead 
of an existing license? How are the users uniquely served by THIS license?

0BSD doesn't have an "or later" either, it's a simple permission grant, 
plus some warantee disclaimer boilerplate that's a really just there to 
make it look like other licenses. (Computer history is a hobby of mine 
and this is essentially a fossilized historical relic, as I explained 
back in https://landley.net/notes-2018.html#13-03-2018 .)

What do you need your new license for that an existing license doesn't 
do? If you expect many people to adopt your license and make it a new 
"standard", then why yours instead of one of the existing ones?

If you yourself are making your own instead of using an existing one, 
why should anyone ELSE use yours instead of making their own? I'm not 
saying there isn't a good answer to that question, but you have to 
articulate it as part of your license promotion strategy, if your goal 
is for people OTHER than you to use it.

(0BSD started started as the license _I_ needed for toybox, and that 
package still uses it. Other people using it really just made _me_ using 
it seem less weird. Unique licenses cause work for lawyers, leading to 
friction in adoption. Common licenses get approved once and then re-used 
by multiple packages. BSD already having 2, 3, and 4 clause versions let 
the zero clause version with an explicit analogy to CC-0 get rubber 
stamped by corporate legal departments a little easier, while not asking 
individual developers to engage significant legal expertise making a 
decision about what to use for their own projects. :)

> I have no information on whether those community members discussed this with an attorney. Personally, I've discussed licensing matters generally with a friend of mine who is a legal counsel, but software licensing is not his specific line of work.

I've studied intellectual property law on and off almost 30 years (back 
before even 
https://www.fool.com/archive/portfolios/rulemaker/2000/05/02/get-your-copyrights-here.aspx 
and https://www.tech-insider.org/linux/research/2005/1002-a.html and so 
on), spent ~3 years working on 
http://www.catb.org/~esr/halloween/halloween9.html related issues, 
talked to a bunch of lawyers like Cathy Raymond and Eben Moglen about IP 
licensing at quite some length, launched the first GPL enforcement suits 
back in 2006 (which ate far too much of my time for something like 4 
years and I most recently got deposed about fallout from all that 
nonsense last year, ala https://landley.net/notes-2024.html#24-06-2024)...

And I very much do NOT consider myself an expert on IP law. The license 
I did was an existing license with half a sentence removed, one 
explicitly chosen for simple straightforward wording and widespread use 
backed by a very large organization, which I still asked for multiple 
expert opinions on my change at the time.

(Lawyers never really say yes. They just refrain from giving a strong no.)

Rob


More information about the License-review mailing list