I agree we should err on the inclusive side. In concept, I think the driver 
should be popularity more than OSS definition. It’s better for users to include 
commonly used open source-like license. JSON and WTFPL (maybe this complies, 
but it’s not on the OSI list anyway), for example. We do need to be mindful of 
the capacity of the team and it makes sense to prioritize compliant licenses, 
but at a high-level the LL is to minimize the need to locally define license in 
SPDX docs and so should cover the licenses that users are likely to run into.
Phil

From: "spdx-legal@lists.spdx.org" <Spdx-legal@lists.spdx.org> on behalf of 
"mdo...@linuxfoundation.org" <mdo...@linuxfoundation.org>
Date: Thursday, May 30, 2019 at 3:44 PM
To: Jilayne Lovejoy <opensou...@jilayne.com>
Cc: "spdx-legal@lists.spdx.org" <Spdx-legal@lists.spdx.org>
Subject: Re: meeting minutes from today



On Thu, May 30, 2019 at 3:23 PM J Lovejoy 
<opensou...@jilayne.com<mailto:opensou...@jilayne.com>> wrote:

3) Need more feedback on documentation updates - see email sent earlier this 
week, comment on PRs in Github
·         discussed licenses that aren't squarely open source and variations on 
how far fall out and how to deal with this potential slippery slope. Some are 
clearly not open source (use, technology restrictions), then there are licenses 
intended to be open source, but impose new copyleft-type conditions (SSPL, 
Parity). We don't want to get into the business of defining what is open 
source, but we have to make a call on these licenses - how to do this without 
getting into the middle of the debate?
_._,_.

I tend to think SPDX should error on being more inclusive as we're expecting 
tool providers, end users and others scanning code to recognize and match the 
licenses. If they're not listed in SPDX, they have to merge in another list. I 
completely agree the current definitions are less inclusive for good reasons, 
but re-defining  "what is open source and others" for an SPDX license list is 
not going to be helpful either. That said, I look at being OSI or FSF approved 
as "properties" of the license. They're a feature. Some will consider them 
important features, but some just need to match the text in a header to some 
list of licenses that knows what this text in the header is. And for that we 
can get functionally practical. My suggestion for defining the rules for what 
gets in or doesn't would follow a similar logic.

SPDX license list may include:

  *   Licenses that show up in public source code repositories,
  *   for source code that is used by a project community directly or as a 
dependency,
  *   and have more than 1 known user of the project codebase
SPDX license list does not include:

  *   commercial licenses for use by a company's licensees
  *   vanity licenses adding a proper noun, minor modification or non-licensing 
elements (e.g. Bitcoin donation address) to a known license
  *   licenses that have no known users
  *   licenses with no known integration or dependency with other codebases
Maybe I'll start a huge debate with this, but if I were to draw a line in the 
sand without any input from others, this is where I'd probably start. So it 
likely needs input from others - let's refine it or throw it out for something 
someone has that's better. I have no emotional or intellectual attachment to 
this line in the sand. :-)



-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#2601): https://lists.spdx.org/g/Spdx-legal/message/2601
Mute This Topic: https://lists.spdx.org/mt/31872337/21656
Group Owner: spdx-legal+ow...@lists.spdx.org
Unsubscribe: https://lists.spdx.org/g/Spdx-legal/unsub  
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to