OpenSSF: Open Source Security Foundation
github.com
github.com
Highlights from the FAQ:
> OpenSSF is focused on improving the security of open source software (OSS) by building a broader community with targeted initiatives and best practices. It will start with a focus on metrics, tooling, best practices, developer identity validation and vulnerability disclosures best practices.
> OpenSSF will be supported by Linux Foundation membership dues with targeted organization contributions to support initiatives
> The founding members are GitHub, Google, IBM, JPMorgan Chase, Microsoft, NCC Group, OWASP Foundation and Red Hat, among others.
Google has project zero and Safe Browsing API which does helpful things to me, GitHub auto scans dependencies, and owasp guidelines are helpful I think
For example, the working group for vulnerability disclosure includes a lot of corporate players, and from what I can tell, not a single security researcher. Only one side of the disclosure process is represented in that working group.
Realizing how allergic major companies are to GPL code really creates some skepticism when they speak about embracing open source.
More importantly, the vulnerability disclosure group does not have a single contributor from the largest consumers of vulnerability disclosures: Linux distribution security teams.
And while I think they can go far improving the current situation, 'creating an API' isn't really the main pain point we are faced with when a legacy entity deals with handing out CVEs. Thus I'm unsure if a corporate-only working group is capable of getting a complete picture of problems regarding F/OSS vulnerability disclosure.
That should not be surprising, as they now all depend on OSS. I think this can be a good thing. The resources they bring can help OSS, and thus all its users and contributors.
> I feel this will be open source in the sense that it is being developed in the open, but not in the sense that they will foster an environment of community contribution.
I think it's reasonable to worry about that. But the Linux Foundation works very hard to foster contributions and comments from all. The Linux kernel, Kubernetes, and many other important OSS projects are in the LF. I think LF has a good track record.
Full disclosure: I just started to work for the Linux Foundation. But I would have said the same before.
If you or anyone else is interested in participating or learning more, please drop me a line at michael /dot/ scovetta {at} microsoft.com.
I once applied for a grant with the Core Infrastructure Initiative, the predecessor of the OpenSSF. It took me about half a day to write and submit a proposal, only to never hear back. After a couple of weeks, I tried to contact the Linux Foundation through their contact form and ask about my grant application, without success. Then I tried to contact some responsible people directly. I never got a single reply.
That's just shameful. Organizations like the Linux Foundation are antithetical to core idea of Open Source which is, well, openness.
But maybe Microsoft can bring some change into these encrusted institutions. BTW, I'm the maintainer of the XML parser and XSLT engine in your latest web browser. Bet you never even heard of me. There are quite a few security-related issues that still need work.
Also, the LF not only helps fund the development of Linux through Linus and other fellows, it hosts a plethora of other open source organizations, like LetsEncrypt which I'm sure you use on a daily basis without knowing.
Chris,
I think that the feedback on the CII grant process deserves to be acknowledged and addressed if possible. I wasn't directly part of the day-to-day efforts of CII, but I (as an individual providing one's opinion and advice) supported the original charter of directly funding security related development of critical software libraries that are often unnoticed like libxml2, OpenSSL, etc.
I am disappointed that CII didn't achieve this objective for more software (there were a few grants that were awarded and completed, but over time this seemed to end, from the public updates I read at the time). I think that it is still a good one to have.
The CII was before my time at the LF so I don't have much to share outside of it was in the mind of the OpenSSF founders to do better.
FYI: the majority of the work was born out of a closed source effort that you weren't involved in: https://github.blog/2020-07-09-what-we-learned-from-building...
While orthogonal to your main point, this sentence conflates Free Software (GPL) with Open Source. It should be emphasised that the GPL is NOT open source, and that Open Source is not Free Software.
https://www.gnu.org/philosophy/open-source-misses-the-point....
Ultimately Free Software and Open Source mean the same thing. They come from philosophically different viewpoints that they emphasize - free software tries to emphasize on freedom, while open source is more a technical approach and often comes from more business oriented people.
But in terms of what licenses actually qualify as either Open Source or Free Software - they don't differ. And the GPL absolutely is an Open Source license. The organization that made the term Open Source popular agrees: https://opensource.org/licenses/gpl-license
To reference the original point, Business may be allergic to the GPL, but at the same love the wider set of Open Source licenses.
Free Software, and Open Source very much do not mean the same thing. Free Software is a subset of open source with an addional layer of very restrictive clauses. It is that extra layer that prevents GPL code being used by most business.
While the difference is based in philosophy, yes, that philisophical difference has a huge impact in how code is used in the real world.
Free Software Foundation disagrees with this view: https://www.gnu.org/licenses/license-list.en.html
Perhaps the difference is most noticeable in libraries (where I do most of my work). Using a GPL library in a commercial app is generally impossible, where BSD (or LGPL) etc is completely acceptable.
In my experience companies are allergic to GPL specifically, whereas other licenses - especially MIT and BSD are far more palatable, and are frequently used.
Meanwhile, free software is really right to repair for software enforced via copyright instead of (dedicated) legislation. It threatens to reduce the power of the corporations, so they will be generally hostile to it unless their end products are the services that only they can build around it.
I'm aware of Google being AGPL-allergic, but are there companies out there that ban everything under the ordinary GPL? They'd have to avoid the Linux kernel, if they really meant it.
Take a look for instance at ETSI TC Cyber, or ETSI NFV Sec.
Even more available in specific domains, such as intelligent transport systems (ISG WG5)
Let's have one more standard promoting another agenda and set of priorities.
Open standards should also promote consolidated standards.
> Take a look for instance at ETSI TC Cyber, or ETSI NFV Sec.